ATP

00:00:00 ◼   ► Hey, have we ever recorded this show before noon?

00:00:04 ◼   ► I would be surprised if Jon has ever recorded a podcast before noon.

00:00:08 ◼   ► No, I've done a bunch of morning ones for usually like people in the UK and stuff.

00:00:12 ◼   ► Have you even shaved yet?

00:00:13 ◼   ► What are you talking about? I don't think you understand my shaving schedule.

00:00:18 ◼   ► Yeah, maybe. You've completed the morning shave, but not the evening shave.

00:00:22 ◼   ► Right, right.

00:00:23 ◼   ► No, that's not... I shave like once a week-ish.

00:00:27 ◼   ► I thought I figured you know your thick Italian heritage is just like just constant.

00:00:32 ◼   ► It is, oh it is.

00:00:34 ◼   ► It's just I don't feel the need to shave it off now that I'm not going into an office

00:00:38 ◼   ► all the time.

00:00:39 ◼   ► Even when I was going to an office honestly it was pretty hit or miss in the later years.

00:00:44 ◼   ► You slowly become like the elder programmer look.

00:00:48 ◼   ► Right?

00:00:49 ◼   ► Yeah, because one or two days it's like stubble and then people start asking you whether you're

00:00:54 ◼   ► throwing a beard, and then you shave,

00:00:55 ◼   ► and then just repeats.

00:00:56 ◼   ► (laughing)

00:00:57 ◼   ► - By the way, if anybody asks,

00:01:00 ◼   ► today is not December 26th, it's December 24th.

00:01:03 ◼   ► And I figure if Roman emperors were able to dictate

00:01:08 ◼   ► whatever they felt like the calendar doing,

00:01:10 ◼   ► I feel like I can do it too.

00:01:12 ◼   ► So today is really actually December 24th in our family,

00:01:15 ◼   ► and we have our own custom calendar, only for this year.

00:01:18 ◼   ► Because this has been,

00:01:22 ◼   ► oh, this has been an experience.

00:01:25 ◼   ► So we drove halfway upstate on the day before Christmas Eve,

00:01:32 ◼   ► on Friday, this past Friday.

00:01:34 ◼   ► The idea was to drive the rest of the way upstate

00:01:36 ◼   ► to our family's place on Saturday, Christmas Eve day.

00:01:41 ◼   ► And we woke up on Saturday morning,

00:01:43 ◼   ► and our kid had a massive fever.

00:01:46 ◼   ► And we were like, oh, no.

00:01:48 ◼   ► So here we are scheduling to go up to visit

00:01:53 ◼   ► the whole family, including his grandparents,

00:01:56 ◼   ► who were, you know, we wanna be careful

00:01:58 ◼   ► with what we bring to them.

00:01:59 ◼   ► And so we were like, oh no, we gotta,

00:02:02 ◼   ► we can't go up like this.

00:02:03 ◼   ► We can't bring this very fevery kid.

00:02:07 ◼   ► First of all, we can't make him sit in a car

00:02:09 ◼   ► for three hours.

00:02:11 ◼   ► That's not super great when you're feeling sick.

00:02:13 ◼   ► Second of all, we don't wanna get them sick.

00:02:15 ◼   ► So we're like, oh God, what do we do?

00:02:16 ◼   ► It's Christmas Eve.

00:02:18 ◼   ► And so we worked out this plan where we were basically

00:02:21 ◼   ► gonna wait until the fever broke and talk to a doctor

00:02:24 ◼   ► and see when it's safe and everything.

00:02:26 ◼   ► And then so as we're working this out,

00:02:27 ◼   ► we get reports from all of his,

00:02:30 ◼   ► from some of his school classmates' parents.

00:02:33 ◼   ► His whole class basically has flu A.

00:02:36 ◼   ► It's the whole class.

00:02:37 ◼   ► - Cool. - It's the real flu.

00:02:39 ◼   ► Many of them got tested.

00:02:40 ◼   ► It's influenza A, real flu,

00:02:43 ◼   ► and now he's showing flu symptoms and we're like,

00:02:44 ◼   ► oh crap.

00:02:46 ◼   ► And it's Christmas Eve, may I remind you.

00:02:48 ◼   ► It's not like anything's easily accessible or open.

00:02:51 ◼   ► So we do like a telehealth thing,

00:02:53 ◼   ► which thank God that even exists.

00:02:55 ◼   ► I mean, that's an amazing thing.

00:02:57 ◼   ► I'm so happy that that's,

00:02:59 ◼   ► like I think one of the,

00:03:01 ◼   ► if there is such a thing as a silver lining

00:03:03 ◼   ► for the COVID era,

00:03:04 ◼   ► it's that I think we have way more access

00:03:07 ◼   ► to virtual services than we used to.

00:03:08 ◼   ► And even though telehealth existed before COVID,

00:03:11 ◼   ► it was a much smaller and less commonly available thing

00:03:13 ◼   ► for most insurance plans and stuff.

00:03:15 ◼   ► So now it's super easy.

00:03:17 ◼   ► So we did telehealth,

00:03:18 ◼   ► and then we had to get a prescription.

00:03:21 ◼   ► And by this time,

00:03:22 ◼   ► the time that we had done the telehealth and done all this,

00:03:25 ◼   ► it was 6 p.m. on Christmas Eve.

00:03:28 ◼   ► - Ugh.

00:03:28 ◼   ► - So try getting a prescription filled

00:03:30 ◼   ► at 6 p.m. on Christmas Eve.

00:03:31 ◼   ► (laughing)

00:03:33 ◼   ► So we eventually found like a 24-hour pharmacy

00:03:36 ◼   ► that was open Christmas Eve, had it sent there.

00:03:39 ◼   ► It was a whole thing, but we got it.

00:03:42 ◼   ► Anyway, so we've had it,

00:03:44 ◼   ► And we're like, all right, the good thing about Christmas

00:03:47 ◼   ► is that unlike, for example, Halloween,

00:03:50 ◼   ► if you are sick on Halloween, you just miss everything

00:03:53 ◼   ► because Halloween is a holiday that depends

00:03:55 ◼   ► on the rest of the surrounding community

00:03:57 ◼   ► celebrating it at the same time you are celebrating it.

00:04:00 ◼   ► Otherwise, you can't really go trick or treating.

00:04:02 ◼   ► If you just show up on strangers' doors

00:04:03 ◼   ► and knock on them on November 3rd,

00:04:05 ◼   ► you're gonna have a bad time, I think.

