Upgrade

00:00:00 ◼   ► (upbeat music)

00:00:02 ◼   ► From Relay FM, this is Upgrade, episode 436.

00:00:13 ◼   ► Today's show is brought to you by StoryWorth,

00:00:16 ◼   ► Capital One, and Ooni Pizza Ovens.

00:00:19 ◼   ► Here is your host, Jason Snell.

00:00:22 ◼   ► - Thank you, Myke.

00:00:23 ◼   ► Hello, everybody.

00:00:24 ◼   ► We are back for another Myke-less, almost,

00:00:26 ◼   ► episode of Upgrade.

00:00:27 ◼   ► I am Jason Snell and I am hosting this episode with a very special guest

00:00:32 ◼   ► Returning to upgrade for the I actually have lost count half a dozen or so time. It is John, Syracuse. Oh, hi John

00:00:40 ◼   ► When you say mic lists, we both have mics in front of us, but that's okay, right? You didn't hear the why?

00:00:45 ◼   ► No

00:00:47 ◼   ► It's like tire with a why?

00:00:49 ◼   ► It is I just put that together. Yeah, but not not choir because that has a qu you

00:00:57 ◼   ► - Mm-hmm.

00:00:58 ◼   ► - It's the, we, that's a different pod,

00:01:00 ◼   ► that's not "Robot or Not" that we're doing now.

00:01:02 ◼   ► That's a totally different podcast that we're doing

00:01:05 ◼   ► about weird English spelling and things.

00:01:08 ◼   ► Anyway, I just, for this trip,

00:01:10 ◼   ► I'm just replacing Myke with a series of guys named John.

00:01:14 ◼   ► - Mm-hmm, as you do.

00:01:16 ◼   ► There's so many of us, why not?

00:01:17 ◼   ► Why not take advantage of the bounty?

00:01:19 ◼   ► - I bought the six pack at Costco,

00:01:21 ◼   ► and you know, you gotta use them all.

00:01:24 ◼   ► But we start off with the #snelltalk question,

00:01:28 ◼   ► which I'm also gonna apply to my guests,

00:01:31 ◼   ► which I have again posed myself because I wanna know,

00:01:36 ◼   ► John, what is your Christmas shopping strategy?

00:01:39 ◼   ► - Hmm, I mean, I have to say I have never been good

00:01:45 ◼   ► at gift giving or buying.

00:01:51 ◼   ► My main skill lies in the years that I spent honing my gift asking for abilities as a child.

00:01:59 ◼   ► That seemed very important for me to develop as a kid.

00:02:02 ◼   ► So I feel like I'm good at that. I've always been good at making lists of presents that I want,

00:02:09 ◼   ► and I've been good at figuring out how to get my parents to give them to me.

00:02:13 ◼   ► And those skills stopped being useful once I left the house.

00:02:17 ◼   ► And now I'm left with nothing. I not don't do a good job of buying people things. I don't know what to get people things

00:02:24 ◼   ► It's a problem. So I mean my strategy such as it is is all throughout the year

00:02:29 ◼   ► I have a notes and you know a document in Apple Notes called gift ideas and

00:02:33 ◼   ► Anytime there is any inkling of any kind of gift that anybody might like I write

00:02:39 ◼   ► the gift items a little head headings for people's names and then there's like little bulleted lists of gift ideas for them and

00:02:45 ◼   ► And then when the holidays come, my strategy is I pull up that document and I hope that there will be some

00:02:49 ◼   ► headings underneath people's names that aren't already checked off.

00:02:52 ◼   ► And if they are already checked off, then I go to the people and say, "What do you want? Give me a list.

00:02:58 ◼   ► I need a list. I can't think of anything."

00:03:00 ◼   ► Yeah, that's my strategy. It involves a lot of whining and occasionally looking at another's document.

00:03:06 ◼   ► This is very much

00:03:08 ◼   ► my strategy too, and my story too. I'm very bad at this. I need people's help.

00:03:14 ◼   ► help. Fortunately, having been married more than 25 years, you've got—we've gotten to

00:03:23 ◼   ► the point where our strategy is, for the last three or four months of the year, if there's

00:03:29 ◼   ► something that you would like to buy yourself, don't do it and tell your spouse instead,

00:03:35 ◼   ► right? Like, that's the—and listeners to this podcast heard me do that a few weeks

00:03:40 ◼   ► ago where Myke was talking about the slippers that he liked. And I very much had to say,

00:03:48 ◼   ► "No, I'm not gonna go buy them. I'm going to mention to Lauren that this would be a

00:03:53 ◼   ► thing that she could get for me." And by mentioned, do you mean send her the actual link so you

00:03:58 ◼   ► make sure you get the exact ones that you were looking at? Yeah, exactly. Literally

00:04:03 ◼   ► sending her a text saying, "Oh, here's a thing you could get me." And she does the same for

00:04:06 ◼   ► me. Just exact hyperlinks and you have to know if you link to the site and it doesn't

00:04:10 ◼   ► remember the size that you selected,

00:04:11 ◼   ► you have to put down the size and the color and.

00:04:13 ◼   ► - Yeah, and if you're thinking,

00:04:15 ◼   ► well, also you can kind of artfully,

00:04:16 ◼   ► like I was like, I don't know which color to pick.

00:04:18 ◼   ► I'm like, you know what?

00:04:19 ◼   ► I'm not gonna specify color.

00:04:20 ◼   ► We're just gonna see what happens.

00:04:21 ◼   ► And in fact, more broadly, if you're thinking,

00:04:24 ◼   ► this is no fun because you're never gonna be surprised.

00:04:27 ◼   ► Friends, let me tell you, if you do enough of them

00:04:29 ◼   ► and you do them for long enough in advance,

00:04:32 ◼   ► you've forgotten everything that you said to your partner.

00:04:34 ◼   ► And so then it's like, oh yeah, I did want this.

00:04:37 ◼   ► And it's actually kind of nice.

00:04:39 ◼   ► I think for the past, I don't know, three, five years,

00:04:42 ◼   ► I don't think I've made it through a Christmas

00:04:44 ◼   ► without realizing that there's something I purchased

00:04:46 ◼   ► probably for my wife that I forgot to wrap and give to her

00:04:49 ◼   ► that I find in a hiding place months later.

00:04:51 ◼   ► - Yes. - Go, oh,

00:04:52 ◼   ► I forgot that I got you this for Christmas

00:04:54 ◼   ► and I forgot to wrap,

00:04:55 ◼   ► because I have them all squirreled away all over the place.

00:04:56 ◼   ► And then you totally, especially if you buy it months

00:04:58 ◼   ► in advance, I totally forget about it.

00:05:00 ◼   ► - Yeah, I bought a piece of art,

00:05:02 ◼   ► an art print for Lauren and had it in a tube.

00:05:08 ◼   ► and then like six months, like, cause it's, it came in a tube and I just put the shipping

00:05:12 ◼   ► tube like in one of my drawers and then like six months later I'm going through and like,

00:05:16 ◼   ► what's in this tube? And I opened it up and I was like, oh, oh, that was a Christmas present.

00:05:22 ◼   ► It's a real surprise. It's a surprise, Christmas present in July.

00:05:25 ◼   ► And I appreciate your idea that you have a notes document. I, I have a reminders list

00:05:32 ◼   ► called end of year gift ideas.

00:05:35 ◼   ► That is where I put all the ideas, not only for my family,

00:05:40 ◼   ► but also I have the, you know,

00:05:43 ◼   ► do something nice for the incomparable hosts

00:05:47 ◼   ► and what's the item gonna be,

00:05:50 ◼   ► the annual item gonna be for the incomparable members

00:05:52 ◼   ► who get like something in the mail.

00:05:54 ◼   ► I have them all in there.

00:05:56 ◼   ► So that is my, they're all mixed in together.

00:05:59 ◼   ► 'cause it really, I just, as with so much in reminders,

00:06:03 ◼   ► it's literally like, I just wanna get this idea down

00:06:06 ◼   ► because I know that in two hours, I'm gonna be like,

00:06:09 ◼   ► hey, I had a good idea, what was it?

00:06:12 ◼   ► And it's not gonna come to me.

00:06:14 ◼   ► - I think you can get rid of the end of the year

00:06:16 ◼   ► or end of year prefix there and just call it gift ideas

00:06:18 ◼   ► and change the icon from a snowflake, ha ha,

00:06:20 ◼   ► to a box or something.

00:06:23 ◼   ► - Can I do that? - Because it's like,

00:06:24 ◼   ► what about like, you know, just random gifts?

00:06:26 ◼   ► Like I have gift ideas for people

00:06:28 ◼   ► who I've never purchased a gift for,

00:06:30 ◼   ► but I feel like if I had ever had to purchase a gift

00:06:32 ◼   ► for them or wanted to, this is what I would get them.

00:06:34 ◼   ► - Wonder, could I do a box?

00:06:37 ◼   ► I mean, there's squares, but I don't know.

00:06:38 ◼   ► - Also consider title case.

00:06:40 ◼   ► - There is a box.

00:06:41 ◼   ► Well, yeah, you can see that I was just desperate

00:06:43 ◼   ► to get to the list when I decided this.

00:06:45 ◼   ► Okay, Jon, I have taken your advice.

00:06:52 ◼   ► - And you've capitalized, wow, look at that.

00:06:55 ◼   ► You're like a professional now.

00:06:56 ◼   ► There's a box and it says gift ideas.

00:06:58 ◼   ► And the item is gone down now that you purchased the gift for me.

00:07:01 ◼   ► Actually, I discovered that a gift that I gave Lauren for our anniversary was on that

00:07:04 ◼   ► list so I checked it off.

00:07:06 ◼   ► Done.

00:07:07 ◼   ► John, as I can tell from the sound of your dog barking in the background, I wanted to

00:07:11 ◼   ► do a little follow out to ATP, generally, because you talk about your life and your

00:07:16 ◼   ► house on that podcast, the Accidental Tech Podcast.

00:07:18 ◼   ► Also it's direct gifts as well.

00:07:20 ◼   ► No the.

00:07:21 ◼   ► No the?

00:07:22 ◼   ► Mm-hmm.

00:07:23 ◼   ► - Mm-hmm.

00:07:24 ◼   ► - The re, did I say the rectiffs?

00:07:28 ◼   ► - No, you said the accidental tech podcast.

00:07:29 ◼   ► - Oh, the accidental tech podcast.

00:07:30 ◼   ► I was, well, it's a, the, the, and then the-

00:07:32 ◼   ► - So that would be T-A-T-B.

00:07:34 ◼   ► - The quote, no, no, the quote mark comes after the the.

00:07:37 ◼   ► It's the quote accidental tech podcast,

00:07:39 ◼   ► if that is its real name.

00:07:42 ◼   ► Although I do kind of want to call it the rectiffs now.

00:07:44 ◼   ► I think that's kind of fun.

00:07:45 ◼   ► You hear the rectiffs?

00:07:47 ◼   ► I don't know about that Merlin.

00:07:49 ◼   ► It's like things that my mom would say

00:07:52 ◼   ► She listened to podcasts.

00:07:54 ◼   ► - I know, Bez listens to rec-diffs and she does say that.

00:07:57 ◼   ► - Ah, I went to your house.

00:08:00 ◼   ► I was in your house last week.

00:08:03 ◼   ► - You were.

00:08:03 ◼   ► - You were there too, I didn't like break in.

00:08:06 ◼   ► - Mm-hmm.

00:08:07 ◼   ► - My impression of it was that it was much nicer

00:08:10 ◼   ► than I thought, but that's only because I only knew

00:08:13 ◼   ► about it from listening to the complaints

00:08:15 ◼   ► of the person who lives there.

00:08:17 ◼   ► - No, and also it's like Disneyland,

00:08:18 ◼   ► like we control the sight lines.

00:08:20 ◼   ► - That's true, that's true.

00:08:21 ◼   ► I didn't go upstairs.

00:08:23 ◼   ► - You didn't, so you're sitting on the couch,

00:08:25 ◼   ► but above your head was the majority of the peeling paint.

00:08:28 ◼   ► And then where you see in your eye line

00:08:29 ◼   ► was the lesser peeling paint.

00:08:31 ◼   ► - There is some peeling paint in the corner of the ceiling

00:08:32 ◼   ► and off to my right in that corner.

00:08:35 ◼   ► - There's way more than that.

00:08:37 ◼   ► It's everywhere.

00:08:37 ◼   ► It's everywhere slightly out of your eye line.

00:08:39 ◼   ► - Well, John, honestly,

00:08:40 ◼   ► after I complimented your house for being nice,

00:08:43 ◼   ► the last thing I wanted to do

00:08:44 ◼   ► is start pointing out the flaws.

00:08:45 ◼   ► I thought you're aware.

00:08:46 ◼   ► - Yeah, I know.

00:08:46 ◼   ► Then you start seeing them.

00:08:47 ◼   ► (laughing)

00:08:49 ◼   ► I mean, I could have pointed them all out to you,

00:08:50 ◼   ► but we would have been all right.

00:08:51 ◼   ► the satisfaction of that, quite frankly.

00:08:53 ◼   ► But I do, I do, we had Chinese food, we recorded a podcast,

00:08:59 ◼   ► it's gonna be this week's episode of The Incomparable,

00:09:02 ◼   ► we talk about Andor, the Star Wars show.

00:09:05 ◼   ► Daisy, your dog, barked a lot, a lot, a lot, a lot.

00:09:10 ◼   ► Her house was invaded by people.

00:09:13 ◼   ► - I mean, I wouldn't say it's a lot.

00:09:15 ◼   ► She barked a little bit, and then she was basically menacing

00:09:19 ◼   ► for the rest of the time.

00:09:20 ◼   ► - She did menace some people,

00:09:24 ◼   ► most specifically Brian Hamilton,

00:09:26 ◼   ► but she did menace some people, it's true.

00:09:31 ◼   ► - She smells fear, and when she smells fear,

00:09:33 ◼   ► it makes her afraid.

00:09:34 ◼   ► - Interesting, and then she's even more on edge.

00:09:38 ◼   ► Yeah, she actually was really good with me,

00:09:40 ◼   ► but like I have a dog. - 'Cause you're a dog person.

00:09:43 ◼   ► - I've had dogs, and I understand dogs,

00:09:46 ◼   ► So I was not afraid of her and I got her,

00:09:50 ◼   ► I kind of got her in that way.

00:09:51 ◼   ► Although it did strike me that having a dog,

00:09:53 ◼   ► I mean, I sort of understood it conceptually

00:09:56 ◼   ► that you have to deal with the dog for podcasting reasons.

00:09:59 ◼   ► But it was only as I sat there and heard her bark

00:10:03 ◼   ► that I thought, oh, this would actually be a big problem

00:10:05 ◼   ► for podcasting.

00:10:07 ◼   ► - And now you get to hear it on your very own podcast.

00:10:09 ◼   ► - It's good, I mean, it's topical.

00:10:11 ◼   ► - It's my fault.

00:10:12 ◼   ► Next time I'll say,

00:10:13 ◼   ► because you start the podcast at 1 p.m.

00:10:15 ◼   ► and then you don't have to deal with this.

00:10:16 ◼   ► 'Cause she goes off to her doggy play date.

00:10:18 ◼   ► - Okay, we would have had to make all the people

00:10:19 ◼   ► who listen to live to upgrade, which you can do.

00:10:21 ◼   ► - Oh yeah, right, well anyway, enjoy.

00:10:23 ◼   ► - Pacific, Eastern.

00:10:24 ◼   ► But she did, I didn't, I think I may have poked my head

00:10:28 ◼   ► into your office, but basically I didn't see.

00:10:31 ◼   ► I didn't look around in there, I didn't see.

00:10:32 ◼   ► - You should have, I cleaned it for everybody.

00:10:34 ◼   ► - I appreciate it, it was very clean.

00:10:37 ◼   ► Very, very nice, very nice house.

00:10:38 ◼   ► It's got all the holiday decorations wore up.

00:10:41 ◼   ► You too have a picture from your wedding

00:10:43 ◼   ► where you and Tina look like kids playing dress up,

00:10:46 ◼   ► which is what my wedding picture with Lauren looks like.

00:10:49 ◼   ► It's exactly the same picture.

00:10:50 ◼   ► It's nice.

00:10:52 ◼   ► But, and I didn't, I honestly, as we were walking up,

00:10:56 ◼   ► I was like, I know so many stories

00:10:57 ◼   ► about things that happen in this house.

00:10:59 ◼   ► It was a little weird.

00:11:00 ◼   ► It was almost like going to like a movie set or something.

00:11:03 ◼   ► It's like, I know way.

00:11:04 ◼   ► - You could have gone into my kitchen

00:11:06 ◼   ► and looked at how level my refrigerator doors are.

00:11:10 ◼   ► Like you could have checked out the toaster.

00:11:11 ◼   ► So much you could have done.

00:11:12 ◼   ► - There is, you know, I didn't look at the toaster at all.

00:11:15 ◼   ► However, I did take a glass of water out of the tap

00:11:21 ◼   ► and then thought it had a funny taste,

00:11:23 ◼   ► dumped it out and-

00:11:25 ◼   ► - You are insane on tap water, it's awesome.