00:04:08 ◼   ► But Christmas, it really just depends

00:04:11 ◼   ► on when your family wants to celebrate it.

00:04:14 ◼   ► It's basically within your own household for the most part.

00:04:16 ◼   ► And so you can kind of celebrate it whenever you want.

00:04:19 ◼   ► So we just delayed it.

00:04:20 ◼   ► So yesterday we sat inside and played video games all day.

00:04:23 ◼   ► Big fan of Project High Rise right now.

00:04:25 ◼   ► It's like a modern Sim Tower.

00:04:27 ◼   ► It's fantastic.

00:04:28 ◼   ► - Oh, I loved Sim Tower.

00:04:29 ◼   ► - Yeah, yeah.

00:04:30 ◼   ► But yeah, and so we just played video games all day.

00:04:33 ◼   ► They tried to teach me how to play Don't Starve together

00:04:35 ◼   ► and I starved.

00:04:36 ◼   ► Well, no, I died almost instantly.

00:04:40 ◼   ► I think I lasted maybe 90 seconds in the game

00:04:43 ◼   ► before dying, so that might not be a strength of mine, but certainly SimTower-like games

00:04:49 ◼   ► are, I'm enjoying. So we just did all that yesterday, and this afternoon we are driving

00:04:53 ◼   ► upstate and today is fake Christmas Eve and tomorrow is fake Christmas and we're gonna

00:04:58 ◼   ► celebrate it, dammit. Anyway, so I'm very thankful for Modern Medicine and for our flexible

00:05:04 ◼   ► family in their scheduling. And yeah, it's been a time. Oh, and 24 Hour Pharmacy is also

00:05:11 ◼   ► Thankful for that.

00:05:12 ◼   ► - Did you give Adam COVID tests?

00:05:15 ◼   ► - Yes, and no.

00:05:17 ◼   ► - And had he gotten his flu shot yet?

00:05:21 ◼   ► - No, this is the thing.

00:05:22 ◼   ► So he had COVID on Halloween.

00:05:25 ◼   ► So that's why I'm such an expert

00:05:27 ◼   ► in what it's like to miss Halloween.

00:05:29 ◼   ► That was awful.

00:05:30 ◼   ► And because we thought we might be exposed to it,

00:05:33 ◼   ► we didn't get any boosters or anything yet,

00:05:35 ◼   ► 'cause we're like, all right,

00:05:36 ◼   ► now let's wait two months, like they say,

00:05:37 ◼   ► and then we'll get all of our boosters.

00:05:39 ◼   ► and that puts us about now.

00:05:42 ◼   ► - Whoops.

00:05:43 ◼   ► - Yeah, I always remember when they would start advertising

00:05:46 ◼   ► the flu shots and they'd be advertising them

00:05:48 ◼   ► and before Halloween I'd be like,

00:05:50 ◼   ► "That seems a little early for flu shots,"

00:05:52 ◼   ► but I eventually gave in and said,

00:05:53 ◼   ► "All right, well I'll just get it as soon as it's available

00:05:55 ◼   ► so I'll let this be a lesson to the listener."

00:05:58 ◼   ► Even though sometimes it seems like they're giving out

00:05:59 ◼   ► the flu shots way too early

00:06:00 ◼   ► and it's not even close to flu season yet,

00:06:02 ◼   ► it's a good idea to get them when you can

00:06:03 ◼   ► 'cause it's easy to forget.

00:06:05 ◼   ► - Yeah, yeah, last year we got them all regularly on schedule

00:06:08 ◼   ► and we had a totally fine, smooth winter.

00:06:10 ◼   ► I mean, you know, and I know it's on 100% or whatever,

00:06:13 ◼   ► but at this point, I'll take any improvement in chances,

00:06:17 ◼   ► even if it's on 100%, I'll take it.

00:06:20 ◼   ► - All right, let's do some follow-up,

00:06:23 ◼   ► and it starts with Brad Crittenden, who writes,

00:06:24 ◼   ► "I got a butterfly keyboard class action lawsuit

00:06:26 ◼   ► "settlement notice.

00:06:28 ◼   ► "It claims that they will pay 300 to $395.

00:06:31 ◼   ► "I have no recollection of this flying by.

00:06:33 ◼   ► "I probably filled it out and just don't remember,

00:06:35 ◼   ► "but I have zero recollection of this."

00:06:37 ◼   ► So here's the thing, I'd never filled it out

00:06:39 ◼   ► and I got one, or rather I got the notice

00:06:41 ◼   ► saying I'm entitled to one or whatever,

00:06:42 ◼   ► but I'm not gonna submit it because,

00:06:45 ◼   ► see I never, as far as I can recall,

00:06:48 ◼   ► I think I only got it repaired once,

00:06:51 ◼   ► or even, did I ever get it repaired or did I just not?

00:06:54 ◼   ► See I can't even remember

00:06:55 ◼   ► if I actually ever got one repaired.

00:06:57 ◼   ► So that's why I'm like, you know,

00:06:58 ◼   ► it's 'cause the class action is basically like

00:07:01 ◼   ► if you got a butterfly keyboard laptop repaired by Apple,

00:07:05 ◼   ► depending on how many times you got it repaired,

00:07:07 ◼   ► you have a certain payout of 100 bucks

00:07:11 ◼   ► or 200 or 300 bucks apparently,

00:07:12 ◼   ► but it's not a huge amount of money.

00:07:15 ◼   ► 'Cause I mean, look, the real problem

00:07:17 ◼   ► with the butterfly keyboards,

00:07:18 ◼   ► well, there were two real problems.

00:07:19 ◼   ► Number one, people getting them repaired out of warranty

00:07:21 ◼   ► had to pay quite a lot of money, and more than that.

00:07:24 ◼   ► Remember, the price was, I believe,

00:07:25 ◼   ► like over $500 to get one repair out of warranty.

00:07:29 ◼   ► So there's that issue, which this, I guess,

00:07:32 ◼   ► partially addresses but doesn't fully address.

00:07:35 ◼   ► The other issue is any replacement you got

00:07:38 ◼   ► would have the same problem eventually.

00:07:40 ◼   ► And so really the whole laptop

00:07:43 ◼   ► was incredibly devalued over time.

00:07:46 ◼   ► That was the big problem.

00:07:48 ◼   ► And so, you know, this is, I'm glad Apple, you know,

00:07:52 ◼   ► got sued over the butterfly keyboards

00:07:53 ◼   ► 'cause frankly they should have.