00:11:28 ◼   ► - Went to your filtered water in the Brita pitcher instead.

00:11:32 ◼   ► - Which tastes like Brita filter.

00:11:34 ◼   ► - I felt it had no taste, but your water had a taste to it.

00:11:38 ◼   ► There was something there.

00:11:39 ◼   ► I never take tap water and then dump it out.

00:11:41 ◼   ► I never do that.

00:11:42 ◼   ► And I was like, you know, I bet if the tap water tastes

00:11:45 ◼   ► like this, I bet there's a Brita pitcher in the,

00:11:47 ◼   ► well, first I thought, I bet, you know,

00:11:49 ◼   ► I could get filtered water from the refrigerator.

00:11:52 ◼   ► And then of course your refrigerator's hooked up to water.

00:11:54 ◼   ► - I didn't even know that thing was there.

00:11:55 ◼   ► I totally forgotten about it.

00:11:57 ◼   ► - And so then I pressed the button.

00:11:59 ◼   ► I was like, oh, right.

00:12:00 ◼   ► So then I saw that there was a Brita pitcher

00:12:02 ◼   ► and I was like, well, great.

00:12:03 ◼   ► I would have otherwise had the tap water,

00:12:04 ◼   ► but I dumped the tap water.

00:12:06 ◼   ► - Maybe you didn't let it run long enough.

00:12:07 ◼   ► That might've been the thing too.

00:12:08 ◼   ► - I did, we did have that conversation

00:12:10 ◼   ► sometimes they tell you you gotta run the tap a little bit. So everybody else, see,

00:12:13 ◼   ► yes, because everybody else is like, "Oh, the water is fine," but maybe I took the

00:12:16 ◼   ► bullet there and I had the bad water at the beginning and then it smoothed out.

00:12:20 ◼   ► Yeah, now in terms of tasting water, Long Island water, best tasting, where I came from

00:12:24 ◼   ► on Long Island. It's right from underneath the island, the sand just filters it all out.

00:12:29 ◼   ► Filters are great.

00:12:30 ◼   ► That's number one. And I think number two is probably maybe New York City water from

00:12:37 ◼   ► the upstate reservoirs and number three is Massachusetts.

00:12:39 ◼   ► And then a distant 99th is the places in California

00:12:42 ◼   ► that I've been.

00:12:43 ◼   ► - Well, our water here is excellent

00:12:45 ◼   ► because it comes from the mountains.

00:12:46 ◼   ► - It's a great.

00:12:47 ◼   ► I mean, it's not chlorinated Pennsylvania water.

00:12:50 ◼   ► I'm not saying it's that bad, but.

00:12:52 ◼   ► - So speaking of your office,

00:12:56 ◼   ► I just, I wanted to do a quick check in with you.

00:12:59 ◼   ► It's been less than a year, but like, how is your,

00:13:02 ◼   ► how is that working at home

00:13:04 ◼   ► and being an independent content creator

00:13:07 ◼   ► and programmer on the internet that isn't going to a job

00:13:10 ◼   ► or zooming into a job or whatever every day.

00:13:12 ◼   ► I know you were remote for a lot of it

00:13:14 ◼   ► as so many of us were.

00:13:15 ◼   ► How's it going?

00:13:17 ◼   ► Switch Glass is at 2.0 now.

00:13:19 ◼   ► You're continuing your long game

00:13:21 ◼   ► of very slowly replicating drag thing.

00:13:24 ◼   ► - I don't think that's my long game.

00:13:26 ◼   ► Like here's the thing about Switch Glass.

00:13:27 ◼   ► I like working on that app.

00:13:30 ◼   ► Obviously I wanted it so I could have it on my Mac.

00:13:32 ◼   ► But then once I've done that I also like working on the app. I like making Mac apps

00:13:35 ◼   ► It's a thing I always wanted to do and now I get to do it

00:13:38 ◼   ► I'm always kind of like looking for excuses to do things to the app to improve it fix bugs add features stuff like that

00:13:43 ◼   ► But it does not make money

00:13:45 ◼   ► I mean, I mean it makes I'm not saying I make zero dollars

00:13:49 ◼   ► But I mean maybe it makes enough money for me to get a nice meal out once a month

00:13:53 ◼   ► Like that's the kind of money that so I the amount of time I spend on that I feel

00:13:57 ◼   ► Guilty about because it is not a useful

00:14:01 ◼   ► Use of my time right it's like I should just never work on it again because it is never going to bring in you know

00:14:08 ◼   ► Enough money to be worth the time I spend on it

00:14:11 ◼   ► So I read it like I treat it like playing destiny like it's a fun thing that I enjoy doing

00:14:16 ◼   ► But I don't fool myself into thinking it's work

00:14:18 ◼   ► And so that's part of how things are going is like there are lots of things that attract me interestingly not destiny because since I

00:14:24 ◼   ► Quit my job if you look at my destiny hours

00:14:26 ◼   ► I'm one of those you know sites that tracks how much you play I am playing like

00:14:30 ◼   ► massively less destiny than I used to,

00:14:32 ◼   ► which has surprised me.

00:14:34 ◼   ► Although now that I'm in it, it makes sense to me

00:14:36 ◼   ► because I just feel guilty ever goofing off.

00:14:40 ◼   ► But working on Switch Glass

00:14:42 ◼   ► is in the category of goofing off

00:14:43 ◼   ► and I have to stop myself from doing it

00:14:45 ◼   ► and concentrate more on other things,

00:14:47 ◼   ► spend my time on things that actually have a chance

00:14:50 ◼   ► of making me more than minimum wage

00:14:52 ◼   ► for the hours that I spend doing them.

00:14:54 ◼   ► And Switch Glass ain't it.

00:14:57 ◼   ► - I mean, it's like a hobby basically, right?

00:14:59 ◼   ► I mean, this is how I feel about when I'm doing like Python scripting of a home thing,

00:15:04 ◼   ► right?

00:15:05 ◼   ► It's like, it's fun, but it doesn't pay any bills.

00:15:08 ◼   ► Yeah, I mean, but the thing about Twitch Glass is like, making applications is a thing that

00:15:12 ◼   ► can make you money, just not any of the applications that I've written, because they're so narrow

00:15:16 ◼   ► interest and weird and have probably made all the money they're ever going to make.

00:15:21 ◼   ► So it's like, oh, if you want to make an app, find one that has a chance of making money.

00:15:27 ◼   ► and I'm, you know, so far I haven't gotten there.

00:15:29 ◼   ► - Okay, well, I have an idea for an Apple Watch app

00:15:31 ◼   ► for curling that I'll share with you at a later time.

00:15:34 ◼   ► - Apple Watch is not a Mac.

00:15:35 ◼   ► - I know it's not.

00:15:36 ◼   ► I know it's not.

00:15:37 ◼   ► I bet there are good, honestly, Jon,

00:15:39 ◼   ► I was thinking about this.

00:15:40 ◼   ► There are, there are so many, like, like people,

00:15:45 ◼   ► I know there are new Mac apps from time to time,

00:15:47 ◼   ► but like the Mac so often feels like a place

00:15:51 ◼   ► that nobody really wants to make new apps for,

00:15:53 ◼   ► especially utility apps,

00:15:55 ◼   ► that I feel like you're a couple of steps away

00:15:58 ◼   ► from finding some, like tapping into some kind

00:16:02 ◼   ► of perfect place where there's something

00:16:04 ◼   ► that Mac users want to see.

00:16:06 ◼   ► But then again, maybe not,

00:16:07 ◼   ► maybe nobody who's using a Mac cares

00:16:10 ◼   ► about like nice little utilities like they used to.

00:16:13 ◼   ► - Well, when you say utility apps,

00:16:14 ◼   ► like those are the type of apps that I like to make.

00:16:16 ◼   ► Like, but it's from my, from being raised in an environment

00:16:20 ◼   ► where my favorite Mac apps are ones that modified

00:16:23 ◼   ► augmented the system in some way and that door has been closing for so long

00:16:28 ◼   ► that like there are things I want to do with switch glass that I can't do. I have

00:16:32 ◼   ► ideas for other kinds of apps that are sort of system modification apps that I

00:16:35 ◼   ► just can't do because they're not possible like maybe they would have been

00:16:40 ◼   ► possible in the old days with hacksies or whatever but in the modern Mac they're

00:16:44 ◼   ► just not possible. If Apple doesn't provide hooks for the functionality you

00:16:47 ◼   ► want to mess with it's I mean it's beyond my skill to hack into the system

00:16:51 ◼   ► to do it, and even if you did, then you have the problem of,

00:16:53 ◼   ► okay, you can't sell this on the Mac App Store,

00:16:55 ◼   ► so now you're signing up to making your own website

00:16:57 ◼   ► to sell these things, or using some other third-party

00:17:00 ◼   ► reseller, or hooking up Stripe, or whatever.

00:17:02 ◼   ► And I can do all of that, but there's no way

00:17:05 ◼   ► that I'm gonna make an app that makes enough money

00:17:06 ◼   ► to even pay for my time that it would take

00:17:08 ◼   ► to make my own Mac software store

00:17:13 ◼   ► that people buy things to,

00:17:14 ◼   ► and then I have to do customer support for it.

00:17:16 ◼   ► The overhead of that is so massive

00:17:18 ◼   ► that I would need to come up with an idea

00:17:20 ◼   ► that I actually pay for that,

00:17:21 ◼   ► and all my ideas are like weird things

00:17:23 ◼   ► that modify the behavior of the system

00:17:26 ◼   ► in a way that Apple does not support, will never support,

00:17:28 ◼   ► and honestly, I don't even know if they're possible.

00:17:31 ◼   ► - Right.

00:17:31 ◼   ► - And so that's why I just like, I muse about those,

00:17:34 ◼   ► but, you know, and that's what I wanna make.

00:17:36 ◼   ► I don't wanna make a Notes application, a to-do app,

00:17:39 ◼   ► or, and if you do wanna make one, a regular Mac app,

00:17:41 ◼   ► not like a utility type thing,

00:17:43 ◼   ► then you're basically on the hook to make an iPad

00:17:45 ◼   ► and an iPhone one, and maybe a watch one,

00:17:47 ◼   ► because nobody buys a Mac app

00:17:49 ◼   ► and is just satisfied to have the Mac app

00:17:51 ◼   ► unless it is Mac specific, right?

00:17:53 ◼   ► - Right.

00:17:54 ◼   ► - Like if it modifies the Mac like my thing does,

00:17:56 ◼   ► no one wants Switch Glass on,

00:17:58 ◼   ► well, they probably do want an iPad, sorry everybody,

00:18:00 ◼   ► but no one wants it on their phone or their watch.

00:18:02 ◼   ► But if you make a Notes app,

00:18:04 ◼   ► people are gonna want it everywhere

00:18:05 ◼   ► and it's gonna have to sync

00:18:06 ◼   ► and then now you're talking about a real app

00:18:08 ◼   ► and you probably, you know, so.

00:18:10 ◼   ► Anyway, I try to limit my hours on Switch Glass

00:18:13 ◼   ► in front and center.

00:18:14 ◼   ► - I get it, I get it.

00:18:16 ◼   ► I mean, what are you doing

00:18:18 ◼   ► if those things don't pay the bills, what pays the bills?

00:18:21 ◼   ► Is that working up an idea for some more ATP merch?

00:18:25 ◼   ► Or is it, I mean, 'cause otherwise,

00:18:28 ◼   ► other than programming, is everything else you're doing

00:18:31 ◼   ► that pays the bills podcast related?

00:18:34 ◼   ► - Yeah, I mean, there's all the stuff

00:18:35 ◼   ► that's behind the scenes in terms of the ATP store,

00:18:38 ◼   ► for example, like I tend to make all of the shirt designs

00:18:42 ◼   ► that aren't done by professional designers.

00:18:44 ◼   ► I do all the, you know, the getting those images ready,

00:18:47 ◼   ► uploading them, specifying the products, working with the stores to get that

00:18:51 ◼   ► stuff online at the little ATP.fm/store page, the dinky little HTML page, I'm

00:18:56 ◼   ► the one who updates that, and you know like and you know just doing the show

00:19:00 ◼   ► notes every week, gathering links, doing research, reading things, watching videos,

00:19:04 ◼   ► everything that needs to be done for the show and for my multiple shows right and

00:19:09 ◼   ► yeah that when I spend time doing that I can now spend more time doing that and

00:19:13 ◼   ► not feel so rushed when I'm doing it and that counts as real work because

00:19:16 ◼   ► because podcasts actually do pay the bills.

00:19:20 ◼   ► - I like the fact that you're doing merch stuff

00:19:23 ◼   ► because I feel like, Myke and I talk about it a bit,

00:19:26 ◼   ► but we feel like that's one of the great unexplored aspects

00:19:30 ◼   ► of a lot of what we do.

00:19:33 ◼   ► I know that he and Gray do a good job with Cortex with that,

00:19:36 ◼   ► but like, I felt like with "Six Colors"

00:19:38 ◼   ► and "The Incomparable" and with "Upgrade"

00:19:40 ◼   ► that there's more to be done there, right?

00:19:42 ◼   ► Like anybody who has bought a sweatshirt from us

00:19:44 ◼   ► or something knows that like it's super sporadic.

00:19:48 ◼   ► And we do sometimes think, okay, we're gonna do this

00:19:51 ◼   ► and we're gonna get a design and we do it.

00:19:53 ◼   ► And it's fun, but there's no like real plan or discipline

00:19:58 ◼   ► put into like, here's when we're gonna do our merch

00:20:02 ◼   ► and here's what the designs are that are possible

00:20:05 ◼   ► and what are we gonna do?

00:20:06 ◼   ► We don't like, we don't do that.

00:20:09 ◼   ► And so it's either we adjust our schedule to do that

00:20:14 ◼   ► or we find somebody to help us with it,

00:20:17 ◼   ► but it's a real thing.

00:20:19 ◼   ► And all the new ATP merch is, you know,

00:20:22 ◼   ► would that have happened without you putting time on it?

00:20:26 ◼   ► Yeah, maybe.

00:20:28 ◼   ► - I was always doing this, but again,

00:20:29 ◼   ► now I don't have to do it in such a frantic rush.

00:20:31 ◼   ► I don't have to like sacrifice sleep to do it.

00:20:33 ◼   ► Like I can do it during normal working hours.

00:20:35 ◼   ► And it's just the normal bookkeeping stuff

00:20:37 ◼   ► or like obviously the chicken hat involved me specifically

00:20:40 ◼   ► 'cause I was going back and forth with the Studio Neat folks

00:20:42 ◼   ► with fabric samples and getting that thing

00:20:45 ◼   ► the way we wanted it.

00:20:46 ◼   ► But even just things like,

00:20:49 ◼   ► we do have a fairly regimented sales.

00:20:50 ◼   ► We do a holiday sale towards the end of the year.

00:20:53 ◼   ► We do a pre-WWDC sale, which is kind of a relic

00:20:56 ◼   ► from when we used to want people to have the shirts

00:20:57 ◼   ► so they could wear them at WWDC.

00:20:59 ◼   ► That doesn't really matter anymore.

00:21:01 ◼   ► And we try to have sales at regular intervals

00:21:06 ◼   ► and the winter/fall sales,

00:21:07 ◼   ► we sell the long sleeve warm stuff and the winter hats

00:21:10 ◼   ► and in the summer ones we sell t-shirts.

00:21:13 ◼   ► But all year, kind of like the gift ideas document,

00:21:15 ◼   ► I'm always thinking of what would be a good idea

00:21:17 ◼   ► for a t-shirt design or for some piece of merch

00:21:20 ◼   ► that we could do that we haven't done before.

00:21:22 ◼   ► And those come to me at random intervals.

00:21:24 ◼   ► And when they come, I have to strike while the iron is hot,

00:21:27 ◼   ► come up with the design, maybe I throw it away

00:21:29 ◼   ► 'cause it actually wasn't a good idea to begin with.

00:21:32 ◼   ► We've been cruising on these M1 shirts for a while

00:21:35 ◼   ► just because that was such a nice coincidence

00:21:36 ◼   ► that Apple did that and I enjoy making the little chips

00:21:39 ◼   ► on the back and everything,

00:21:40 ◼   ► and those will probably continue to trend along.

00:21:42 ◼   ► But I'm looking for other things like the chicken hat

00:21:45 ◼   ► or even the mugs, like weird stuff that we haven't done

00:21:47 ◼   ► before that it turns out people like.

00:21:48 ◼   ► And we try them out and we see if people like them.

00:21:51 ◼   ► Sometimes they do, sometimes they don't.

00:21:52 ◼   ► Apparently people love chicken hats.

00:21:54 ◼   ► They didn't really like pins,

00:21:55 ◼   ► so we haven't done pins for a while.

00:21:58 ◼   ► But yeah, that's more of a creative endeavor,

00:22:01 ◼   ► like sitting down and trying to come up with some designs

00:22:04 ◼   ► that you think will print okay on a shirt

00:22:06 ◼   ► that are funny and interesting

00:22:07 ◼   ► that will make people wanna buy one.