00:07:55 ◼   ► One issuance of that keyboard was bad enough

00:08:00 ◼   ► and then to keep issuing it year after year

00:08:02 ◼   ► after year after year claiming they'd fixed it

00:08:04 ◼   ► was beyond negligent and borderline fraudulent.

00:08:08 ◼   ► So they should have been held liable for that

00:08:11 ◼   ► and it seems like they've been held

00:08:12 ◼   ► a little bit liable for that.

00:08:14 ◼   ► But ultimately, if you were an owner

00:08:16 ◼   ► of a butterfly keyboard laptop,

00:08:18 ◼   ► this kind of helps a little bit,

00:08:20 ◼   ► but it doesn't really address the actual value

00:08:23 ◼   ► that you lost by having that laptop

00:08:25 ◼   ► and having to eventually sell or trade it in

00:08:27 ◼   ► or get it repaired out of warranty.

00:08:29 ◼   ► I was excited to get my money for this

00:08:31 ◼   ► and then I realized, oh wait,

00:08:32 ◼   ► my butterfly keyboard that broke wasn't mine,

00:08:34 ◼   ► it was on my work's laptop.

00:08:35 ◼   ► It was annoying though.

00:08:36 ◼   ► - Oh no.

00:08:37 ◼   ► - You gotta get your work laptop taken away

00:08:38 ◼   ► and you gotta have a replacement

00:08:40 ◼   ► and all your stuff is on it.

00:08:40 ◼   ► It's even more annoying with all the work

00:08:42 ◼   ► 'cause they own the machine

00:08:43 ◼   ► so it's not like you can just copy stuff off of it

00:08:45 ◼   ► 'cause they don't let you attach external drives

00:08:47 ◼   ► and blah, blah, blah.

00:08:47 ◼   ► So anyway, I hope my old workplace gets the money

00:08:51 ◼   ► they have coming to them

00:08:52 ◼   ► for all the butterfly keyboards that broke.

00:08:55 ◼   ► - Yeah, I didn't end up getting any of these repaired.

00:08:57 ◼   ► In fact, I'm trying to remember.

00:08:59 ◼   ► So we still have, you know, Aaron, what was my adorable now, Aaron's adorable in the house,

00:09:03 ◼   ► and every great moment, like once every few months, she'll be like, "Oh, my such and such

00:09:07 ◼   ► key is sticking."

00:09:08 ◼   ► Yep, sure is.

00:09:09 ◼   ► I bet you're right.

00:09:11 ◼   ► But anyways, but that never got a repair.

00:09:14 ◼   ► And then I think the one or two other MacBook Pros that I had during the butterfly era were

00:09:19 ◼   ► either works or have already been sold off.

00:09:22 ◼   ► So yeah, I'd never had a repair, and that probably explains why I never got this notice.

00:09:26 ◼   ► But I'm glad to hear that this is a thing,

00:09:28 ◼   ► and I agree with you Marco that,

00:09:30 ◼   ► you know, it's one thing to release a keyboard

00:09:32 ◼   ► that's flawed and maybe you didn't know it,

00:09:34 ◼   ► maybe you thought, oh it'll work itself out,

00:09:36 ◼   ► and okay fine.

00:09:38 ◼   ► The thing that bothered me about this whole fiasco,

00:09:40 ◼   ► well one of the things that bothered me

00:09:41 ◼   ► about this whole fiasco was the fact that,

00:09:43 ◼   ► like you said, they just kept saying,

00:09:44 ◼   ► no, no, no, it's fine, no, it's fine,

00:09:47 ◼   ► no it's fine, no, it's fine.

00:09:49 ◼   ► - This time we fixed it.

00:09:50 ◼   ► - Oh yeah, oh we added a gasket or whatever it was,

00:09:52 ◼   ► I don't remember the details,

00:09:53 ◼   ► we added a gasket, that'll fix it.

00:09:55 ◼   ► Because you're right, they were also saying,

00:09:57 ◼   ► "Nothing's wrong, keep changing it.

00:09:58 ◼   ► "Nothing's wrong, oh wait, we changed it again.

00:10:00 ◼   ► "Nothing's wrong, what are you talking about?

00:10:02 ◼   ► "There's no flaw, we fixed the flaw.

00:10:03 ◼   ► "What are you talking about?

00:10:04 ◼   ► "There's no flaw, oh, we fixed it again."

00:10:05 ◼   ► Like, yeah, it was a rough time.

00:10:08 ◼   ► - Huge Apple energy.

00:10:09 ◼   ► - Not to mention the fact that it sucked

00:10:10 ◼   ► even when it was working, but we'll move beyond that for now.

00:10:14 ◼   ► - I mean, what are you gonna do?

00:10:14 ◼   ► But no, I'm glad to see that this is finding,

00:10:19 ◼   ► as late as it is, I'm glad to see it's finding

00:10:20 ◼   ► some sort of resolution for regular people.

00:10:23 ◼   ► - Two related things to this.

00:10:24 ◼   ► I don't know we don't couldn't find this feedback for the show

00:10:26 ◼   ► But we did get some feedback from somebody through some channel live streams that said

00:10:30 ◼   ► That they had a Mac with a butterfly keyboard

00:10:33 ◼   ► And it went bad and they brought it to the Apple store

00:10:36 ◼   ► And they were informed that oh the Apple repair program for the butterfly keyboards that said hey if you have a problem

00:10:42 ◼   ► You know whatever it's called the extended repair program if you have a problem or replace it that program has ended

00:10:46 ◼   ► So this person had a butterfly keyboard

00:10:48 ◼   ► Brought it in and and they're no longer

00:10:51 ◼   ► eligible to get it repaired for free or for cheap because the Apple, we talked about this

00:10:54 ◼   ► when Apple rolled out the program, it's like, okay, at a certain point, Apple says, yeah,

00:10:58 ◼   ► we have this repair program because we realized we put out a defective keyboard, but at a

00:11:01 ◼   ► certain point, we just want to stop paying for that, so we're going to.

00:11:03 ◼   ► And I think we're past that point now, which is pretty crappy, so if you do want it repaired,

00:11:07 ◼   ► you're on the hook for whatever the, you know, $500 repair bill, really just probably want

00:11:11 ◼   ► to throw that laptop into the sea at that point.

00:11:13 ◼   ► And the second thing is that, just from some brief Googling, I believe I still have a butterfly

00:11:19 ◼   ► keyboard Mac that's in semi-active use. The original Retina MacBook Air had it, right?

00:11:25 ◼   ► The Intel one? Yeah, the first one, the 2018 model. But the

00:11:29 ◼   ► 2019 did not. Yeah, maybe I got the 2019 one, I'll have

00:11:33 ◼   ► to look it up. There was an Intel MacBook Air Retina that

00:11:37 ◼   ► didn't have the butterfly keyboard. Yeah, I'll double check, but I was thinking,

00:11:42 ◼   ► "Oh, maybe if that one ever goes bad, I'll get paid," but apparently I won't, again,

00:11:45 ◼   ► according to that feedback that I can't find. So that's pretty crappy. I understand the

00:11:49 ◼   ► the repair programs can't last forever, but it's kind of like, what's the point of the

00:11:53 ◼   ► repair program if you're saying, "We're sorry for putting out a defective laptop, but we

00:11:56 ◼   ► have a timer on the lifetime of your computer."

00:11:59 ◼   ► So, I mean, it seems like the keyboard is even more likely to fail after the three or

00:12:04 ◼   ► four or five or however many years after that time period is up, but they're just saying,

00:12:07 ◼   ► "Yeah, we can't pay for that forever," because we basically know that there's probably some

00:12:11 ◼   ► graph of over time, what are the odds that a keyboard will fail, and at a certain point,

00:12:15 ◼   ► it becomes like ruinous, not ruinous,

00:12:16 ◼   ► but very expensive for Apple to repair

00:12:19 ◼   ► because the percentage of them that are gonna go bad

00:12:21 ◼   ► after 10 years is approaching like 25% or something.

00:12:25 ◼   ► - Well, but it does make sense to have a time limit on it.

00:12:28 ◼   ► I just think this time limit was probably too short.

00:12:31 ◼   ► 'Cause at some point they have to stop manufacturing

00:12:33 ◼   ► and keeping the parts and the service equipment

00:12:36 ◼   ► and stuff like that.

00:12:36 ◼   ► So it does make sense to have an expiration date,

00:12:39 ◼   ► but that expiration date should be however many years

00:12:43 ◼   ► their products are expected to be in use

00:12:45 ◼   ► that number of years after they stop being for sale.

00:12:49 ◼   ► And so maybe it's like seven years, you know,

00:12:51 ◼   ► whatever it is, like as long as they can expect

00:12:53 ◼   ► a laptop to reasonably still be in use by most people,

00:12:57 ◼   ► it should have covered that duration, and it didn't.

00:13:00 ◼   ► - They don't have to keep making

00:13:01 ◼   ► the butterfly keyboard parts,

00:13:02 ◼   ► like that's not the only solution available to them.

00:13:05 ◼   ► They can do what they have done in the past,

00:13:06 ◼   ► which is that you bring something in to get repaired,

00:13:08 ◼   ► and they're like, oh, we don't make parts for that anymore,

00:13:11 ◼   ► and we don't make that product anymore,

00:13:13 ◼   ► but we recognize that you've got a broken thing

00:13:15 ◼   ► and it's our fault, so here is the new version.

00:13:17 ◼   ► We've heard lots and lots of stories of Apple doing that.

00:13:19 ◼   ► Like they basically give you a new laptop

00:13:20 ◼   ► or a new iPod back in the day.

00:13:22 ◼   ► That's not the one that you had

00:13:23 ◼   ► that is better than the one you had,

00:13:24 ◼   ► but it's the only option available to them

00:13:26 ◼   ► because they don't make the old one anymore.

00:13:27 ◼   ► They could always do that,

00:13:28 ◼   ► but of course that would cost money.

00:13:30 ◼   ► - And that would cost way more money.

00:13:32 ◼   ► 'Cause like-- - Yeah.

00:13:33 ◼   ► Well, you stopped shipping a defective keyboard for years.

00:13:35 ◼   ► - I know, but like, but you know,

00:13:36 ◼   ► this was not a small number of affected products,

00:13:39 ◼   ► even though they kept saying,

00:13:40 ◼   ► "It's only a tiny percentage,

00:13:42 ◼   ► but we all know it's a much bigger percentage.

00:13:44 ◼   ► - A tiny percentage after six months,

00:13:46 ◼   ► is it a tiny percentage after six years?

00:13:47 ◼   ► - Right, yeah, that's the thing.

00:13:49 ◼   ► This would have been a massive number

00:13:52 ◼   ► if they actually did that.

00:13:52 ◼   ► And I've heard of that happening,

00:13:55 ◼   ► and one time it even happened for me

00:13:57 ◼   ► back in the white plastic MacBook days.

00:14:00 ◼   ► But it's not super common.

00:14:03 ◼   ► I feel like you have to have a fairly exceptional case,

00:14:05 ◼   ► and I think it has to go through some levels of approval

00:14:08 ◼   ► before they just give you a new thing

00:14:11 ◼   ► replace your old broken thing. So it's not that common of a thing.

00:14:16 ◼   ► We are brought to you this week by Nebula, a streaming service created and owned by some

00:14:21 ◼   ► of the Internet's most thoughtful creators, offering full length and ad-free videos earlier

00:14:26 ◼   ► than anywhere else as well as Nebula exclusives. Personally, I love Nebula because here's what

00:14:32 ◼   ► happened. I'm a YouTube video addict like many of us and I started noticing that many

00:14:39 ◼   ► of my favorite creators who would have all this cool, like, usually it's like educational

00:14:43 ◼   ► explainer kind of content. People like Wendover Productions and Real Engineering and Practical

00:14:48 ◼   ► Engineering and Georgia Dow. I was seeing all these people who, videos I was watching,

00:14:54 ◼   ► and they would keep saying, "Hey, by the way, we started the streaming service. We're over

00:14:57 ◼   ► here on Nebula and all of our stuff there is ad-free and, you know, you get some bonus

00:15:01 ◼   ► content or some exclusive features." And I was like, "Okay, I gotta check this out."

00:15:06 ◼   ► And it is so great.

00:15:08 ◼   ► They have so many creators.

00:15:09 ◼   ► The ones I already knew about, plus so many more that I've discovered that I also like

00:15:13 ◼   ► because it's kind of in a similar vein or similar quality levels of what I was hoping

00:15:18 ◼   ► for.

00:15:19 ◼   ► And everything there is ad-free and it's full length.

00:15:22 ◼   ► It's like getting the director's cut or getting bonus features.

00:15:26 ◼   ► Plus they have these exclusives that are even bigger and higher production value or more

00:15:32 ◼   ► in depth than you could often go in a place like YouTube.

00:15:34 ◼   ► And so I am super happy with Nebula.

00:15:37 ◼   ► They have fully produced originals,

00:15:39 ◼   ► things like Anita Sarkeesian's That Time When or Impact,

00:15:42 ◼   ► Renee Ritchie's retrospective on the launch of the iPhone.