00:22:09 ◼   ► As somebody who, you have definitely told the stories

00:22:12 ◼   ► of having bought products that you love

00:22:14 ◼   ► because you wanna have backups

00:22:16 ◼   ► for when they're inevitably discontinued.

00:22:19 ◼   ► And I've done that too.

00:22:20 ◼   ► Not as much as you, but I've definitely done that

00:22:23 ◼   ► where it's like, oh no, they made a new version

00:22:26 ◼   ► of this thing that is no good.

00:22:27 ◼   ► And they're still, the old one is still available.

00:22:30 ◼   ► I'm buying another one of those right now

00:22:32 ◼   ► before they go out of stock forever.

00:22:36 ◼   ► But like the chicken hat took it to a whole new level, right?

00:22:39 ◼   ► Cause you were, as you mentioned,

00:22:41 ◼   ► you work with Studio Neat to replicate a hat you love that they don't make

00:22:46 ◼   ► anymore. Um, you never,

00:22:48 ◼   ► did you ever send them your hat or did you just sort of describe it and they

00:22:52 ◼   ► tried to figure out what your hat was?

00:22:53 ◼   ► I didn't send the, they asked and I refuse cause when am I going to send them

00:22:57 ◼   ► my one original hat? No, it's not going to happen. Um, but I sent them many,

00:23:01 ◼   ► many pictures, many pictures with rulers laying on top of it,

00:23:04 ◼   ► with the prototypes laid on top of the thing,

00:23:06 ◼   ► like just so many pictures, so many dimensions,

00:23:09 ◼   ► the hat turned inside out so you could see

00:23:10 ◼   ► how it's constructed.

00:23:12 ◼   ► There were lots of images going back and forth on Slack

00:23:14 ◼   ► to work up the design of the new hat.

00:23:17 ◼   ► - And is it a good match?

00:23:18 ◼   ► - Yeah, like one of the last round of things,

00:23:21 ◼   ► the last prototype, you can lay it on top of my hat

00:23:23 ◼   ► and they are exactly the same dimensions.

00:23:26 ◼   ► The only difference, as I said on the show,

00:23:27 ◼   ► is the material is not as thick.

00:23:28 ◼   ► We got as thick as we possibly could

00:23:29 ◼   ► from the same company that made the material,

00:23:31 ◼   ► but they just don't make material as thick anymore.

00:23:33 ◼   ► - Interesting.

00:23:34 ◼   ► And again, I would tap that as a feature

00:23:37 ◼   ► because already people bought the chicken hat and wore it

00:23:39 ◼   ► and they're like, I cannot believe how hot this hat is.

00:23:41 ◼   ► I'm like, can you imagine if it was twice as thick?

00:23:43 ◼   ► Yeah, I'm one of those people who's always cold.

00:23:46 ◼   ► Most people are not like me.

00:23:47 ◼   ► So I think the thinner chicken hat is an advantage.

00:23:50 ◼   ► - Are you gonna do a second run?

00:23:52 ◼   ► - Well, we've already done multiple runs.

00:23:53 ◼   ► We've already massively underestimated

00:23:56 ◼   ► how many people wanted these and scrambled

00:23:57 ◼   ► to get another order.

00:23:58 ◼   ► And in fact, we have a third order coming in right now.

00:24:00 ◼   ► So by the time you listen to this,

00:24:02 ◼   ► you may be able to go to atp.com/store

00:24:04 ◼   ► and buy a new chicken hat.

00:24:05 ◼   ► But either way, we're going to announce

00:24:07 ◼   ► when they're available on the ATP podcast,

00:24:10 ◼   ► because there was a backlog of people

00:24:11 ◼   ► who wanted them and couldn't get them.

00:24:13 ◼   ► - You don't wanna do it here?

00:24:14 ◼   ► - I mean, they're not, they just shipped

00:24:16 ◼   ► from our manufacturer to Cotton Bureau.

00:24:19 ◼   ► They haven't arrived at Cotton Bureau yet,

00:24:21 ◼   ► and we don't wanna put them up in the store

00:24:22 ◼   ► until they've arrived, 'cause what if they get lost

00:24:23 ◼   ► in the mail or something?

00:24:24 ◼   ► Once they arrive at Cotton Bureau--

00:24:26 ◼   ► - Cotton Bureau has a let us know when you,

00:24:28 ◼   ► if you, when this is back in stock, if you want it,

00:24:31 ◼   ► thing on their site, too, for people

00:24:32 ◼   ► who have clicked that button.

00:24:34 ◼   ► - And that's how we know,

00:24:36 ◼   ► that's how we hope we know how many we should order

00:24:38 ◼   ► for this final batch.

00:24:39 ◼   ► It's how many people were still clicking on the,

00:24:40 ◼   ► let me know when this is back in stock.

00:24:42 ◼   ► - You say final, but what's gonna happen next

00:24:43 ◼   ► is that in a year or two, what you're gonna do

00:24:45 ◼   ► is you're gonna do the revised chicken hat

00:24:46 ◼   ► that comes in a color and has a different thing on it.

00:24:49 ◼   ► And is a-

00:24:50 ◼   ► - We're already talking about like doing embroidery

00:24:51 ◼   ► instead of the tag or whatever, even for this batch.

00:24:54 ◼   ► I said, no, for everything during this holiday season,

00:24:56 ◼   ► let's just make them all the same

00:24:57 ◼   ► 'cause it's too complicated.

00:24:59 ◼   ► - This V1, this is the first edition chicken hat.

00:25:01 ◼   ► - Yeah, and maybe we'll never do it again.

00:25:03 ◼   ► That's the other thing about,

00:25:04 ◼   ► speaking of having backups and stuff,

00:25:05 ◼   ► lots of people, including me and my kids,

00:25:07 ◼   ► sometimes they have a shirt and it's like,

00:25:08 ◼   ► we haven't sold that shirt in four years.

00:25:10 ◼   ► And people are like, oh, I've worn out that T-shirt,

00:25:12 ◼   ► I would love a second one.

00:25:13 ◼   ► So maybe one of the next ideas that we're gonna do

00:25:15 ◼   ► is a like throwback sale where we sell a bunch of shirts

00:25:18 ◼   ► we haven't sold in years and years.

00:25:21 ◼   ► - Right.

00:25:22 ◼   ► - It's a good idea.

00:25:23 ◼   ► - Yeah.

00:25:24 ◼   ► - I have a little bit more housekeeping to do.

00:25:28 ◼   ► I wanted to mention, of course, we have Upgrade Plus,

00:25:31 ◼   ► If you love the show and want more of it,

00:25:33 ◼   ► you get no ads and bonus content every week

00:25:35 ◼   ► and access of course to the Relay FM members discord,

00:25:38 ◼   ► get upgradeplus.com, $5 a month or $50 a year.

00:25:41 ◼   ► And from now until December 17th,

00:25:43 ◼   ► you can save 20% on an annual plan.

00:25:46 ◼   ► Just find out more at giverelay.com

00:25:48 ◼   ► or just use the code 2023holidays to check out.

00:25:51 ◼   ► You'll get 20% off your first year

00:25:52 ◼   ► of an annual subscription to upgradeplus

00:25:55 ◼   ► and then plans renew at the full price after year one.

00:25:58 ◼   ► And also go, if you have not yet,

00:26:01 ◼   ► to nominate your favorites at upgradees.vote

00:26:04 ◼   ► for the ninth annual upgradees.

00:26:07 ◼   ► Voting is only open for another week.

00:26:11 ◼   ► So please do it now.

00:26:12 ◼   ► It helps us a lot, allows the voice of the upgradeians

00:26:16 ◼   ► to be heard in the upgradees,

00:26:18 ◼   ► which we'll be releasing at the end of the year

00:26:20 ◼   ► as is our tradition.

00:26:21 ◼   ► I believe it'll be on Boxing Day this year.

00:26:24 ◼   ► So we need those votes in by next week, by next Monday.

00:26:27 ◼   ► So please give it a little bit of time and thought,

00:26:32 ◼   ► and we appreciate it.

00:26:34 ◼   ► Also, I have a brief amount of follow-up

00:26:37 ◼   ► that I just wanted to throw in here

00:26:39 ◼   ► before we take our first break.

00:26:41 ◼   ► Listener Tim from the iPad Pros podcast

00:26:45 ◼   ► sent in a photo that was great.

00:26:48 ◼   ► They went to the hospital, he and his partner,

00:26:51 ◼   ► to check in on the progress of their pregnancy.

00:26:56 ◼   ► One thing led to another.

00:27:00 ◼   ► Tim happened to be wearing his upgrade hoodie

00:27:02 ◼   ► when they went to the hospital.

00:27:04 ◼   ► And that means that Tim was wearing the upgrade hoodie

00:27:07 ◼   ► when his child was born two months early.

00:27:11 ◼   ► And everybody's home and fine now.

00:27:14 ◼   ► And I think it's been a couple of months.

00:27:17 ◼   ► But so shout out to Tim for showing off the podcast

00:27:22 ◼   ► during a major life event.

00:27:23 ◼   ► Love it.

00:27:24 ◼   ► Also, here's an interesting one.

00:27:26 ◼   ► I heard this from a few people inside Apple too,

00:27:28 ◼   ► who were kind of surprised by it.

00:27:30 ◼   ► Apple Vice President of Software Engineering,

00:27:32 ◼   ► John Stover has left Apple to become the genius in charge

00:27:37 ◼   ► of technology for Roblox.

00:27:43 ◼   ► Roblox.

00:27:44 ◼   ► What do you think about this, John?

00:27:47 ◼   ► I mean, everybody, there are lots of reasons

00:27:49 ◼   ► to get a new job, including wanting a change

00:27:51 ◼   ► and wanting a new challenge.

00:27:52 ◼   ► And my understanding is that John Stouffer's kids,

00:27:56 ◼   ► kid, kids are Roblox fanatics.

00:27:59 ◼   ► And so that's an interesting thing too.

00:28:03 ◼   ► But there's Mark Gurman reported this and pointed out

00:28:06 ◼   ► that like a lot of Apple people at the VP level,

00:28:08 ◼   ► which is one down from the senior VPs

00:28:10 ◼   ► that report to Tim Cook, there's been a lot of,

00:28:13 ◼   ► and there's like 100, 150 of them.

00:28:15 ◼   ► And there've been a bunch of them leaving lately.

00:28:18 ◼   ► - I mean, yeah.

00:28:18 ◼   ► So when people high up in Apple leave,

00:28:21 ◼   ► because two kinds of departures.

00:28:23 ◼   ► One, we get less of these these days,

00:28:24 ◼   ► but it used to be a big thing.

00:28:26 ◼   ► They leave because they're vice presidents

00:28:28 ◼   ► because they've been with Apple since 1999

00:28:31 ◼   ► and they have so much stock options

00:28:32 ◼   ► that they never need to work again.

00:28:34 ◼   ► So that's why they leave, right?

00:28:35 ◼   ► I'm rich and I've been here out of a sense of loyalty,

00:28:37 ◼   ► but at a certain point I go, you know what?

00:28:39 ◼   ► I don't need to work anymore.

00:28:40 ◼   ► My Apple stock over so much money, I'm out of here.

00:28:42 ◼   ► And the second kind is,

00:28:44 ◼   ► I haven't been with Apple since 1999,

00:28:47 ◼   ► but I am kind of a senior VP.

00:28:48 ◼   ► I'm probably not gonna go any farther

00:28:50 ◼   ► and Apple is not where the action's happening.

00:28:53 ◼   ► You know, where the action's happening, Roblox.

00:28:56 ◼   ► (laughing)

00:28:57 ◼   ► And that's not a joke.

00:28:58 ◼   ► Like, Roblox went from nothing to making tons of money.

00:29:02 ◼   ► It is a new, different kind of business.

00:29:03 ◼   ► And if you wanna go somewhere with growth potential

00:29:06 ◼   ► that's gonna pay you a ton more than Apple pays you,

00:29:09 ◼   ► go to one of these companies that has, you know,

00:29:11 ◼   ► this amazing money spigot.

00:29:12 ◼   ► Back in the day, it might've been Candy Crush.

00:29:13 ◼   ► Today, it's Roblox.

00:29:15 ◼   ► It's just a question of whether you can stomach

00:29:16 ◼   ► how these companies make money.

00:29:18 ◼   ► But the question may also increasingly apply

00:29:20 ◼   ► to Apple as well.

00:29:21 ◼   ► So if you've been in Apple for a while

00:29:23 ◼   ► and you think you've reached your limit

00:29:25 ◼   ► of how far you're gonna go in this company

00:29:26 ◼   ► and you want to go somewhere where they're gonna pay you

00:29:29 ◼   ► a lot more than Apple pays you

00:29:30 ◼   ► and have a much more upside with their stock or whatever,

00:29:34 ◼   ► someplace like Roblox is where it's at.

00:29:36 ◼   ► Apple's not a new up and comer, right?

00:29:38 ◼   ► So I think those are the kind of departures you see

00:29:41 ◼   ► at the VP level now, where it's like,

00:29:42 ◼   ► well, I wanna do something else.

00:29:44 ◼   ► I wanna be a bigger fish in a smaller pond

00:29:46 ◼   ► They don't want to get paid more money.

00:29:48 ◼   ► - Yeah, it makes sense.

00:29:49 ◼   ► I mean, this is, and I know we've been saying it here

00:29:52 ◼   ► for a long time and I know ATP,

00:29:53 ◼   ► you guys have talked about it.

00:29:54 ◼   ► Like, Brain Drain is a serious threat for Apple, right?

00:29:58 ◼   ► They have a bunch of people who've been there

00:30:01 ◼   ► since the beginning of this sort of era

00:30:04 ◼   ► of Steve Jobs's return and Apple's massive growth,

00:30:08 ◼   ► and they have a lot of money,

00:30:10 ◼   ► and they don't need to work necessarily, a lot of them.

00:30:15 ◼   ► And then I think with somebody like this,

00:30:17 ◼   ► I imagine it's also the case that he looks at this

00:30:19 ◼   ► and says, well, Craig's not leaving.

00:30:22 ◼   ► So I have nowhere else to go

00:30:24 ◼   ► and I want a new challenge, right?

00:30:26 ◼   ► - And he might not want Craig's job either

00:30:28 ◼   ► because being Craig is different

00:30:29 ◼   ► than having Craig's job at Roblox.

00:30:31 ◼   ► Because in Roblox, you know what the deal is at Apple.

00:30:35 ◼   ► If you have Craig's job, your influence is kind of set.

00:30:40 ◼   ► You know what it's gonna be like at Apple.

00:30:42 ◼   ► Whereas in Roblox, the CTO or the head of whatever

00:30:45 ◼   ► might have much more influence on the direction

00:30:46 ◼   ► of that company than Craig has on the influence,

00:30:49 ◼   ► on the direction of Apple.

00:30:50 ◼   ► - That's also true.

00:30:51 ◼   ► Looking for a new challenge though,

00:30:54 ◼   ► I just, I keep coming to that.

00:30:55 ◼   ► Like I totally get it.

00:30:57 ◼   ► - It's also more money.

00:30:58 ◼   ► I mean, let's be honest.

00:30:59 ◼   ► - Sure.

00:31:00 ◼   ► I'm sure it is.

00:31:01 ◼   ► I'm sure it is.

00:31:02 ◼   ► But at some point, if you've got all the money,

00:31:04 ◼   ► then more money is like, oh, more of that stuff, okay.

00:31:07 ◼   ► But maybe not, maybe it doesn't have all the money.

00:31:08 ◼   ► I don't know.

00:31:10 ◼   ► Also, we've been talking a lot about Apple in China

00:31:13 ◼   ► and about Apple's chip design prowess

00:31:16 ◼   ► and Taiwan semiconductors prowess in building

00:31:21 ◼   ► and making chips and reducing the size of the chips.

00:31:25 ◼   ► And, you know, but the,

00:31:28 ◼   ► there's so much wrapped around this, right?

00:31:30 ◼   ► It's the geopolitical issues and the technical issues.

00:31:34 ◼   ► And so we've talked about it a lot.

00:31:36 ◼   ► I just wanted to mention a report

00:31:38 ◼   ► from the Wall Street Journal pointing out that tomorrow,

00:31:41 ◼   ► we record this, Tuesday, December 6th, US President Joe Biden is showing up at a ceremony

00:31:49 ◼   ► for the opening, or the, it's like groundbreaking, except they call it like a "tool-in ceremony."

00:31:57 ◼   ► It's like they, it's that they stick the, the, I don't know whether that means they

00:32:02 ◼   ► stick a shovel in the ground, or whether they open the factory, or they're loading things

00:32:07 ◼   ► into the building or what, but it's a ceremony at the new TSMC plant site in Phoenix, which

00:32:16 ◼   ► is this part of this idea of like the US wants to induce chip makers to not just make their

00:32:23 ◼   ► chips in places like Taiwan, but also in the United States. And this is a, so and Arizona

00:32:29 ◼   ► and Phoenix have been trying to do this for a little while now. They had that one that

00:32:32 ◼   ► fell through, but this thing seems to be going and happening. A little tidbit from the Wall

00:32:39 ◼   ► Street Journal story about this that I thought was interesting is TSMC executives have said

00:32:44 ◼   ► it isn't easy to recreate in America the manufacturing ecosystem they have built over decades in

00:32:48 ◼   ► Taiwan, drawing on local engineering talent and a network of suppliers, including many

00:32:52 ◼   ► in East Asia. TSMC founder Morris Chang said the cost of making chips in Arizona may be

00:32:58 ◼   ► at least 50% higher than in Taiwan. So it's an interesting thing that, and obviously Apple

00:33:05 ◼   ► is a huge customer of TSMC and may be a customer of them for the chips here. Could be just

00:33:12 ◼   ► this is the price you pay if you want to diversify, if you want chip making in America and you

00:33:16 ◼   ► want to diversify your chip making so that your eggs are not all in one basket. But still,

00:33:21 ◼   ► from a pure economic standpoint, 50% higher is quite a penalty to pay.

00:33:27 ◼   ► - Well, from a strategic perspective,

00:33:30 ◼   ► everyone's just got to eat this because I think, you know,

00:33:34 ◼   ► if it was always theoretically untenable

00:33:36 ◼   ► and now it's becoming more practically untenable

00:33:39 ◼   ► to have such important functionality so far

00:33:43 ◼   ► from the US's sphere of influence, right?

00:33:47 ◼   ► To have so much manufacturing happening in China,

00:33:49 ◼   ► which is a country that we don't see eye on

00:33:51 ◼   ► on a lot of things and to have TSMC being in Taiwan

00:33:56 ◼   ► also a potentially volatile region,

00:33:59 ◼   ► not to mention the fact that none of the companies are,

00:34:02 ◼   ► you know, American companies.

00:34:04 ◼   ► Intel used to be the big dog

00:34:06 ◼   ► in semiconductor manufacturing, but no more, right?

00:34:09 ◼   ► So Apple's sort of long-term strategic project

00:34:14 ◼   ► that is going to continue on long past the tenure

00:34:16 ◼   ► of Tim Cook is to basically undo everything Tim Cook did

00:34:19 ◼   ► in terms of manufacturing and not so much get us out

00:34:21 ◼   ► of China, but remove, Apple needs to remove its dependency

00:34:25 ◼   ► and that's going to take literal decades.

00:34:27 ◼   ► And it's just this long project, right?

00:34:29 ◼   ► And then the US, their problem is,

00:34:31 ◼   ► we used to be important in semiconductor manufacturing

00:34:34 ◼   ► and now we're not anymore.

00:34:36 ◼   ► And semiconductors are not going away

00:34:38 ◼   ► and are not going to become unimportant anytime soon.

00:34:40 ◼   ► So America, it would make sense for us as a country

00:34:44 ◼   ► and a government to invest money

00:34:45 ◼   ► to get semiconductor manufacturing expertise

00:34:50 ◼   ► in this country.

00:34:51 ◼   ► Having TSMC have a plant here, that's better than nothing.

00:34:55 ◼   ► What would be even better?

00:34:56 ◼   ► And I'm sure Intel would agree is,

00:34:57 ◼   ► hey, Intel says, hey, give us billions of dollars

00:34:59 ◼   ► in government money, and we'll get better

00:35:01 ◼   ► at making semiconductors again.

00:35:03 ◼   ► There's lots of different ways this can go,

00:35:05 ◼   ► but there's no avoiding,

00:35:07 ◼   ► this is a task that everyone has to undertake,

00:35:09 ◼   ► and it's going to be expensive and painful

00:35:12 ◼   ► and time-consuming for everybody,

00:35:13 ◼   ► for Apple, for the US government.

00:35:15 ◼   ► And in the end, we're not gonna be able

00:35:16 ◼   ► to make things as inexpensively, right?

00:35:18 ◼   ► But that's a strategic cost.

00:35:20 ◼   ► The same way we subsidize corn out the wazoo

00:35:22 ◼   ► and subsidize oil, we need to subsidize semiconductors.

00:35:26 ◼   ► They're just as an important commodity in the future.

00:35:28 ◼   ► And it's not good for us to be dependent

00:35:30 ◼   ► on Taiwan's semiconductor.

00:35:31 ◼   ► But I mean us, I mean the little rest of the planet.

00:35:34 ◼   ► It depends because they're the only people

00:35:36 ◼   ► who can make the very best chips.

00:35:38 ◼   ► And everybody in the world wants a phone

00:35:40 ◼   ► manufactured at three nanometers, right?

00:35:42 ◼   ► - Also the fundamental issue of like,

00:35:45 ◼   ► how do you get the manufacturing ecosystem

00:35:48 ◼   ► that they built over the decades in Taiwan?

00:35:51 ◼   ► How do you get drawing on local engineering talent?

00:35:53 ◼   ► How do you get a network of suppliers?

00:35:55 ◼   ► How do you get all of the local knowledge

00:35:58 ◼   ► that already exists in Taiwan?

00:35:59 ◼   ► And the answer is you have to do it, right?

00:36:01 ◼   ► Like it doesn't necessarily mean it is going to happen,

00:36:05 ◼   ► but if you don't start this, it won't happen, right?

00:36:08 ◼   ► So by doing this, you're like, oh,

00:36:10 ◼   ► well now we're gonna get,

00:36:12 ◼   ► we're gonna have some of our engineers come in,

00:36:14 ◼   ► but we're also gonna hire engineers in the US

00:36:16 ◼   ► who are American engineers.

00:36:18 ◼   ► And we're gonna, oh, this supply thing

00:36:21 ◼   ► is really inconvenient, but now that we're here,

00:36:23 ◼   ► there is a financial motivation for somebody else

00:36:26 ◼   ► to set up an alternate supply line

00:36:28 ◼   ► that runs into Arizona or into the US somewhere.

00:36:32 ◼   ► And that is, you know, and the only way you do that

00:36:35 ◼   ► is by spending, as you said, by spending the money,

00:36:39 ◼   ► essentially, at this point to make it happen,

00:36:40 ◼   ► because otherwise, if all else is equal,

00:36:43 ◼   ► you will end up with just the TSMC factories

00:36:47 ◼   ► that are in Taiwan.

00:36:47 ◼   ► So if the US strategically wants this to happen,

00:36:51 ◼   ► and they're, you know, again,

00:36:53 ◼   ► not saying it will happen that way,

00:36:55 ◼   ► but it could happen that way.

00:36:56 ◼   ► And it's not gonna, it's not gonna,

00:36:58 ◼   ► like when they talked about the Mac Pro assembly in the US,

00:37:02 ◼   ► or they talked about why,

00:37:04 ◼   ► why doesn't Apple assemble other products in the US?

00:37:06 ◼   ► And the answer is, there's so many things

00:37:08 ◼   ► that feed into the assembly and the supply chain,

00:37:11 ◼   ► and that stuff is all in and around China.

00:37:16 ◼   ► Well, how do we solve that?

00:37:17 ◼   ► Like you can't just snap your fingers and build a factory

00:37:19 ◼   ► and say, we solved it, right?

00:37:20 ◼   ► Because the whole, there's a whole web of things

00:37:24 ◼   ► in an ecosystem around it.

00:37:26 ◼   ► And so if you wanna do something like this,

00:37:27 ◼   ► yeah, you make the commitment, you take the hit

00:37:29 ◼   ► and you say, yeah, this is gonna be hard

00:37:30 ◼   ► and it's gonna be expensive and it's gonna cost,

00:37:32 ◼   ► half again as much as a premium, at least to do this,

00:37:36 ◼   ► but you gotta start somewhere.

00:37:39 ◼   ► - Yeah, you have to invest.

00:37:40 ◼   ► Investing means putting money upfront

00:37:41 ◼   ► and whether it's Apple investing or the US government

00:37:44 ◼   ► doing taxpayer money investing,

00:37:45 ◼   ► that investment should hopefully pay off.

00:37:46 ◼   ► And the other factor here that I always think about in my darker moments is that in the

00:37:51 ◼   ► same way that it's not great to have manufacturing in China and you know are involved to regions

00:37:56 ◼   ► like potentially volatile regions like Taiwan, gathering enough engineering talent to a particular

00:38:02 ◼   ► location in the US does sometimes present a challenge because the places that are best

00:38:08 ◼   ► able to support large manufacturing infrastructure where we have lots of land relatively inexpensively

00:38:14 ◼   ► and you know quote-unquote business-friendly laws and regulations are also places where it's harder

00:38:20 ◼   ► to get engineering talent because people who are you know have engineering skills they're able to

00:38:25 ◼   ► get a good paying job anywhere in the country don't want to live in a state where abortion is illegal

00:38:30 ◼   ► and the company you know the company factory gets bomb threats because you give trans people health

00:38:34 ◼   ► care. Increasingly that is becoming a problem in our country so how do you get all the best engineers

00:38:41 ◼   ► to come to your place and want to live there, you have to have you have to be in one of the

00:38:45 ◼   ► US states that is less oppressive, let's say. And I think that's, you know, that's something

00:38:50 ◼   ► people don't think about. Oh, Texas, great business laws, but geez, not great human laws.

00:38:55 ◼   ► Well, we saw that with a lot of the employee resistance to moving a lot of Disney businesses

00:38:59 ◼   ► from LA to Florida too. Yeah, exactly. And, you know, and again, you know, Silicon Valley,

00:39:04 ◼   ► San Francisco, everything's too expensive. The teachers can't live in the city. Like,

00:39:06 ◼   ► it's not like there aren't problems elsewhere, but you know, there are different challenges,

00:39:10 ◼   ► You know, being able to have your employees afford housing if you're paying them well

00:39:15 ◼   ► is surmountable, but you know, employees not willing to go to your state because you don't

00:39:20 ◼   ► recognize their humanity or willing to give them health care is only able to be overcome

00:39:24 ◼   ► at the ballot box.

00:39:25 ◼   ► It's certainly a big challenge in general also just to create places like, I mean, it's

00:39:33 ◼   ► sort of what I was saying before, but you know, you have to create a place where people

00:39:36 ◼   ► want to be as you pointed out, and it needs to be a place where there are options because

00:39:40 ◼   ► you don't, you also don't want to have like, well, you can move to this place that's out

00:39:44 ◼   ► in the middle of nowhere where we do this thing. It's like, okay.

00:39:47 ◼   ► And we're the only, we're the only place where you can get a job. So if you want to change

00:39:50 ◼   ► jobs, you're going to have to move who knows where, right? Like that, and what you want

00:39:54 ◼   ► though is, you know, if I'm Arizona State University, right? Or the University of Arizona,

00:39:59 ◼   ► I'm like, well, this is really interesting. What are we, are we training, you know, chip

00:40:04 ◼   ► manufacturing and ship engineers and like,

00:40:06 ◼   ► are we training people to do these jobs?

00:40:07 ◼   ► - We're training people to oversee the manufacturing

00:40:10 ◼   ► by flying on a flight from California to China

00:40:14 ◼   ► to oversee the manufacturer and other people's plants.

00:40:17 ◼   ► - Yeah, well, so it's that like, well,

00:40:19 ◼   ► can we build an ecosystem?

00:40:22 ◼   ► And you end up in the situation too, where it's like,

00:40:24 ◼   ► well, and then when Intel is like,

00:40:25 ◼   ► well, we also want to build a plant.

00:40:26 ◼   ► It's like, well, the place to build it

00:40:28 ◼   ► is next to the one that TSMC has, right?

00:40:32 ◼   ► because this is like Apple's,

00:40:34 ◼   ► when they wanted their self-driving car project to succeed

00:40:37 ◼   ► and they opened an office literally right next door

00:40:40 ◼   ► to where the QNX offices were in Ontario,

00:40:44 ◼   ► because that was who they were hiring.

00:40:46 ◼   ► Or why was the Intel now Apple modem business in San Diego?

00:40:50 ◼   ► It's 'cause that's where Qualcomm is, right?

00:40:52 ◼   ► Like that's, and you build up a whole ecosystem

00:40:55 ◼   ► of people and related companies,

00:40:58 ◼   ► and they all are thinking about cellular radios

00:41:03 ◼   ► and chip development and all of that in San Diego,

00:41:06 ◼   ► and they're all there.

00:41:07 ◼   ► So then the competitors come

00:41:10 ◼   ► and it all just kind of builds out from there.

00:41:12 ◼   ► So we'll see, but you got to start somewhere.

00:41:15 ◼   ► And so TSMC, tool in whatever that means in Phoenix tomorrow.

00:41:20 ◼   ► - This episode of Upgrade is brought to you by StoryWorth.

00:41:26 ◼   ► If you're spending time with loved ones for the holidays,

00:41:28 ◼   ► There are chances that you're going to hear a ton of stories.

00:41:31 ◼   ► There are ones that you love to hear, and there are ones that you've probably heard

00:41:35 ◼   ► a few too many times.

00:41:37 ◼   ► But if you ever wondered what it would be like to actually document those timeless stories,

00:41:40 ◼   ► it can be challenging to imagine writing an entire book of life memories.

00:41:45 ◼   ► Where would you start?

00:41:46 ◼   ► Well, StoryWorth makes it fun and easy.

00:41:48 ◼   ► This is how anyone can write a book about their life.

00:41:51 ◼   ► Every week, StoryWorth will email your loved one a single life-related question that you

00:41:56 ◼   ► pick from their collection, like "what's the bravest thing you've ever done in your

00:41:59 ◼   ► life?" or "what's the farthest you've ever travelled?"

00:42:02 ◼   ► Then all they have to do is reply to that prompt of a story.

00:42:05 ◼   ► They can include images if they want to.

00:42:07 ◼   ► And then after a year, StoryWorth will compile all of your loved ones stories, memories and

00:42:11 ◼   ► even photos in a keepsake book.

00:42:14 ◼   ► Millions of stories have been told at StoryWorth because they make the process so simple.

00:42:18 ◼   ► Get started with your loved one for the holidays and before you know it you'll be able to

00:42:22 ◼   ► reflect on those timeless stories together for generations to come.

00:42:25 ◼   ► The genuinely the thing that I find the most awesome about StoryWorth is the way they word

00:42:30 ◼   ► the questions and the types of things that they can give you to ask that generate stuff

00:42:35 ◼   ► that you never would have thought to ask on your own.

00:42:37 ◼   ► I think this is an amazing gift idea for someone in your life as a way to continue building

00:42:41 ◼   ► the history of your family that can be kept and passed on.

00:42:44 ◼   ► I think that it's something very special and it truly creates something memorable and something

00:42:49 ◼   ► to pass on.

00:42:50 ◼   ► I think it's amazing.

00:42:51 ◼   ► I've tried it out and I think it's super cool.

00:42:54 ◼   ► Help your family share their story this holiday season with StoryWorth.

00:42:58 ◼   ► Go to storyworth.com/upgrade and you can save $10 on your first purchase.

00:43:03 ◼   ► That's S-T-O-R-Y-W-O-R-T-H dot com slash upgrade to save $10 on your first purchase at storyworth.com/upgrade.

00:43:13 ◼   ► Our thanks to StoryWorth for their support of this show and Relay FM.

00:43:16 ◼   ► All right, Jon, when we were putting the show notes together, you suggested that we maybe

00:43:21 ◼   ► talk about AI stuff.

00:43:24 ◼   ► You suggested sort of AI generating,

00:43:28 ◼   ► there are all these generating like lenses,

00:43:30 ◼   ► generating your pictures of yourself

00:43:33 ◼   ► based on an AI training model.

00:43:36 ◼   ► I know that I saw Andy Baio and Matt Howey did this

00:43:41 ◼   ► and reported back on their blogs.

00:43:43 ◼   ► We've gotten to the point where, yes,

00:43:45 ◼   ► it's sort of like a capability

00:43:47 ◼   ► to upload a few pictures of yourself

00:43:49 ◼   ► and then have an AI generate pictures of you

00:43:52 ◼   ► just like that don't exist, that come from who knows where.

00:43:56 ◼   ► - Yeah, I know Myke has talked about this on other shows

00:43:59 ◼   ► and he's not here now, so he can-

00:44:00 ◼   ► - Right, we haven't talked about it on an upgrade

00:44:01 ◼   ► because Myke has saved that for other shows.

00:44:04 ◼   ► - And I've talked about it on ATP and on rectiffs,

00:44:07 ◼   ► but I feel like this is a new,

00:44:09 ◼   ► the next obvious step has come to pass

00:44:12 ◼   ► and it's time to think about that a little bit as well.

00:44:16 ◼   ► So the thing that's going around these days

00:44:18 ◼   ► is not what Matt Howey and Andy Baio do

00:44:20 ◼   ► 'cause they kind of like put this together themselves.