00:15:45 ◼   ► They have Nebula classes featuring all kinds

00:15:46 ◼   ► of online courses led by creators,

00:15:48 ◼   ► including introduction to programming called What is Code

00:15:51 ◼   ► from creator and NYU professor Daniel Shiffman,

00:15:53 ◼   ► and just so much more bonus and exclusive content.

00:15:56 ◼   ► So see for yourself at nebula.tv/atp.

00:16:00 ◼   ► Now normally this is five bucks a month or 50 bucks a year,

00:16:03 ◼   ► but they're having a sale for us, 40 bucks for a year.

00:16:05 ◼   ► So 40 bucks for a year by going to nebula.tv/atp.

00:16:10 ◼   ► Thank you so much to Nebula for sponsoring our show.

00:16:14 ◼   ► - Let's move on and talk about stable diffusion, baby.

00:16:20 ◼   ► And apparently artists can now opt out

00:16:23 ◼   ► of the next version of stable diffusion.

00:16:26 ◼   ► This is an article on the Technology Review.

00:16:28 ◼   ► Artists will have the chance to opt out

00:16:29 ◼   ► of the next version of stable diffusion.

00:16:30 ◼   ► The company behind it has announced stability.ai

00:16:32 ◼   ► will work with Spawning, an organization founded

00:16:35 ◼   ► by artists couple Matt Dreiherst and Holly Herndon,

00:16:37 ◼   ► who have built a website called Have I Been Trained.

00:16:40 ◼   ► That's in the spirit of Have I Been Pwned, isn't it?

00:16:42 ◼   ► That's so good.

00:16:42 ◼   ► - Oh, all right. - That allows artists

00:16:43 ◼   ► to search for their works in the dataset

00:16:47 ◼   ► that was used to train stable diffusion.

00:16:49 ◼   ► Artists will be able to select which works

00:16:51 ◼   ► they want to exclude from the training data.

00:16:53 ◼   ► - I mean, you don't even need to read the summary

00:16:54 ◼   ► or the article, just the headline,

00:16:56 ◼   ► which I've copied and pasted directly into the notes,

00:16:57 ◼   ► just the headline is enough for you to,

00:16:59 ◼   ► it's kind of like the Ian Beteridge law,

00:17:00 ◼   ► if you're a user, just say no.

00:17:02 ◼   ► Artists can now opt, no, you're not getting it.

00:17:05 ◼   ► The idea is not, oh, we'll steal all your stuff

00:17:07 ◼   ► and build a commercial product off of it and make money,

00:17:10 ◼   ► but you can always track us down,

00:17:12 ◼   ► find out we're doing that and tell us to stop.

00:17:14 ◼   ► Everything's fine now, right?

00:17:16 ◼   ► No, it's not.

00:17:17 ◼   ► We'll siphon money from your bank account slowly over time,

00:17:20 ◼   ► but if you catch us doing that and tell us to stop,

00:17:22 ◼   ► we'll stop.

00:17:24 ◼   ► Everything's fine now, right?

00:17:25 ◼   ► No problem anymore?

00:17:26 ◼   ► No, this is not the solution.

00:17:28 ◼   ► You can't take people's stuff, use it,

00:17:30 ◼   ► and say, "But we'll stop if you ask us to."

00:17:33 ◼   ► It should not be an opt-out.

00:17:34 ◼   ► Would you like to opt out

00:17:35 ◼   ► of having your intellectual property stolen?

00:17:37 ◼   ► Please offer that plan to Disney.

00:17:39 ◼   ► Hey, Disney, we'll be using your intellectual property,

00:17:41 ◼   ► but if you ask us to stop, we will.

00:17:43 ◼   ► - I can't.

00:17:45 ◼   ► - It's hilarious that someone thought this was a good idea.

00:17:48 ◼   ► - Well, I can, I mean, look,

00:17:51 ◼   ► I know this is probably unpopular,

00:17:52 ◼   ► so maybe I shouldn't even bother,

00:17:54 ◼   ► but I think the issue of copyright and rights

00:17:59 ◼   ► in terms of whether AI can be trained

00:18:02 ◼   ► on your copyrighted material is a weird blurry situation.

00:18:06 ◼   ► I don't think it's clear cut.

00:18:08 ◼   ► 'Cause if you think about what training is

00:18:11 ◼   ► and how it might relate to real world analogs

00:18:13 ◼   ► to which we have precedent,

00:18:14 ◼   ► I am able to view a copyrighted work

00:18:19 ◼   ► if it's being shown to me in a legal way.

00:18:21 ◼   ► I'm able to view it and remember it and learn from it.

00:18:25 ◼   ► Artists everywhere are able to look at other artists' work

00:18:29 ◼   ► be inspired by them and generate new things in their spirit.

00:18:33 ◼   ► And so it is kind of a weird blurry line

00:18:36 ◼   ► in the sense that the AI models

00:18:38 ◼   ► that are being trained in this data,

00:18:39 ◼   ► they're not making illegal copies

00:18:41 ◼   ► and distributing illegal copies of the original works.

00:18:45 ◼   ► They are, in a way, viewing them.

00:18:47 ◼   ► - But they're not people, so they're not.

00:18:49 ◼   ► - Right, and so that's,

00:18:50 ◼   ► and I know there was recently a court thing somewhere

00:18:52 ◼   ► that said that they basically failed,

00:18:54 ◼   ► that AI stuff, the output of AI models

00:18:57 ◼   ► does not qualify for copyright

00:18:58 ◼   ► 'cause it was not created by a human.

00:19:00 ◼   ► And I think that's also gonna be worked out over time

00:19:02 ◼   ► in the core systems, I think that's also probably--

00:19:05 ◼   ► - I think that's actually, that ruling's actually

00:19:07 ◼   ► more questionable than, in my opinion,

00:19:09 ◼   ► than the whether they're allowed to take your stuff.

00:19:11 ◼   ► Because the not created by a human thing is fuzzy in the,

00:19:14 ◼   ► okay, so the program is in human,

00:19:17 ◼   ► but a human runs the program and tweaks the output

00:19:19 ◼   ► and so on and so forth, so you have to say,

00:19:22 ◼   ► this isn't copyrightable because it wasn't created

00:19:23 ◼   ► by a human, it's like, well, the program didn't create it

00:19:25 ◼   ► on its own, it's just sitting there waiting for a human

00:19:27 ◼   ► It's like saying, "This is not copyrightable because it was created by Photoshop and Photoshop's

00:19:31 ◼   ► not a human."