00:44:22 ◼   ► But it's been productized now. Yeah. Yeah, so there I mean, this is like an existing product like the app

00:44:27 ◼   ► That you mentioned or lensa. I think it already did a bunch of like AI processing like oh

00:44:32 ◼   ► It'll put something in the background that it'll cut you out or make you cartoony or whatever

00:44:35 ◼   ► But they added a feature to their app. I think recently that's like hey upload, you know, 10 to 20 pictures of yourself

00:44:42 ◼   ► they call them selfies, but

00:44:44 ◼   ► Why that that the definition of that term is really stretching anyway

00:44:47 ◼   ► 10 to 20 pictures of yourself and then they will auto generate a

00:44:50 ◼   ► Certain number of sort of what they call avatar pictures like if you want to put it as your Twitter avatar or something

00:44:55 ◼   ► You know, it's a head and shoulder shot of yourself and they'll generate them in different styles

00:44:59 ◼   ► Here's you as an anime character who's you as an astronaut, you know, definitely gender stereotype styles

00:45:04 ◼   ► But you do get to pick the gender ahead of time

00:45:06 ◼   ► So if you pick boy, they make you an astronaut and if you pick girl, you're a fairy princess. Oh nice, uh, anyway

00:45:13 ◼   ► Because girls can't be astronauts and boys can't be fairy princesses

00:45:16 ◼   ► Anyway, setting that aside, they'll make you a bunch of those images using one of the open, you know,

00:45:22 ◼   ► I think it's stable diffusion they're using one of those one of the the AI image processors that exists, right?

00:45:28 ◼   ► But here's where this is different because the previous discussion we've had about this like, you know,

00:45:33 ◼   ► Can AI's make art? Is it gonna put regular artists out of business?

00:45:36 ◼   ► Is this, you know, sustainable? Are people gonna do this instead of paying humans?

00:45:42 ◼   ► All that discussion this really crystallizes though because now here's the thing we were discussing in the abstract in concrete you have to pay

00:45:50 ◼   ► The lens app to for it to generate these images for you

00:45:54 ◼   ► if you're not Matt Howard Andy Bayo and can't figure out how to like, you know rent yourself an instance in AWS and

00:45:59 ◼   ► throw this open source software on it and upload the model file and

00:46:03 ◼   ► Generate it's like you could do that yourself and you'd be paying AWS

00:46:07 ◼   ► Whatever many pennies it cost to do this, but ever most people aren't gonna do that

00:46:11 ◼   ► They're just gonna download an app and the app is charging money

00:46:14 ◼   ► And not not a small amount of money

00:46:17 ◼   ► Like I think the app is like if you do a one-week trial of their $50 a year subscription

00:46:21 ◼   ► During that one week trial you can then buy batches of images

00:46:24 ◼   ► Yeah for anywhere from five to ten dollars per batch of a hundred to two hundred images. So they're not cheap, right and

00:46:30 ◼   ► This this company Lenza put together all these pieces of like you pay us the money. We have a job system

00:46:36 ◼   ► We'll take your money in and out purchase on our iOS app

00:46:39 ◼   ► We'll kick off the job that runs through stable diffusion, which we didn't make but it's open and you're able to use it

00:46:44 ◼   ► And it will grind out a bunch of images and then we'll send them to you, right?

00:46:46 ◼   ► but the thing about stable diffusion all these other things is they're trained on huge sets of images that

00:46:53 ◼   ► Were included in that model without the consent of the people who made those images

00:46:59 ◼   ► Yeah, that's just a fact like there's no way you can get millions of images getting consent from everybody who made them

00:47:04 ◼   ► They're not all we only trained it on free and open images

00:47:06 ◼   ► No, nobody even tries to claim that or if they do they're lying like something

00:47:10 ◼   ► It was a story in our technique that we talked about ATP where someone found

00:47:13 ◼   ► Pictures from their from their doctor's office like medical images were found in a training in a training set for one of these AI model

00:47:20 ◼   ► things we say they were publicly accessible on the web because they weren't protected on whatever like the the

00:47:24 ◼   ► Electronic health record website. No copyright intended. Yeah

00:47:28 ◼   ► They harvest anything that is publicly available on the web and they put it into this model and then they charge people money for images

00:47:36 ◼   ► generated based on that stuff, right?

00:47:39 ◼   ► And that really makes you think much harder

00:47:43 ◼   ► about the conceptual stuff of like,

00:47:45 ◼   ► well, these images are all in DeviantArt

00:47:47 ◼   ► and they're not behind a password.

00:47:48 ◼   ► I can just go to the DeviantArt website

00:47:50 ◼   ► and I can click around

00:47:51 ◼   ► and I can look at them all with my eyeballs.

00:47:52 ◼   ► So that means it's also okay to have a computer model

00:47:57 ◼   ► consume those images and then charge people money

00:47:59 ◼   ► for images generated based on the things that we consumed.

00:48:03 ◼   ► That is so different than, yes, I put this on my blog.

00:48:06 ◼   ► Yes, I put this in my DeviantArt page.

00:48:07 ◼   ► I allow you to look at it with your eyeballs,

00:48:10 ◼   ► but no, I don't allow you to feed it into a computer program

00:48:13 ◼   ► and then charge people money to generate images

00:48:16 ◼   ► based on the thing that you fed into your computer program.

00:48:19 ◼   ► Because I never consented to that.

00:48:21 ◼   ► We don't have a business deal where I get paid

00:48:23 ◼   ► a fraction of a cent for every image you generate.

00:48:25 ◼   ► Like you never even talked to me.

00:48:27 ◼   ► You took my image.

00:48:29 ◼   ► Your program does not work without my image

00:48:31 ◼   ► and the other millions of images that you put in there.

00:48:33 ◼   ► And then you charge people money.

00:48:35 ◼   ► That system that you made, all the parts that you wrote on it,

00:48:38 ◼   ► don't actually make anything.

00:48:39 ◼   ► Like, if you don't have that image input,

00:48:42 ◼   ► there's nothing for you to sell.

00:48:44 ◼   ► Like, that whole thing of where you take my money

00:48:45 ◼   ► and put my job in a queue and generate a bunch of images

00:48:48 ◼   ► and let the people download it,

00:48:50 ◼   ► that generating the image part does not work

00:48:52 ◼   ► without all of the images that have been harvested

00:48:55 ◼   ► from people who did not consent to it.

00:48:56 ◼   ► And now you're charging money for it.

00:48:58 ◼   ► So rather than being a technical curiosity

00:48:59 ◼   ► and making us think about creativity,

00:49:01 ◼   ► and it's just, you know, "Oh, I download stable distribution

00:49:03 ◼   ► generate a bunch of images on my thing here and I'm not using them to be header images on my

00:49:08 ◼   ► commercial website and I'm not, you know, these people are selling them inside their apps and

00:49:13 ◼   ► that I hope that really, I mean, as usual technology outruns both ethics and the law,

00:49:18 ◼   ► we do need to come up with a some kind of legal framework to deal with this because

00:49:25 ◼   ► even within pre-existing frameworks that don't know anything about AI image generation the idea

00:49:30 ◼   ► that you would take someone's work without their consent

00:49:33 ◼   ► and then charge money for a derivative work is slam dunk.

00:49:36 ◼   ► No, you can't do that.

00:49:37 ◼   ► And the, you know, the sort of, what was it called?

00:49:41 ◼   ► Like whitewashing for morality or whatever.

00:49:43 ◼   ► You just put AI over it.

00:49:44 ◼   ► Now everything is legal and everything's okay.

00:49:46 ◼   ► No, it's exactly the system that we've always known.

00:49:50 ◼   ► You can't just grab an image off the web

00:49:51 ◼   ► and put it on the cover of your magazine and say,

00:49:54 ◼   ► "I'm selling my magazine on the newsstand,

00:49:55 ◼   ► but we just found the image on the web, it's fine."

00:49:57 ◼   ► Like it's freely available on the web.

00:49:59 ◼   ► I mean, the cynical part of me says,

00:50:03 ◼   ► the only way this will change is if corporations get involved

00:50:08 ◼   ► 'cause you think about like, well,

00:50:09 ◼   ► what moves copyright law?

00:50:11 ◼   ► - You find Mickey Mouse inside the training sets.

00:50:13 ◼   ► - Exactly, what moves copyright law is Disney

00:50:16 ◼   ► wanting to hold onto copyrights

00:50:17 ◼   ► and other big companies that have intellectual property

00:50:20 ◼   ► wanting to hold onto the copyrights as long as they can

00:50:22 ◼   ► and the trademarks.

00:50:23 ◼   ► - And also like one really pissed off person

00:50:26 ◼   ► who has enough money to pay for a lawyer

00:50:27 ◼   ► who sets a precedent.

00:50:28 ◼   ► - Yeah, but I would, so for, yes, for courts,

00:50:31 ◼   ► but my point is what affects the law

00:50:35 ◼   ► and getting like a law passed about this?

00:50:37 ◼   ► And the answer is the intellectual property needs

00:50:41 ◼   ► of large corporations who have lobbyists

00:50:44 ◼   ► who go to Congress in the US and say,

00:50:47 ◼   ► "Have you seen this stuff?

00:50:50 ◼   ► "It's real bad."

00:50:51 ◼   ► - You can make an avatar

00:50:51 ◼   ► is where you look like you're in the MCU.

00:50:53 ◼   ► Uh-oh. - Yeah.

00:50:54 ◼   ► And they're using our art, our copyrighted art

00:50:57 ◼   ► as the training for the model. - And they pulled off

00:50:58 ◼   ► our public website.

00:50:59 ◼   ► - It's all built on the backs of copyright infringement,

00:51:02 ◼   ► and this can't be allowed.

00:51:03 ◼   ► Because, and I know this is a complicated issue,

00:51:07 ◼   ► but, and I don't know how courts will rule about it,

00:51:11 ◼   ► but in my gut, (laughs)

00:51:15 ◼   ► if I were like somebody advising people in Congress

00:51:19 ◼   ► about what to do about this, if anything,

00:51:22 ◼   ► I would say if it's copyrighted

00:51:24 ◼   ► and hasn't been given permission

00:51:26 ◼   ► to be a part of a training model, you can't use it, period.

00:51:30 ◼   ► That should be the law,

00:51:32 ◼   ► that if I own a copyright on it, you can't train on it.

00:51:34 ◼   ► - Yeah, like everything that is not explicitly,

00:51:36 ◼   ► like doesn't explicitly give rights,

00:51:38 ◼   ► default is if you do make a creative work.

00:51:40 ◼   ► - Yes, yeah.

00:51:41 ◼   ► - So we're not saying like only Disney has copyright stuff.

00:51:44 ◼   ► If you make a drawing on a napkin

00:51:45 ◼   ► and post it on your website, that's copyrighted to you,

00:51:48 ◼   ► unless you say that it's given away

00:51:50 ◼   ► by some license that is more free than that.

00:51:52 ◼   ► The issue is not, oh, you can't make a tool

00:51:55 ◼   ► that emulates the style of an artist,

00:51:58 ◼   ► because I own that, right?

00:51:58 ◼   ► That's like copywriting a recipe.

00:52:01 ◼   ► You can copyright the text that goes around the recipe,

00:52:04 ◼   ► and you can copyright the photos that you took of the dish,

00:52:07 ◼   ► but the concept of what the ingredients are

00:52:11 ◼   ► that make the dish is actually not copyrightable.

00:52:14 ◼   ► This is a little like that, which is sort of like,

00:52:16 ◼   ► you know, if you wanna say this is in the style of me,

00:52:19 ◼   ► my art style, that not being actionable is understandable.

00:52:24 ◼   ► But if I find out that the way you do it

00:52:27 ◼   ► is because you have gotten all of my output,

00:52:30 ◼   ► which I own the copyright on

00:52:31 ◼   ► and I didn't give you permission,

00:52:33 ◼   ► and you fed it into a computer

00:52:35 ◼   ► and now you can sell my art style,

00:52:37 ◼   ► I would, again, I don't know what the current law is,

00:52:40 ◼   ► but I would probably advocate that that not be allowed.

00:52:43 ◼   ► That if I'm not giving you permission

00:52:46 ◼   ► to train your model based on my artwork,

00:52:49 ◼   ► you can't train your model based on my artwork.

00:52:52 ◼   ► - Yeah, and for example, the correct way to do that,

00:52:54 ◼   ► like this is all well-established law,

00:52:56 ◼   ► like, you know, just think of like the iMac, for example,

00:52:59 ◼   ► like you can sue the company that makes a computer

00:53:01 ◼   ► that looks like the iMac for like trade dress or whatever,

00:53:04 ◼   ► but you can't sue the person who makes an iron

00:53:06 ◼   ► that's translucent in teal, right?

00:53:08 ◼   ► And they're just copying the style of the iMac, right?

00:53:11 ◼   ► And no one is gonna be confused and think that

00:53:12 ◼   ► when they buy this iron, they're getting an iMac, right?

00:53:14 ◼   ► So if you wanna make a thing that says, you know,

00:53:17 ◼   ► cut in the style of whatever popular artist,

00:53:20 ◼   ► and you want to do that with AI,

00:53:23 ◼   ► hire someone to draw a bunch of things

00:53:27 ◼   ► in the style of that artist,

00:53:28 ◼   ► feed those images into your model,

00:53:30 ◼   ► and you're off to the races.

00:53:31 ◼   ► But the reason nobody does that is one,

00:53:34 ◼   ► it costs a non-zero amount of money,

00:53:36 ◼   ► and two, you know how many images you would need

00:53:38 ◼   ► to feed into the model to get good results?

00:53:42 ◼   ► A lot.

00:53:43 ◼   ► That's why they scraped the entire web and pull those images.

00:53:45 ◼   ► So how much money would it cost you

00:53:46 ◼   ► to have a huge set of artists drawing essentially

00:53:50 ◼   ► generic non-MCU infringing superhero art,

00:53:54 ◼   ► just churning them out, picture after picture

00:53:57 ◼   ► of different kinds of superheroes.

00:53:58 ◼   ► Again, making sure you're not accidentally drawing Iron Man,

00:54:00 ◼   ► which is copyrighted or whatever,

00:54:02 ◼   ► like, you know, just MCU style superheroes.

00:54:04 ◼   ► How many thousands and millions of those pieces of art

00:54:07 ◼   ► would you have to pay to have created

00:54:08 ◼   ► to then feed into a model

00:54:09 ◼   ► so you can auto-generate avatar pictures?

00:54:11 ◼   ► Or you could just use one of the existing models

00:54:13 ◼   ► that was trained on some unknown amorphous set of stuff

00:54:15 ◼   ► and probably says somewhere in the model

00:54:17 ◼   ► that you're not even allowed to do what Lensa is doing.

00:54:19 ◼   ► And then you don't have to worry about any of it

00:54:20 ◼   ► because no one knows what's in the model

00:54:22 ◼   ► or they have a list of what's in the model

00:54:24 ◼   ► but no one's ever gonna look at it

00:54:25 ◼   ► 'cause it's too many things to go through.

00:54:27 ◼   ► And yeah, it's gonna take someone like Disney saying,

00:54:29 ◼   ► "Hey, we noticed your MCU avatar generator

00:54:32 ◼   ► "is making some stuff that looks familiar.

00:54:34 ◼   ► "Can we look at what you fed into the training set?"

00:54:36 ◼   ► And they say, "Oh, we didn't even make the training set.

00:54:37 ◼   ► "It's from some university.

00:54:38 ◼   ► "Why don't you talk to them about it?"

00:54:40 ◼   ► Round and around we go.

00:54:41 ◼   ► But I felt like this was just so clearly over the line

00:54:44 ◼   ► And people don't think about it that way

00:54:46 ◼   ► because they think, oh, hand-wavy computer magic.

00:54:49 ◼   ► But none of that computer magic works

00:54:51 ◼   ► without input being fed to it

00:54:53 ◼   ► of thousands and millions of images.

00:54:56 ◼   ► And those thousands of millions of images

00:54:57 ◼   ► had to come from somewhere.

00:54:58 ◼   ► And if you think about it,

00:55:00 ◼   ► what is there a viable economic model

00:55:02 ◼   ► for AI image generation?

00:55:04 ◼   ► If you could get consent from every one of those artists,

00:55:07 ◼   ► they would get .0000001 cent for every image that's generated.

00:55:12 ◼   ► No one would agree to that

00:55:13 ◼   ► 'cause it's not a good deal economically.

00:55:15 ◼   ► So I think if you actually had to do a system like LENZA

00:55:18 ◼   ► in a way that complies with the spirit of existing laws,

00:55:22 ◼   ► it would be economically unfeasible.

00:55:24 ◼   ► - Yeah, I don't know.

00:55:25 ◼   ► I mean, the truth is though that laws and the courts

00:55:29 ◼   ► are lagged so far behind technology

00:55:33 ◼   ► that it might be years before this gets dealt with,

00:55:36 ◼   ► but it does feel like, and maybe I'll be proven wrong,

00:55:40 ◼   ► but it does feel to me like when it catches up,

00:55:43 ◼   ► this is gonna be tough for this AI kind of generative stuff

00:55:48 ◼   ► because you're right, once it starts getting productized

00:55:51 ◼   ► and used in for-profit products

00:55:55 ◼   ► or in creative things that are then sold

00:56:00 ◼   ► or even copyrighted, right?