00:19:32 ◼   ► Well, someone ran Photoshop and used it and clicked around and did stuff, right?

00:19:36 ◼   ► How many tweaks to the output do you have to do for it to suddenly become a copyrightable,

00:19:39 ◼   ► right?

00:19:40 ◼   ► I think that is actually more fuzzy.

00:19:41 ◼   ► What is not fuzzy, in my opinion, and again, who knows what the law will be, but in my

00:19:45 ◼   ► personal opinion, is not fuzzy at all that you cannot take—that you can't take works

00:19:51 ◼   ► that you don't have the rights to and use it to train an AI model.

00:19:54 ◼   ► and I talked about this on upgrade a while back,

00:19:56 ◼   ► like it just, in my opinion, there's no,

00:20:00 ◼   ► like I see where people think it might be fuzzy,

00:20:02 ◼   ► but the difference is that it's not a person doing,

00:20:04 ◼   ► you're feeding it into a program,

00:20:05 ◼   ► and the value of that program is nothing

00:20:08 ◼   ► without being fed input.

00:20:10 ◼   ► You have to feed input into the program.

00:20:12 ◼   ► If I just give you that program and say,

00:20:14 ◼   ► well, it's a program that I wrote,

00:20:15 ◼   ► so anything that comes out of it is mine,

00:20:16 ◼   ► sure, okay, go ahead, but it's like,

00:20:18 ◼   ► okay, well, my program doesn't work unless I feed it data.

00:20:20 ◼   ► Okay, well, now what data,

00:20:21 ◼   ► where are you getting that data from?

00:20:23 ◼   ► I'll just feed it any data I feel like.

00:20:25 ◼   ► The point of the acquisition of data,

00:20:27 ◼   ► you are taking that data and using it.

00:20:29 ◼   ► It's like taking a font and using it to make a sign.

00:20:32 ◼   ► Like fonts, this is more complicated

00:20:34 ◼   ► 'cause you can't copyright a font.

00:20:35 ◼   ► But anyway, if you wanna use the official version

00:20:37 ◼   ► of a font or whatever, you have to pay for it.

00:20:39 ◼   ► You can use a clone of the font or whatever,

00:20:41 ◼   ► which gets things complicated.

00:20:42 ◼   ► But anyway, I don't think you can take people's

00:20:45 ◼   ► copyrighted material and use it to train an AI model.

00:20:47 ◼   ► I just don't, without some kind of compensation for them,

00:20:50 ◼   ► because you are directly deriving value from their work.

00:20:55 ◼   ► And I don't think the analogy to people makes any sense

00:20:57 ◼   ► because the program's not a person.

00:20:58 ◼   ► Yes, a person can look at art and make derivative art,

00:21:00 ◼   ► right, but it's not a person.

00:21:02 ◼   ► It's a person using a program, right?

00:21:04 ◼   ► It's like, well, if I use a program to copy this,

00:21:07 ◼   ► I didn't copy your art and resell it.

00:21:09 ◼   ► I used the CP program, and the CP program copied your art,

00:21:12 ◼   ► but then once I had it from the CP program,

00:21:14 ◼   ► then I sold it, but it's fine

00:21:16 ◼   ► because the CP program did just what a person does,

00:21:18 ◼   ► looks at the art, right?

00:21:19 ◼   ► No, it's just a more complicated version of the CP.

00:21:22 ◼   ► Now, will lawyers agree?

00:21:25 ◼   ► I think it really depends on, especially in our country,

00:21:27 ◼   ► how rich the corporation is on each side of the case

00:21:30 ◼   ► when it comes to, like if Disney arrives and says,

00:21:32 ◼   ► "Hey, we see you've trained your AI model

00:21:34 ◼   ► "on all of our intellectual property,"

00:21:35 ◼   ► probably Disney's gonna win, based on the precedent.

00:21:39 ◼   ► And in this case, I think Disney would be right

00:21:41 ◼   ► that you can't train your model on stuff made by Disney.

00:21:45 ◼   ► But as I said, an upgrade, Disney can train an AI model

00:21:48 ◼   ► on stuff made by Disney.

00:21:49 ◼   ► Like I'm not saying AI models are useless.

00:21:51 ◼   ► Disney can train all the models they want

00:21:53 ◼   ► on all their own intellectual property

00:21:55 ◼   ► and use it to generate stuff.

00:21:57 ◼   ► And I think the stuff that Disney generates

00:21:59 ◼   ► from Disney's own intellectual property

00:22:01 ◼   ► should be copyrightable because humans are running that

00:22:04 ◼   ► and doing the prompts and tweaking the output and whatever.

00:22:06 ◼   ► And I think that should be copyrightable.

00:22:08 ◼   ► So, so far, the one legal case that we've heard of

00:22:12 ◼   ► has gone the opposite way as I expected.

00:22:14 ◼   ► But this is a, you know, I think you're right, Marco,

00:22:16 ◼   ► that lots of people have different opinions on this.

00:22:18 ◼   ► I just feel like mine is very strongly against using other people's intellectual property.

00:22:23 ◼   ► And I think the problem with this stable diffusion thing is it's kind of admitting that, hey,

00:22:29 ◼   ► we probably shouldn't be using your stuff because they will immediately stop using it

00:22:32 ◼   ► if you ask them.

00:22:33 ◼   ► So based on that silly policy, they're saying, oh, we understand that you have the right

00:22:38 ◼   ► to make us stop.

00:22:39 ◼   ► Or maybe they're just trying to say we're being nice and we'll stop if you want.

00:22:42 ◼   ► But given my opinion that it is totally wrong for them to be taking that stuff in the first

00:22:46 ◼   ► I think this is a dumb policy.

00:22:48 ◼   ► - I think this is one of those areas where

00:22:51 ◼   ► we're gonna look back at whatever our opinions are now,

00:22:54 ◼   ► we're gonna look back in five or 10 years

00:22:55 ◼   ► and be like, man, we were way off.

00:22:57 ◼   ► And I don't know which direction we're gonna go in,

00:22:59 ◼   ► but I think this is gonna be a rapidly evolving policy

00:23:04 ◼   ► and what's acceptable and what's legal

00:23:08 ◼   ► are both also going to rapidly evolve.

00:23:10 ◼   ► I agree with you on the output being copyrightable.