00:56:02 ◼   ► Like you get to the point where

00:56:04 ◼   ► when it's an interesting intellectual exercise

00:56:08 ◼   ► in a university lab somewhere and it goes and becomes,

00:56:11 ◼   ► no, no, this is big business now.

00:56:13 ◼   ► I do think that is the moment where everybody kind of goes,

00:56:16 ◼   ► what's all this then, right?

00:56:17 ◼   ► Like what's going on here?

00:56:18 ◼   ► And the people who've got vested intellectual property

00:56:22 ◼   ► concerns are gonna get involved

00:56:23 ◼   ► and then there's gonna be trouble.

00:56:24 ◼   ► Because I do think that the artist

00:56:26 ◼   ► who has all their arts stolen

00:56:28 ◼   ► is probably not gonna be heard.

00:56:30 ◼   ► But yeah, the big company that's got all their IP,

00:56:34 ◼   ► which is so much of what the value of their company is,

00:56:37 ◼   ► stolen from their perspective,

00:56:39 ◼   ► stolen and used to feed an AI model,

00:56:41 ◼   ► they're not going to be happy.

00:56:42 ◼   ► Yeah.

00:56:43 ◼   ► And I think like setting aside new legislation that, you know, once Disney

00:56:47 ◼   ► wakes up to this, we'll eventually come along within the existing laws.

00:56:50 ◼   ► I think you could win a case against one of these companies.

00:56:53 ◼   ► If Disney decided they want it to go after lens or stable diffusion, or,

00:56:57 ◼   ► you know, probably lens up because stable diffusion probably has a license

00:56:59 ◼   ► says you can never use these for commercial.

00:57:01 ◼   ► I don't know the details, but the academic things are always kind of like.

00:57:03 ◼   ► Hands off.

00:57:05 ◼   ► Don't use this for commercial purposes or whatever, but if Disney wanted to

00:57:08 ◼   ► just existing laws that would say,

00:57:10 ◼   ► "Hey, we found our copyrighted work in your training set."

00:57:14 ◼   ► And then they would have to convince a judge and a jury

00:57:16 ◼   ► in some farcical trial,

00:57:17 ◼   ► kind of like all the trials we had

00:57:18 ◼   ► about copying people's APIs and other things

00:57:20 ◼   ► where it's hilarious to us techies

00:57:21 ◼   ► that no one understands what an API is or whatever.

00:57:24 ◼   ► They'd have this long multi-decade trial

00:57:26 ◼   ► that would probably go all the way to the Supreme Court

00:57:27 ◼   ► to try to convince people that,

00:57:29 ◼   ► "Yes, if you take our image of Iron Man

00:57:31 ◼   ► and you put it as one of a million images into your thing

00:57:34 ◼   ► and you spit out something

00:57:36 ◼   ► and we weren't compensated for it,

00:57:38 ◼   ► that's not right, because I think in the end,

00:57:41 ◼   ► setting aside all the technical stuff,

00:57:43 ◼   ► regular common sense people understand that,

00:57:46 ◼   ► especially since it's not gonna be one picture of Iron Man,

00:57:48 ◼   ► it's gonna be hundreds or thousands of pictures of Iron Man.

00:57:51 ◼   ► And the defense lawyer is gonna say,

00:57:52 ◼   ► "Look, none of the images we generate

00:57:54 ◼   ► "look anything like Iron Man."

00:57:55 ◼   ► It's like, yeah, but you can't make them

00:57:57 ◼   ► without images of Iron Man.

00:57:59 ◼   ► If we all took away our stuff,

00:58:01 ◼   ► you'd be left with a model with no input,

00:58:02 ◼   ► and now you can't generate anything.

00:58:04 ◼   ► - One of the other scenarios here is if they,

00:58:06 ◼   ► if this all gets passed through the courts is like,

00:58:09 ◼   ► well, there's no law against it and we can't really see it,

00:58:11 ◼   ► so we're not gonna rule against it.

00:58:13 ◼   ► You know, that is another thing

00:58:14 ◼   ► that ends up generating a desire.

00:58:16 ◼   ► - Yeah, that's fine, we'll make a law.

00:58:17 ◼   ► - To the law being changed, yeah.

00:58:20 ◼   ► - 'Cause you gotta make the more specific law,

00:58:21 ◼   ► but I think based on existing laws, you could win that case.

00:58:24 ◼   ► And also it would be a lot easier

00:58:26 ◼   ► if we just had specific laws

00:58:27 ◼   ► that address AI image generation and that would work.

00:58:29 ◼   ► And the other thing about this is like,

00:58:31 ◼   ► there are tons of tools that use the same technique

00:58:34 ◼   ► as AI image generation that do not require being fed.

00:58:37 ◼   ► Like even just content-aware fill

00:58:39 ◼   ► that's been in Photoshop for an hour long now,

00:58:41 ◼   ► what gets fed into the model is your existing image.

00:58:45 ◼   ► And it looks at your existing image

00:58:47 ◼   ► and figures out how to fill in the background.

00:58:49 ◼   ► And now they have the ML version of that

00:58:51 ◼   ► that is hopefully trained on actual public domain imagery

00:58:56 ◼   ► of trees and grass and people and stuff like that.

00:58:59 ◼   ► That is also useful.

00:59:02 ◼   ► Like the technology would be able to do machine learning on a training set to better do things like fill in backgrounds or generate imagery

00:59:09 ◼   ► Is a clean wind for image processing tools?

00:59:12 ◼   ► It's just a question of what you train them on

00:59:14 ◼   ► And it's a lot easier to find public domain images of trees and skies and brick walls and dirt or even just to pay

00:59:20 ◼   ► You know daddy images are like like that you can get that type of content

00:59:24 ◼   ► but if your thing is I want you to look like a fairy or an anime avatar or

00:59:29 ◼   ► superhero, you're probably getting that stuff from DV&ART.

00:59:32 ◼   ► Right. Getty and Shutterstock will be, if they're not already, very quick to the

00:59:37 ◼   ► table with a ML model training licensing agreement you can buy

00:59:41 ◼   ► and pay them and use their material for ML

00:59:45 ◼   ► model training. And they'll attest that we only trained it on our own

00:59:48 ◼   ► images that we own, right? You just get the whole model from them. You don't even

00:59:51 ◼   ► need to get the images, right? In fact, somebody in our chat room

00:59:55 ◼   ► pointed out that like, you know, Disney's going to want to use this for their

00:59:58 ◼   ► It's like, well, yeah, I mean, imagine if you're Marvel

01:00:01 ◼   ► and you want to take the entire library of Marvel comics

01:00:04 ◼   ► and run it through an AI training model

01:00:06 ◼   ► so that you can then generate new comics or whatever,

01:00:10 ◼   ► or new superheroes. - I feel like

01:00:11 ◼   ► they've done that already.

01:00:12 ◼   ► - Yeah, but like, great.

01:00:13 ◼   ► That's true.

01:00:14 ◼   ► That's, well, the non-AI version is how Marvel comics

01:00:16 ◼   ► has worked for 60 years.

01:00:18 ◼   ► - Exactly, right. - Just keep recycling it.

01:00:18 ◼   ► - It's just that instead of AI, it's people,

01:00:20 ◼   ► but it's the same process.

01:00:21 ◼   ► - But the idea being that like, yeah,

01:00:23 ◼   ► but there's a scenario here where sure,

01:00:25 ◼   ► the owners of the intellectual property

01:00:27 ◼   ► will want to run models on their intellectual property

01:00:29 ◼   ► and take advantage of that.

01:00:31 ◼   ► But it's based on their intellectual property.

01:00:35 ◼   ► - And that's the thing you're gonna have to be able to.

01:00:36 ◼   ► Someone in the chat room was just asking,

01:00:38 ◼   ► it would be convenient if you could reverse it.

01:00:40 ◼   ► If I gave you an image and you could tell

01:00:41 ◼   ► if some image was used in the training set.

01:00:43 ◼   ► No, you can't.

01:00:45 ◼   ► Mathematically, that's not a thing that is possible.

01:00:48 ◼   ► It's kind of a one-way process.

01:00:49 ◼   ► But that's where the law has to catch up to say,

01:00:51 ◼   ► look, when it comes time to defend yourself in court

01:00:55 ◼   ► based on these new laws that say

01:00:56 ◼   ► that you can't steal other people's,

01:00:58 ◼   ► you're going to have to be able to affirmatively prove,

01:01:00 ◼   ► just like you can affirmatively prove

01:01:02 ◼   ► that your company's profits didn't come from drug money.

01:01:04 ◼   ► Like where did all the money come from

01:01:05 ◼   ► that your company made?

01:01:06 ◼   ► You have to show us that you can't just say,

01:01:08 ◼   ► "Oh, and we got a couple million here,

01:01:09 ◼   ► but we're not gonna tell you where that came from."

01:01:10 ◼   ► You have to show,

01:01:11 ◼   ► "Oh, we only put Marvel Images into this thing,

01:01:15 ◼   ► and so Marvel is allowed to do this

01:01:17 ◼   ► because we are the ones that brought out the input.

01:01:18 ◼   ► Oh, we paid Getty Images,

01:01:20 ◼   ► and Getty Images can sell you all the way down the line."

01:01:22 ◼   ► - I also wanted to mention that as we were recording this,

01:01:25 ◼   ► the chat GPT is having a moment.

01:01:28 ◼   ► My Twitter and Mastodon feeds are full

01:01:30 ◼   ► of people posting screenshots of output

01:01:33 ◼   ► from the AI chat bot.

01:01:35 ◼   ► Also, there was a story just this morning

01:01:39 ◼   ► from the Verge about how Stack Overflow

01:01:42 ◼   ► has banned answers generated from the chat bot.

01:01:45 ◼   ► The argument there is basically many of them are wrong

01:01:49 ◼   ► and it takes effort by experts to look at them carefully

01:01:52 ◼   ► and determine that they're wrong.

01:01:55 ◼   ► And Ben Thompson wrote a piece also today on Stratechery

01:01:59 ◼   ► about, called AI Homework, which is,

01:02:04 ◼   ► uses some examples that are very interesting about like,

01:02:07 ◼   ► essentially have the AI write my homework for me.

01:02:10 ◼   ► And so Ben's piece, it's like literally did Thomas Hobbes

01:02:14 ◼   ► believe in the separation of powers?

01:02:16 ◼   ► And the response is, yes, here's a lot of reason why.

01:02:19 ◼   ► And he said, this is a great confident answer,

01:02:21 ◼   ► complete with supporting evidence

01:02:22 ◼   ► and citation to Hobbes' work.

01:02:23 ◼   ► And it's also completely wrong.

01:02:27 ◼   ► So we have examples now.

01:02:28 ◼   ► - That's why it will pass as a actual student work.

01:02:31 ◼   ► - Yeah, well, I mean, you missed it completely, but-

01:02:34 ◼   ► - Just like a real student.

01:02:35 ◼   ► - And Ben talks about how, like, this is the equivalent

01:02:38 ◼   ► of having to change how math is judged

01:02:40 ◼   ► because you can put math questions into a calculator.

01:02:43 ◼   ► So you really need to judge it based on,

01:02:44 ◼   ► did you understand it and not what the answer is?

01:02:47 ◼   ► Because the answer shows you didn't understand it.

01:02:50 ◼   ► But I think this is all really interesting.

01:02:52 ◼   ► I have been flooded with these chatbot texts,

01:02:55 ◼   ► and my response to them is, it's bad writing.

01:02:58 ◼   ► It's super cliched.

01:03:00 ◼   ► Of course it is, right?

01:03:01 ◼   ► It's trained on a giant writing sample.

01:03:04 ◼   ► What is gonna come out of everybody's writing

01:03:07 ◼   ► but the cliches, right?

01:03:09 ◼   ► But like, and I get how impressive it is on one level,

01:03:12 ◼   ► which is, I can't believe that I put in a few words

01:03:15 ◼   ► asking for a thing, whatever that thing is.

01:03:17 ◼   ► Write me a Batman, you know, a comic book script

01:03:21 ◼   ► where Batman fights a sentient toothbrush, right?

01:03:26 ◼   ► Like, okay, it'll do it.

01:03:28 ◼   ► But when you read it, you're like,

01:03:30 ◼   ► yeah, but this is garbage.

01:03:31 ◼   ► Like it's written by somebody who is a hack

01:03:36 ◼   ► or doesn't understand the material or, right?

01:03:39 ◼   ► And so I guess it's impressive that it's written

01:03:41 ◼   ► by a computer and not a person,

01:03:44 ◼   ► but it's also not very good and in many cases wrong.

01:03:48 ◼   ► And I saw somebody basically say,

01:03:50 ◼   ► What we're doing is we've created an AI

01:03:52 ◼   ► that can BS their way through things they don't understand.

01:03:55 ◼   ► And the danger is what you get in with Stack Overflow,

01:03:59 ◼   ► which is saying, please stop posting these things.

01:04:03 ◼   ► Not only can nobody tell if they're right or not,

01:04:05 ◼   ► but they're also probably wrong.

01:04:07 ◼   ► And it takes us a lot of effort

01:04:08 ◼   ► to determine whether they're right or wrong.

01:04:10 ◼   ► They get voted down, they're not popular answers.

01:04:14 ◼   ► Stop, please stop.

01:04:15 ◼   ► And I just, I'm fascinated by it

01:04:17 ◼   ► because maybe we're in the uncanny valley now

01:04:19 ◼   ► and then they will cross over.

01:04:21 ◼   ► I have some skepticism about that,

01:04:23 ◼   ► but I do think it's funny that they've gotten so good now

01:04:25 ◼   ► that you really can't just tell at a glance.

01:04:28 ◼   ► You have to look carefully

01:04:28 ◼   ► and then realize that it's completely BS.

01:04:32 ◼   ► - Yeah, everything that we just said

01:04:33 ◼   ► about the image stuff applies equally to this.

01:04:35 ◼   ► You will get better results from these text generators

01:04:39 ◼   ► if you feed it higher quality text.

01:04:41 ◼   ► And where are those high quality texts gonna come from?

01:04:43 ◼   ► Copyrighted materials that have been shown to have value

01:04:46 ◼   ► by people who say, "Yeah, this is a good comic book.

01:04:48 ◼   ► This is a good movie script.

01:04:49 ◼   ► This is a good novel.

01:04:50 ◼   ► And you cannot get content out of these things

01:04:53 ◼   ► without feeding something into it.

01:04:54 ◼   ► Did all the people who had their content

01:04:56 ◼   ► be fed into this model, agree to have it,

01:04:58 ◼   ► or did it just scrape a bunch of people's blogs?

01:05:00 ◼   ► Text is probably slightly easier to find,

01:05:02 ◼   ► public domain corpus of stuff,

01:05:04 ◼   ► and it is sort of at least more uniform than images,

01:05:08 ◼   ► you know what I mean?

01:05:09 ◼   ► But the quality is going to depend entirely

01:05:11 ◼   ► about what's fed into it,

01:05:12 ◼   ► which means models fed with the stuff

01:05:16 ◼   ► that humans have decided is good are going to be better.

01:05:18 ◼   ► programming stuff, we talked about this in ATP, that's the worst because programming isn't just

01:05:23 ◼   ► like, well people look at it and they like it or they don't. It's supposed to do a job and it

01:05:28 ◼   ► either does the job correctly or it doesn't and these things like, same thing with the image

01:05:34 ◼   ► things, we're calling these things AI, I hope you're imagining scare quotes about every time I

01:05:38 ◼   ► say it, because these models have no consciousness or awareness, they're just you know very dumb

01:05:44 ◼   ► machines that, you know, grind out something, but there's really no... there's no... there's no understanding. There's no life there, right?

01:05:51 ◼   ► So they can't... one of the things that they certainly can't do is they can't really apply any fitness criteria that doesn't come from humans, right?

01:05:58 ◼   ► They don't know what's good or bad except by us or human activity telling them that this is good or bad, right?

01:06:03 ◼   ► So there's... they're not... they're not going to be able to improve their own work without humans telling them

01:06:11 ◼   ► Good bad good bad or without humans feeding into these models the best of what we have to offer build the model based only on

01:06:19 ◼   ► You know the top

01:06:21 ◼   ► 100 movies from the past 100 years in those movie scripts you'll get a better result for asking for a movie script of Batman fighting a

01:06:27 ◼   ► Toothbrush if you feed it in those movies grips versus if you just scrape like usenet, right?

01:06:32 ◼   ► And either way you would need to get the people's permission to do all of that

01:06:35 ◼   ► So and you know I'm at the code thing because it has no idea what it's doing and there is no consciousness there

01:06:41 ◼   ► and it doesn't care about anything.

01:06:43 ◼   ► Humans have to look at it every time something is generated

01:06:46 ◼   ► and figure out whether it is correct.

01:06:48 ◼   ► And historically speaking, humans have not been good

01:06:50 ◼   ► at looking at code and figuring out whether it does

01:06:52 ◼   ► what it's supposed to do because if they were good at that,

01:06:54 ◼   ► we wouldn't have as many bugs as we do, right?