00:23:15 ◼   ► again, it's like, if you use Photoshop to flood fill an area with a pattern, does that

00:23:19 ◼   ► not qualify for copyright because the program made the pattern and you didn't? So that

00:23:24 ◼   ► obviously, I think using it, I think the output is copyrightable, but I think the input is

00:23:29 ◼   ► really questionable and it's not clear cut. I think one area where it could kind of fall

00:23:37 ◼   ► legally into some kind of settlement is that when you watch a movie and you get the little

00:23:43 ◼   ► FBI warrant at the beginning, there are certain rights that you have and certain rights that

00:23:47 ◼   ► you don't. You are not, for instance, allowed to buy a copy of the movie on Blu-ray or whatever

00:23:53 ◼   ► and then show it to an audience who has paid to be there. You aren't allowed to use it

00:23:57 ◼   ► commercially. You aren't allowed to do public performances. And the same thing applies to

00:24:03 ◼   ► most music and things like that. And so you could argue, if you are a copyright holder,

00:24:08 ◼   ► You could maybe argue, well, the use of your model viewing or listening to my content is

00:24:15 ◼   ► not licensed because it's not personal, private home use or whatever.

00:24:20 ◼   ► So they could maybe get something in there, but I think overall the concept, I still currently

00:24:25 ◼   ► at least, I still fall on the side of the AI is just watching and learning from things

00:24:31 ◼   ► in a similar, well, maybe not too similar, but it's in an analogous way to how humans

00:24:37 ◼   ► might see something and be inspired by it over time.

00:24:40 ◼   ► But what do you have to say about the idea that the program doesn't have any rights?

00:24:43 ◼   ► Because in our law system, computer programs don't have rights right now, right? And I

00:24:48 ◼   ► think this specific computer program definitely shouldn't have rights, because it's not a

00:24:51 ◼   ► sentient being. So yes, it is doing something similar to what humans do, broad strokes,

00:24:56 ◼   ► but it doesn't have any rights, because it is not a sentient being, and all of our laws

00:25:00 ◼   ► apply to sentient beings, right? No, for that, I mean, I guess I agree with

00:25:04 ◼   ► but I don't see this as whether the AI has rights itself.

00:25:09 ◼   ► I see this as like--

00:25:11 ◼   ► - You're trying to give it creative control.

00:25:12 ◼   ► It's saying it created this thing,

00:25:14 ◼   ► therefore now the program has rights to it.

00:25:15 ◼   ► I mean, if you wanted to say that, I would say great.

00:25:17 ◼   ► Now the program can do whatever it wants with it.

00:25:19 ◼   ► You're like, well, the program doesn't wanna do anything

00:25:22 ◼   ► with it, the human that ran the program, though,

00:25:24 ◼   ► wants to sell it for an in and out purchase.

00:25:26 ◼   ► So it's like, okay, if you actually want to assign rights,

00:25:29 ◼   ► assign them to the program, and then you'd be like,

00:25:30 ◼   ► well, that's dumb, the program can't do anything.

00:25:32 ◼   ► - Exactly, that's why I don't have any rights.

00:25:35 ◼   ► - Well, but I think it's more complicated than that.

00:25:38 ◼   ► Like when you dive into the technical part of it,

00:25:40 ◼   ► like okay, well what about when Google indexes

00:25:43 ◼   ► fact data off of pages, and then you do a fact query

00:25:46 ◼   ► and it shows one of those Google knowledge things.

00:25:48 ◼   ► That's a questionable area of copyright as well, right?

00:25:51 ◼   ► But I think that's very, very similar to this.

00:25:53 ◼   ► - It is, but see, here's the thing.

00:25:55 ◼   ► You're saying how this turns out,

00:25:56 ◼   ► it can go a couple different ways.

00:25:58 ◼   ► It really depends on, in the case of the Google thing,

00:26:02 ◼   ► When that case comes--

00:26:03 ◼   ► I don't remember what the resolution of the case--

00:26:05 ◼   ► I think Google won that one.

00:26:06 ◼   ► But when the cases come out, there are two strong forces

00:26:08 ◼   ► here.

00:26:09 ◼   ► One is rich corporations want something.

00:26:12 ◼   ► In this case, Google is the rich corporation.

00:26:14 ◼   ► It wanted to be able to take summaries of web pages

00:26:17 ◼   ► and show them as its own output, which is arguably,

00:26:20 ◼   ► you're right, very similar to what these models do.

00:26:22 ◼   ► But I'm pretty sure--

00:26:23 ◼   ► and we'll have all of this for a moment--

00:26:25 ◼   ► I'm pretty sure Google won that case.

00:26:27 ◼   ► And did they win it because they were right,

00:26:28 ◼   ► or did they win it because they're Google,

00:26:30 ◼   ► and they're rich, and they have a lot of lawyers?

00:26:31 ◼   ► Right.

00:26:32 ◼   ► see also Disney and copyright.

00:26:34 ◼   ► - Right, exactly, right.

00:26:35 ◼   ► So I think, but in using the analogy to what Google does,

00:26:38 ◼   ► I kind of agree that Google, it's a little bit fuzzier,

00:26:42 ◼   ► but because Google's whole thing is there,

00:26:44 ◼   ► every one of their search results

00:26:45 ◼   ► is arguably a summary or whatever,

00:26:47 ◼   ► but Google is trying to be a venue for you to go somewhere,

00:26:51 ◼   ► and by putting the summary things,

00:26:52 ◼   ► they're making you not go there.

00:26:53 ◼   ► When a program is summarizing

00:26:58 ◼   ► and providing like an index or whatever,

00:27:00 ◼   ► If you follow that road too far down, you're like,

00:27:02 ◼   ► oh, you can't have search engines anymore

00:27:03 ◼   ► because you're not allowed to reproduce

00:27:05 ◼   ► any portion of my work, even a summary of it

00:27:07 ◼   ► if it's programmatically generated anywhere.

00:27:08 ◼   ► And that gets absurd.

00:27:09 ◼   ► Obviously, we think we should have search engines,

00:27:11 ◼   ► so let's not make a dumb law like that.

00:27:12 ◼   ► - You can't even have browser caches

00:27:14 ◼   ► if you follow that law too closely.

00:27:16 ◼   ► - Right, but then there, but you know,

00:27:17 ◼   ► this is the kind of technical nuance

00:27:19 ◼   ► that laws aren't great with, okay?

00:27:20 ◼   ► But what about when they take my webpage

00:27:21 ◼   ► and they don't even provide a link to it

00:27:23 ◼   ► and they just summarize the content

00:27:24 ◼   ► and re-smash it through and say, oh, here's the answer

00:27:27 ◼   ► and it's provided by Google

00:27:28 ◼   ► and they don't give credit to my answer

00:27:29 ◼   ► I'm the one who wrote that like, you know, "Grapes are poison to dogs" on my webpage,

00:27:33 ◼   ► and they just took my paragraph and summarized it or massaged it and put it on there, right?