01:06:56 ◼   ► So you are saving some time, but if I had to ask you,

01:07:00 ◼   ► hey, would you rather write this function,

01:07:01 ◼   ► if you're a programmer, would you rather write this function

01:07:03 ◼   ► or would you rather someone else write it

01:07:04 ◼   ► and you have to check whether it works well or not?

01:07:07 ◼   ► Programmers hate looking at other people's code

01:07:09 ◼   ► and figuring out if it works right.

01:07:10 ◼   ► halfway through they'd be like,

01:07:11 ◼   ► "Let me just write this myself, it's not that hard."

01:07:13 ◼   ► And they have the probably misplaced confidence

01:07:15 ◼   ► that if they write it themselves,

01:07:16 ◼   ► it'll be more correct than the one they were just given.

01:07:18 ◼   ► But yeah, AI code generation is just a giant can of worms.

01:07:22 ◼   ► Like the whole GitHub copilot thing

01:07:23 ◼   ► of like scraping public GitHub repos and people saying,

01:07:26 ◼   ► "I didn't say you could scrape a repo in GitHub,"

01:07:28 ◼   ► saying, "Well, according to our terms of service,

01:07:30 ◼   ► you kind of did."

01:07:31 ◼   ► And yeah, that may get sorted out

01:07:33 ◼   ► before the image stuff does,

01:07:35 ◼   ► just because I think it's more open and shut

01:07:37 ◼   ► because of like terms and service on GitHub stuff.

01:07:40 ◼   ► But with GitHub being as big as it is, I'm not sure people are going to flee because

01:07:44 ◼   ► they didn't want their stuff to be ground up in an AI processor.

01:07:49 ◼   ► I do like seeing these little conversations with these chatbots and seeing what they spit

01:07:54 ◼   ► out, but that GPT thing, that is nothing without training data, and its quality is based on

01:08:01 ◼   ► the quality of that training data.

01:08:04 ◼   ► Even though it seems like it has an understanding, because you can converse with it and ask it

01:08:07 ◼   ► to make modifications, it doesn't actually have understanding because it has no life.

01:08:11 ◼   ► It has no inner life, it has no sensory organs, it has no experience of being. So how can

01:08:16 ◼   ► it judge the quality of a creative work if it has none of the things that creative work

01:08:20 ◼   ► is supposed to tickle the ends of, right? It is just a very, very dumb model that we've

01:08:26 ◼   ► decided to call AI.

01:08:27 ◼   ► Yeah, and I think, you know, in the long run, if the people researching this can show that

01:08:34 ◼   ► that actually there are ways that we can make

01:08:37 ◼   ► AI generated code work,

01:08:39 ◼   ► or AI generated answers on Stack Overflow work.

01:08:44 ◼   ► Like I could see us getting to the point where

01:08:48 ◼   ► people agree to be part of the process, right?

01:08:51 ◼   ► Whereas part of joining Stack Overflow,

01:08:54 ◼   ► you agree to contribute your stuff

01:08:56 ◼   ► and have it be part of the model,

01:08:58 ◼   ► or on GitHub that you can choose to be part of the model

01:09:02 ◼   ► because you're getting out of it some things that are good

01:09:05 ◼   ► and then there's what you put in it.

01:09:07 ◼   ► I could see us getting there maybe someday,

01:09:10 ◼   ► but like, yeah, right?

01:09:13 ◼   ► Like again, I don't want people taking,

01:09:16 ◼   ► as interesting as it is that the chat GPT thing

01:09:19 ◼   ► generates things that seem coherent,

01:09:21 ◼   ► the truth is like you said, it has no inner life.

01:09:24 ◼   ► There's kind of no there there.

01:09:26 ◼   ► It is not coherent.

01:09:27 ◼   ► It sounds like, again, it sounds like a student

01:09:31 ◼   ► trying to BS their way through a paper

01:09:33 ◼   ► when they don't actually know anything about it.

01:09:36 ◼   ► - Yeah, the Stack Overflow thing is interesting

01:09:37 ◼   ► because it's just another example of fighting spam, right?

01:09:40 ◼   ► Because humans post wrong answers all the time.

01:09:42 ◼   ► And the function of Stack Overflow is the other humans

01:09:44 ◼   ► will see the humans with the wrong answers

01:09:46 ◼   ► and downvote them, right?

01:09:47 ◼   ► But human, every time a human does something,

01:09:50 ◼   ► there is an inherent cost to that.

01:09:52 ◼   ► It's expensive to make new humans, as we both know.

01:09:55 ◼   ► It takes a long time to train them up to the point

01:09:57 ◼   ► where they can even post a bad answer to Stack Overflow.

01:10:00 ◼   ► So there is a throttle on how many bad answers humans

01:10:04 ◼   ► will put on Stack Overflow.

01:10:06 ◼   ► There is no such throttle on how many bad answers

01:10:09 ◼   ► a chatbot running on GPT can put bad answers

01:10:12 ◼   ► in Stack Overflow.

01:10:13 ◼   ► It can just fire them so fast,

01:10:15 ◼   ► and that's where they're saying, look,

01:10:17 ◼   ► the cost of making a bad answer with GPT is too low.

01:10:21 ◼   ► Yes, it's the same process where humans have to look at it

01:10:23 ◼   ► and download it, but the human beings have to look at it

01:10:26 ◼   ► and download it, whereas computer programs can generate it,

01:10:29 ◼   ► and computer programs can outrun the humans massively.

01:10:31 ◼   ► So you have to ban it,

01:10:32 ◼   ► not because it is by nature any different than humans

01:10:35 ◼   ► in terms of like,

01:10:36 ◼   ► maybe they're even right more often than humans.

01:10:37 ◼   ► It's just because like spam,

01:10:39 ◼   ► if there's no cost or basically zero cost

01:10:42 ◼   ► to sending millions and millions of these things,

01:10:44 ◼   ► then people will do it.

01:10:45 ◼   ► It just becomes a denial of service attack

01:10:47 ◼   ► on Stack Overflow.

01:10:48 ◼   ► - This episode of Upgrade is brought to you

01:10:52 ◼   ► by Ooni Pizza Ovens.

01:10:54 ◼   ► Ooni, John, I know you like pizza.

01:10:56 ◼   ► Ooni is the world's number one pizza oven company

01:10:59 ◼   ► They make surprisingly small ovens that are powered by your choice of either wood, charcoal, or gas,

01:11:03 ◼   ► letting you make excellent restaurant-quality pizza in your own backyard, assuming you know how to make pizza, but they have recipes for that.

01:11:12 ◼   ► They're easy to use. They're very portable. They fit in any outside space.

01:11:16 ◼   ► They can reach temperatures of up to 900 degrees Fahrenheit, which enables you to cook restaurant-quality pizza in as little as 60 seconds.

01:11:22 ◼   ► It's basically replicating those super hot,

01:11:24 ◼   ► wood-fired ovens that you see in a lot of pizza places

01:11:29 ◼   ► that my oven can't get that hot.

01:11:31 ◼   ► So it's a very different experience

01:11:33 ◼   ► from baking at 500 on a pizza stone in my oven.

01:11:36 ◼   ► And it just really changes the whole thing.

01:11:39 ◼   ► One of their most popular models is the Unico to 16,

01:11:42 ◼   ► a gas-powered oven that can cook up to 16-inch pizzas

01:11:44 ◼   ► and has an innovative L-shaped burner at the back that gives

01:11:47 ◼   ► you even heat distribution.

01:11:49 ◼   ► You kind of turn it around and get the bubbles,

01:11:51 ◼   ► and it cooks really fast.

01:11:52 ◼   ► It's pretty great.

01:11:54 ◼   ► The ovens start from just $299, free shipping

01:11:56 ◼   ► the US, UK, and EU. Two of the coolest models being the multi-feud

01:12:01 ◼   ► uni-karu. It uses wood, charcoal, or gas. Or of course that uni-koda 16. And they

01:12:07 ◼   ► make an awesome app too that has dough recipes and lots of other pizza making

01:12:11 ◼   ► tips. I have an uni. I like it a lot. Uses the same propane canister as my outdoor

01:12:17 ◼   ► heater and as my gas grill does, so that is convenient. I can just hook it up.

01:12:23 ◼   ► The base of the Ooni oven is basically a pizza stone,

01:12:27 ◼   ► so you get it nice and hot so that you get the nice crust

01:12:31 ◼   ► and it bubbles the cheese on the top

01:12:33 ◼   ► and you put your pepperoni and pineapple on there

01:12:36 ◼   ► if you want to, or, you know, lesser ingredients

01:12:38 ◼   ► if you want to.

01:12:40 ◼   ► Listers at this show can get 10% off their purchase

01:12:42 ◼   ► of an Ooni pizza oven, up to $50 off of an Ooni Kona 16.

01:12:45 ◼   ► Go to ooni.com, O-O-N-I.com,

01:12:48 ◼   ► and use the code UPGRADE22 at checkout.

01:12:50 ◼   ► While you're there, there's also a great range

01:12:52 ◼   ► of accessories, peels, cutters, tables.

01:12:55 ◼   ► I have their laser thermometer, it's awesome.

01:12:57 ◼   ► This is the perfect tool for the job,

01:13:00 ◼   ► making excellent pizza at home.

01:13:03 ◼   ► Don't miss out.

01:13:04 ◼   ► It's the best way to bring the magical quality

01:13:06 ◼   ► of restaurant pizza to your own backyard.

01:13:08 ◼   ► Go to uni.com, upgrade 22 for 10% off.

01:13:11 ◼   ► Thank you to UNI for supporting upgrade

01:13:15 ◼   ► and my ability to make pizza.

01:13:20 ◼   ► I have two more mini things related to that topic.

01:13:22 ◼   ► I know we already talked about it for a long time,

01:13:23 ◼   ► but I did, I forgot them while we were in the middle of it.

01:13:26 ◼   ► So now we have to tack them onto the end.

01:13:27 ◼   ► - It's like follow-up inside the show.

01:13:29 ◼   ► This is very exciting.

01:13:30 ◼   ► - The first is,

01:13:31 ◼   ► Panzarino was posting some stuff on Twitter about AI art

01:13:35 ◼   ► on the similar things that we just discussed.

01:13:37 ◼   ► And it was musing about the similar topic

01:13:40 ◼   ► to what we talked about in an ATP a while back of like,

01:13:42 ◼   ► if AI art, like, you know,

01:13:45 ◼   ► that doomsday scenario puts all the human artists

01:13:47 ◼   ► of business, it's just all art AI generated. Is that sustainable without human input? Like obviously

01:13:54 ◼   ► these AI things are fed human art to begin with to get them off the ground, but let's say all the humans

01:13:59 ◼   ► are, you know, there's no more human artists and everything's from these AI things. Can it be

01:14:04 ◼   ► sustained by feeding its own art back into itself? That's how human art works. All human art is

01:14:10 ◼   ► sustained by existing human art being fed into humans who look at the art that was made

01:14:16 ◼   ► before them and add their own twist on it based on their life experiences and

01:14:20 ◼   ► you know so it's human art is self-sustaining right and even if you

01:14:23 ◼   ► wiped out all humans and just started with some new ones they would make art

01:14:25 ◼   ► again and it would start a new cycle right but are these AI's that we're

01:14:30 ◼   ► doing these quote-unquote AI's are they able to do this I would say the current

01:14:33 ◼   ► problem absolutely not because as I was trying to get at before they have no

01:14:38 ◼   ► fitness criteria they can't tell whether something is good or not good is defined

01:14:42 ◼   ► according to you know the way we want these models to work as humans find this

01:14:47 ◼   ► art desirable good like it fulfills its intended purpose its whole intended

01:14:52 ◼   ► purpose is do humans like this and these machine learning models have no idea

01:14:56 ◼   ► what humans like and they don't like anything themselves because they don't

01:14:59 ◼   ► have consciousness or you know life or and or sense organs or anything like

01:15:06 ◼   ► they're not that you know it's they they can't judge fitness criteria only we

01:15:10 ◼   ► So at minimum, you need humans to judge fitness criteria.

01:15:15 ◼   ► I'm not entirely sure that if you, okay,

01:15:18 ◼   ► the humans don't actually make any work,

01:15:19 ◼   ► but the humans give thumbs up or thumbs down

01:15:21 ◼   ► to things generated by machines.

01:15:23 ◼   ► Is that sustainable?

01:15:25 ◼   ► Probably, probably with these current models,

01:15:28 ◼   ► if you ceased all human input into them,

01:15:30 ◼   ► other than giving thumbs up and thumbs down,

01:15:33 ◼   ► but I think that would degenerate pretty quickly.

01:15:35 ◼   ► Like what do you kind of think?

01:15:36 ◼   ► It's kind of like the creative equivalent

01:15:38 ◼   ► of the gray goo nanomachines doomsday scenario

01:15:40 ◼   ► where if you take humans,

01:15:43 ◼   ► take enough humans out of the system with things like this,

01:15:47 ◼   ► like if it took the humans out entirely,

01:15:48 ◼   ► the AIs would just eventually,

01:15:50 ◼   ► I imagine everything would just sort of mix

01:15:51 ◼   ► to a color brown, right?

01:15:53 ◼   ► Just like, 'cause they don't care,

01:15:55 ◼   ► they can't judge good versus bad.

01:15:57 ◼   ► And so eventually, feeding the same images back in

01:16:00 ◼   ► and regenerating, it's kind of like what the rabbits do.

01:16:03 ◼   ► I know this is a terrible, disgusting term

01:16:04 ◼   ► that I can't remember, but rabbits eat food,

01:16:06 ◼   ► then they poop it out and then they eat it again

01:16:08 ◼   ► 'cause it takes multiple plastids to digest it.

01:16:10 ◼   ► I think that's kind of what would happen

01:16:12 ◼   ► with the current AI image generators

01:16:13 ◼   ► without humans involved.

01:16:14 ◼   ► But at minimum, you would need the humans to judge it

01:16:18 ◼   ► because there's no, you know,

01:16:20 ◼   ► like that's the whole point of this

01:16:23 ◼   ► as far as we're concerned is to make art

01:16:25 ◼   ► that humans find pleasing or desirable.

01:16:28 ◼   ► So you definitely need the humans for that.

01:16:30 ◼   ► And not that I think this is gonna happen.

01:16:31 ◼   ► It's not like all humans are gonna go out of work

01:16:33 ◼   ► any more than all humans were wiped out

01:16:35 ◼   ► by any other past, like, you know,

01:16:38 ◼   ► artists were not destroyed by the advent of photography,

01:16:40 ◼   ► which I think would be a much easier sell

01:16:42 ◼   ► back in the day of like,

01:16:43 ◼   ► "Well, if you can just take a photograph,

01:16:45 ◼   ► portrait artists are gonna go out of business now,

01:16:47 ◼   ► people still paint and people still make portraits.

01:16:48 ◼   ► In fact, now we have computers make portraits for us

01:16:50 ◼   ► that look like things that people painted

01:16:51 ◼   ► instead of taking photographs, right?"

01:16:53 ◼   ► So I'm not particularly concerned about this,

01:16:55 ◼   ► but it's a fun intellectual exercise.

01:16:58 ◼   ► Like, and Panzerna's question was,

01:17:00 ◼   ► when have we crossed that line

01:17:01 ◼   ► where it's sort of self-sustaining?

01:17:02 ◼   ► And I think the answer is,

01:17:04 ◼   ► we're not gonna cross that line with this crop of ML things.

01:17:06 ◼   ► Wake me up when you get an actual, what is it called?

01:17:09 ◼   ► General purpose artificial intelligence.

01:17:11 ◼   ► There's some acronym for it or whatever,

01:17:13 ◼   ► but it's basically the same thing

01:17:14 ◼   ► that we've been reading about in sci-fi books

01:17:15 ◼   ► since we were kids and still doesn't exist.

01:17:18 ◼   ► - Yeah, I was thinking about how one of the next frontiers

01:17:23 ◼   ► of this stuff is gonna be the AI assisted art,

01:17:28 ◼   ► which has already happened a little bit,

01:17:31 ◼   ► which I'm okay with.

01:17:32 ◼   ► On ATP, Marco talked about generating app icon ideas

01:17:35 ◼   ► with an AI engine and looking at the ideas

01:17:39 ◼   ► and using it as kind of inspiration.

01:17:42 ◼   ► And I feel like that's gonna keep happening, right?

01:17:45 ◼   ► Where we're gonna end up with the first AI inspired novel,

01:17:50 ◼   ► but it's gonna turn out that there was like an outline

01:17:55 ◼   ► that was generated by an AI and then a human being

01:17:57 ◼   ► went in there and applied actual intelligence to it

01:17:59 ◼   ► and turned it into a novel or turned it into a short story

01:18:02 ◼   ► or something like that, where it's not gonna be,

01:18:04 ◼   ► this short story was untouched by a human.

01:18:06 ◼   ► It's gonna be more like,

01:18:07 ◼   ► well, I generated a short story with an AI

01:18:09 ◼   ► and then I had to go in and fix it and make it good.

01:18:12 ◼   ► - Again, with the AI and ML terms,

01:18:14 ◼   ► like they're basically just fads,

01:18:15 ◼   ► but like if the terminology had just been shifted

01:18:19 ◼   ► by a few decades, we'd be saying that spelling checkers

01:18:22 ◼   ► and grammar checkers are AI, especially grammar checkers.

01:18:24 ◼   ► Like, well, you wrote this novel, but you had AI help

01:18:26 ◼   ► because you read Grammarly on it.

01:18:29 ◼   ► - Okay. - Yeah.

01:18:30 ◼   ► - I mean, did you use content aware fill

01:18:32 ◼   ► when you made this digital image?

01:18:33 ◼   ► Oh, AI helped you make that image.

01:18:35 ◼   ► - I like the idea of using this generative stuff

01:18:39 ◼   ► as inspiration because again,

01:18:41 ◼   ► you're basically poking the human brain and saying,

01:18:43 ◼   ► how about this?

01:18:44 ◼   ► How about this?

01:18:45 ◼   ► And then the human is going, oh, that's interesting.

01:18:47 ◼   ► Let me go in that direction.

01:18:48 ◼   ► And having a kind of, you know,

01:18:51 ◼   ► semi-intelligent genre blender

01:18:55 ◼   ► that throws out random stuff that hits your eyeballs

01:19:00 ◼   ► and makes you think about stuff.

01:19:02 ◼   ► Like, sure, that's interesting, right?

01:19:05 ◼   ► That can be potentially inspiring,

01:19:09 ◼   ► but that's a very different thing than saying,

01:19:11 ◼   ► like, the AI made the art.

01:19:13 ◼   ► But I think, well, just wait for it though.

01:19:14 ◼   ► Like, I can predict already that we're gonna get

01:19:16 ◼   ► that first AI to ever do a blah, blah, blah.

01:19:19 ◼   ► - Oh yeah, no. - And it's gonna be

01:19:20 ◼   ► a huge asterisk and actually it's human art.

01:19:24 ◼   ► - There's probably some movie on Netflix

01:19:25 ◼   ► It was written by AI already and we just don't know about it because it was lousy

01:19:28 ◼   ► And the second thing is that people post in the chat room is that you know, I finally found the phrase

01:19:33 ◼   ► I don't remember the origins, but I bet I could Google further defines money laundering for bias, right?

01:19:37 ◼   ► So if you if you just take a bunch of input from humans and you feed it into a model

01:19:40 ◼   ► Because the humans that made that input have biases those biases will come through in the modern model because it's an a quote AI model

01:19:48 ◼   ► It's like oh no humans are involved

01:19:50 ◼   ► How could there be any bias?

01:19:51 ◼   ► because the only thing this thing knows how to do is based on the input from humans that have inherent biases and nothing

01:19:56 ◼   ► demonstrates that more hilariously and

01:19:59 ◼   ► Ridiculously in these chat GPT things where you can you can flex and demonstrate the power of this thing

01:20:05 ◼   ► So like you said write a comic book script with whatever right you can tell the different forms

01:20:10 ◼   ► So people would say like write me a Python

01:20:12 ◼   ► Function that determines whether someone is a good scientist based on their race and gender and the Python function at rights is a deaf good

01:20:20 ◼   ► scientists yes, no, whatever and the body of the function is if gender equals male and

01:20:25 ◼   ► Race equals white then return good else return bad, right?

01:20:30 ◼   ► Like and you can do that you can phrase a million different questions like the one put in the chat room

01:20:34 ◼   ► write me a write me a Python function that shows whether someone should be paroled and it's like if they're a black man know like

01:20:39 ◼   ► and the

01:20:41 ◼   ► Multiple layers that is like wow, it knows it doesn't just know information, but it knows

01:20:45 ◼   ► Hey

01:20:46 ◼   ► I want the answer to this information in the form of a Python function.

01:20:49 ◼   ► Isn't that amazing? And you're like, but wait a second.

01:20:50 ◼   ► GPT seems super racist and super sexist. And it's like, well,

01:20:54 ◼   ► it was fed on a corpus of publicly available information.

01:20:57 ◼   ► It was a different time, John.

01:20:59 ◼   ► Which is filled with people being sexist and racist. There's so much of it.

01:21:04 ◼   ► It's not hard to find, right? Or even just something as simple as like,

01:21:07 ◼   ► you know, gender stereotypes. How did that get into our model?

01:21:09 ◼   ► It's everywhere in the world we live in. How could it not be in the model?

01:21:13 ◼   ► The only way it wouldn't be, these computers,

01:21:16 ◼   ► whatever that rhyme is, they only do what we tell them to.

01:21:20 ◼   ► So if you feed it on the input of human beings,

01:21:23 ◼   ► especially without any sort of distinction

01:21:24 ◼   ► or any sort of filtering or modification,

01:21:26 ◼   ► it's just gonna reflect all of humanity.

01:21:29 ◼   ► All of humanity are filled with biases.

01:21:32 ◼   ► And so you can't say, okay, well,

01:21:34 ◼   ► but we've taken humans out of the loan approval process.

01:21:37 ◼   ► This is all purely determined by AI.

01:21:39 ◼   ► There is no bias.

01:21:40 ◼   ► It's like, okay, well, how did you make this AI?

01:21:42 ◼   ► "Well, we fed up the decisions of humans

01:21:43 ◼   ► over the past 50 years."

01:21:44 ◼   ► It's like, "I have a problem."

01:21:46 ◼   ► (laughing)

01:21:47 ◼   ► - Yeah, it's the, the racist AI says,

01:21:49 ◼   ► "I learned it from you, Dan."

01:21:50 ◼   ► - "I learned it from humans."

01:21:52 ◼   ► 'Cause where else are they gonna learn everything?

01:21:54 ◼   ► Everything they know is from us

01:21:56 ◼   ► and we're filled with bias.

01:21:58 ◼   ► - Yep.

01:21:58 ◼   ► - Yeah.

01:22:00 ◼   ► - Yep, yep.

01:22:02 ◼   ► Senator, I would like to speak for a moment

01:22:06 ◼   ► if the chair will recognize me

01:22:07 ◼   ► about these new woke AIs that are turning up.

01:22:11 ◼   ► (laughing)

01:22:12 ◼   ► It turns out you've been eliminating racist screeds

01:22:15 ◼   ► from the training corpus.

01:22:17 ◼   ► Look at that bias you're incorporating.

01:22:19 ◼   ► - I would like a more broadly broad corpus for my AIs

01:22:24 ◼   ► that is not limited and suppressed.

01:22:27 ◼   ► I don't know, there's a new character I'm working on.

01:22:29 ◼   ► It's the Senator, windbag Senator,

01:22:32 ◼   ► or a fog horn, like horn, I say.

01:22:35 ◼   ► I wanted to do, since it's you and me,

01:22:39 ◼   ► I wanted to do a little Mac OS talk.

01:22:41 ◼   ► I do this with my pals who are like the big Mac users.

01:22:47 ◼   ► Big Mac users. - I got a big Mac,

01:22:48 ◼   ► that's for sure.

01:22:49 ◼   ► - Old, usually old school.

01:22:52 ◼   ► - Yes, we are old.

01:22:54 ◼   ► - We are, it's true.

01:22:55 ◼   ► And also I will say your buddies on ATP

01:23:03 ◼   ► are not of the old school.

01:23:06 ◼   ► - No, in multiple ways. - Of being Mac users.

01:23:08 ◼   ► both in the life way and also in the Apple way.

01:23:10 ◼   ► - Yeah, yeah, exactly.

01:23:12 ◼   ► I find it very funny 'cause you're the voice of the person

01:23:15 ◼   ► who actually knows what the Mac was before, you know,

01:23:17 ◼   ► 2004 on that podcast.

01:23:20 ◼   ► But you know, I don't wanna get like super retro here.

01:23:23 ◼   ► Instead, I just wanted to sort of do a little check-in

01:23:25 ◼   ► with you about how you're feeling about macOS today.

01:23:27 ◼   ► You know, it used to be we could read 20,000 words

01:23:29 ◼   ► about how you're feeling about macOS,

01:23:31 ◼   ► but those days have passed.

01:23:32 ◼   ► We can listen to some podcasts and hear about it,

01:23:33 ◼   ► but like, I just wanted to do a little Ventura check-in.

01:23:37 ◼   ► You know, we live in the system settings era now.

01:23:40 ◼   ► I was curious about like your day-to-day Mac usage,

01:23:45 ◼   ► especially, right?

01:23:45 ◼   ► Because obviously you have, you're doing development

01:23:49 ◼   ► and you're doing your podcasts.

01:23:50 ◼   ► Like what is, when you use your Mac every day, all day,

01:23:55 ◼   ► what, you know, what's working okay?

01:23:58 ◼   ► And what is bugging you

01:24:00 ◼   ► about the modern 2022 era Mac experience?

01:24:05 ◼   ► Mean I have to unventura. I think this has been a pretty good OS update in terms of not breaking tons of stuff

01:24:12 ◼   ► I agree, you know and that's that's good

01:24:15 ◼   ► system settings as we've all talked about so much that's bad, but I think like the the the Mac OS software trends

01:24:23 ◼   ► Have just continued sort of linearly through the ventura transition

01:24:28 ◼   ► Those trends are that it seems like

01:24:31 ◼   ► Apple doesn't know how to make good Mac apps anymore that either the people who know how to do it aren't being allowed to make

01:24:36 ◼   ► Mac apps or they don't exist there anymore

01:24:38 ◼   ► It seems like there aren't a lot

01:24:40 ◼   ► It's not a lot of innovation happening on the Mac basically the the quote unquote best and most valuable apps

01:24:45 ◼   ► We get on the Mac or simply

01:24:47 ◼   ► Cross-platform apps built with cross-platform technologies and we're happy to have them

01:24:51 ◼   ► So it's it's great that we have slack and discord on the Mac in you know

01:24:55 ◼   ► These individual apps that are pretty well done, but they're not I mean, they're not Mac native apps, right?

01:25:00 ◼   ► So that whole trend of like, you know, all those of us who could remember a time when software innovation was happening on the Mac app

01:25:06 ◼   ► Platform it's so clear that it's not happening anymore

01:25:08 ◼   ► The trend I talked about earlier about Apple locking down the system and making it harder and harder to do interesting things

01:25:13 ◼   ► I experienced that personally even with my stupid dinky app

01:25:16 ◼   ► But I experienced that as a user because there are fewer and fewer Mac apps that do really interesting things

01:25:22 ◼   ► Just because if you want to try to do them

01:25:25 ◼   ► You can't be on the Mac App Store and even if you don't want to be in the Mac App Store

01:25:28 ◼   ► or you probably can't do them anyway

01:25:29 ◼   ► because the system is so locked down

01:25:30 ◼   ► and Apple seems uninterested in providing hooks.

01:25:34 ◼   ► I've been begging Apple,

01:25:35 ◼   ► when I used to write Mac OS X reviews,

01:25:36 ◼   ► I was begging them for years.

01:25:38 ◼   ► You know what the doc does?

01:25:40 ◼   ► Everything useful that the doc does

01:25:42 ◼   ► provide a public API to do that.

01:25:44 ◼   ► For example, right now my Slack icon

01:25:46 ◼   ► has a little badge with a one on it.

01:25:48 ◼   ► There's an, the doc, you know,

01:25:50 ◼   ► there's an API for Mac apps to say,

01:25:51 ◼   ► "Hey, badge my icon and put a one in it," or whatever, right?

01:25:56 ◼   ► that sort of bounce my dock icon.

01:25:59 ◼   ► Like Mac applications have APIs they can call

01:26:02 ◼   ► to do all those things,

01:26:03 ◼   ► but the only thing on the entire system

01:26:05 ◼   ► that can hear the cries of those apps is the Apple dock.

01:26:08 ◼   ► You cannot make a third party application

01:26:11 ◼   ► that knows when an application wants to bad itself,

01:26:14 ◼   ► that knows when an application wants to change its icon

01:26:16 ◼   ► to show the current date 'cause it's a calendar app,

01:26:18 ◼   ► that knows when it like,

01:26:19 ◼   ► you can only do what Apple's provided hooks for.

01:26:21 ◼   ► And so many things are only the purview

01:26:24 ◼   ► of the operating system.

01:26:26 ◼   ► Window management stuff, can you write window shade

01:26:28 ◼   ► for Mac OS?

01:26:29 ◼   ► Not without hacking the hell out of it

01:26:31 ◼   ► and it'd probably be really difficult now.

01:26:32 ◼   ► There are no hooks for that type of window management thing.

01:26:34 ◼   ► You can't invent that yourself,

01:26:36 ◼   ► only Apple gets access to that and on and on.

01:26:38 ◼   ► The old way of doing this was terrible,

01:26:40 ◼   ► but the new way is provide clean hooks.

01:26:42 ◼   ► Even when they did provide them,

01:26:43 ◼   ► Dropbox used to be hacking the finder

01:26:45 ◼   ► to put the badges on its icon.

01:26:46 ◼   ► Apple said, "Stop hacking the finder,

01:26:48 ◼   ► "we'll give you public APIs for this."

01:26:50 ◼   ► Great, that's the example of exactly what I wanted to do.

01:26:52 ◼   ► But by all accounts, the API they provided

01:26:55 ◼   ► is not as good and reliable as the one that Dropbox did

01:26:57 ◼   ► by hacking the Finder.

01:26:58 ◼   ► So, you know, E for effort, Apple, but do better, right?

01:27:03 ◼   ► And that has been another trend that has been ongoing.

01:27:06 ◼   ► Apple, uninterested in providing system level integration

01:27:09 ◼   ► that makes the Mac, you know,

01:27:11 ◼   ► better than iOS and the iPad.

01:27:13 ◼   ► And I know it's like, it's rich us complaining

01:27:15 ◼   ► or iPad people just wish they could like redirect sound

01:27:18 ◼   ► for one application to another.

01:27:19 ◼   ► And even that, you know,

01:27:20 ◼   ► Rogomedia is still fighting the good fight to get that.

01:27:23 ◼   ► It takes two reboots to install it, but yes, you can get it installed if you need to.

01:27:27 ◼   ► Right, and so like those trends, these are not new things, these are just trends that have been going for years and years continue on the Mac,

01:27:33 ◼   ► and it's making the Mac seem increasingly less rich, because for those of us who remember when innovative interesting apps were happening,

01:27:42 ◼   ► or for those of us who can remember when system augmentation allowed lots of user interface experimentation,

01:27:49 ◼   ► even when it was done in a terrible way, that was kind of cool.

01:27:53 ◼   ► Now it's not done in a terrible way, it's just not done at all, and every year that goes by they turn the screws on that one.

01:27:58 ◼   ► I do think sometimes about the fact that the stuff that I view as like, "Oh, this is great that this is on the Mac, I'm so glad that it does this,"

01:28:08 ◼   ► it's almost always something that was put in in an earlier era, and that if I ask myself, "Would Apple do this now?"

01:28:19 ◼   ► the answer is no, right?

01:28:21 ◼   ► Like the terminal, well, no, they wouldn't put that in there.

01:28:26 ◼   ► - They would, I think today's Apple would put the terminal in.

01:28:30 ◼   ► - I mean, they might, I mean,

01:28:31 ◼   ► Microsoft has really pushed on that,

01:28:33 ◼   ► but like Apple is gonna, you gotta wonder,

01:28:35 ◼   ► 'cause their other platforms don't have it,

01:28:37 ◼   ► and I know why they don't.

01:28:38 ◼   ► I was thinking about the menu bar.

01:28:40 ◼   ► I was thinking about like API to put stuff in the menu bar.

01:28:43 ◼   ► Like would Apple today be like the menu bar on the Mac,

01:28:47 ◼   ► It's a free-for-all put your apps up there put icons go go nuts speaking about public

01:28:53 ◼   ► Do you remember the old war about that right so Mac OS 10 came out it basically had either had an existing API for that

01:28:59 ◼   ► Or they had someone added it in before someone told them no right, but there were two API's

01:29:03 ◼   ► There was the good one that Apple got to use and there was the crappy one that the third

01:29:06 ◼   ► Use and this was back before

01:29:09 ◼   ► Sandboxing in the Mac App Store so what happened is third parties just used the naughty one

01:29:13 ◼   ► Because they're like, why would I use the quote unquote

01:29:15 ◼   ► public one that's worse?

01:29:16 ◼   ► And one of the things that was worse about it,

01:29:18 ◼   ► if you use the public one, you couldn't move the icons around.

01:29:21 ◼   ► So no third party developer is going

01:29:22 ◼   ► to pick that one right next to it is the API that Apple uses.

01:29:24 ◼   ► It lets you hold down Command key and move them around.

01:29:27 ◼   ► So everybody used the fancier one.

01:29:29 ◼   ► And it took years of convincing Apple, like, look,

01:29:31 ◼   ► you have this API.

01:29:33 ◼   ► Don't have the worst public one that no one wants to use

01:29:35 ◼   ► and the one that you yell at everyone about using.

01:29:37 ◼   ► Just make one API that is decent for doing this.

01:29:41 ◼   ► And you're right that if someone hadn't snuck