00:27:38 ◼   ► And I, you know, again, I think Google won that case because they're rich, not because

00:27:41 ◼   ► they're right.

00:27:42 ◼   ► I would have decided that differently, but it's a nuanced ruling in that case, very nuanced,

00:27:45 ◼   ► because I don't want to outlaw search engines or browser caches.

00:27:47 ◼   ► In the case of the, you know, Disney finds someone's been training their AI models on,

00:27:52 ◼   ► you know, Marvel superheroes, which absolutely has already happened, I'm sure, right?

00:27:57 ◼   ► And Disney gets angry and sues.

00:27:59 ◼   ► Disney's probably going to win that because they have all the politicians in their pocket

00:28:02 ◼   ► or whatever the paraphrase of the line from The Godfather is.

00:28:07 ◼   ► But on the other side of that, unlike the Google case, there is another side to that

00:28:11 ◼   ► that is potentially just as powerful, which is people like free stuff.

00:28:16 ◼   ► And combined with the fact that it's very difficult to – it's not immediately obvious

00:28:21 ◼   ► what has been used to train an AI model.

00:28:23 ◼   ► And then mixed into that, these sort of academic things, which is like, well, they're not

00:28:26 ◼   ► selling anything that's just part of academic research and they should have broader purview

00:28:32 ◼   ► to do things for non-commercial purposes or whatever, right?

00:28:34 ◼   ► But the fact that it's just going to be like, "Well, Disney can sue, but they can try to

00:28:39 ◼   ► make laws in the US, but the world is bigger than the US and anything that gets on the

00:28:43 ◼   ► internet is everywhere, so good luck trying to stop people from making pictures of pregnant

00:28:47 ◼   ► Sonic."

00:28:48 ◼   ► Right?

00:28:49 ◼   ► Yeah, pictures of pregnant Sonic probably Sega would not like to exist, but people are

00:28:52 ◼   ► going to make them and it's a derivative work.

00:28:54 ◼   ► That's probably a bad example.

00:28:55 ◼   ► But anyway, these AI models are going to create images, and they're going to go everywhere

00:29:01 ◼   ► if people want them.

00:29:02 ◼   ► If it is a useful thing to have, it's going to be so hard to stop them.

00:29:05 ◼   ► So the countervailing force here is at a certain point the case is coming, they're going to

00:29:10 ◼   ► say, "Well, we think this should probably be illegal, but it's just so common and people

00:29:13 ◼   ► really like it."

00:29:14 ◼   ► And I know courts don't put an opinion that says, "Well, everybody does it, so it's fine."

00:29:18 ◼   ► But practically speaking, you can end up with decisions that kind of lean in that direction.

00:29:23 ◼   ► I'm sure there's some legal precedent of like, you know, like people who squat on a piece

00:29:27 ◼   ► of land and at a certain point they own it.

00:29:29 ◼   ► That doesn't make any sense.

00:29:30 ◼   ► You're like, "That doesn't seem fair."

00:29:31 ◼   ► But it's like, "But we have laws like that because it's even more absurd to do the opposite

00:29:34 ◼   ► in some cases."

00:29:35 ◼   ► Well, we've been living on this land for 100 years, but you know, you're going to kick

00:29:38 ◼   ► us off because, well, again, in this country there's terrible analogies for that as well.

00:29:42 ◼   ► Anyway, I'm not doing well with analogies.

00:29:44 ◼   ► It's early in the morning.

00:29:45 ◼   ► What can I tell you?

00:29:46 ◼   ► So it depends on when these cases come out.

00:29:48 ◼   ► I feel like there actually is something that may be equivalent to Disney trying to say,

00:29:54 ◼   ► "Hey," again legally, "Hey, we've decided that this is okay."

00:29:58 ◼   ► Regardless of what the law ends up being, I think it is not right to take people's copyrighted

00:30:02 ◼   ► work, train a model on it, and sell the results of that model's output.

00:30:06 ◼   ► If that model ever becomes a sentient being, which is plausible in the future, then we'll

00:30:10 ◼   ► have a different discussion.

00:30:11 ◼   ► That discussion will be what kind of rights do sentient computer programs have.

00:30:14 ◼   ► And that is a very different discussion,

00:30:16 ◼   ► but I want to be clear,

00:30:17 ◼   ► none of these programs are sentient,

00:30:19 ◼   ► no one is arguing that they are.

00:30:21 ◼   ► So we can set that aside for a sci-fi future,

00:30:24 ◼   ► but that's not what we're dealing with now.

00:30:26 ◼   ► - Yeah, see, I see where you're coming from,

00:30:28 ◼   ► I don't agree.

00:30:29 ◼   ► I still think that--

00:30:30 ◼   ► - Would you agree if you were an artist

00:30:32 ◼   ► and your work was being taken

00:30:33 ◼   ► and then the results of it being sold?

00:30:35 ◼   ► - We are artists.

00:30:36 ◼   ► So what would you think about somebody downloading all,

00:30:39 ◼   ► all episodes of ATP we've ever made,

00:30:40 ◼   ► all the podcasts we've ever made,

00:30:42 ◼   ► and training a model on that?

00:30:44 ◼   ► - Yeah, I don't think that's something,

00:30:45 ◼   ► and selling the results,

00:30:46 ◼   ► I don't think that's something you should do.

00:30:47 ◼   ► What if they took the source code for Overcast

00:30:50 ◼   ► and fed it into a program that tweaked it

00:30:52 ◼   ► and then started selling new copies of Overcast?

00:30:55 ◼   ► - Well, I don't release the source code,

00:30:56 ◼   ► so that's a little bit different, but you know--

00:30:57 ◼   ► - Yeah, it can reverse compile it,

00:30:59 ◼   ► 'cause you released the binary

00:31:00 ◼   ► and they can just generate the source code from that.

00:31:01 ◼   ► You've all seen that, and the program doesn't care

00:31:03 ◼   ► what the variable names are all, A, B, C, D, and stuff,

00:31:05 ◼   ► because the program doesn't care about that,

00:31:06 ◼   ► but the program refacto, decompiles it,

00:31:09 ◼   ► refactores it, changes it a little bit,

00:31:11 ◼   ► and sells a new thing.

00:31:12 ◼   ► I said, "Derivative work."

00:31:12 ◼   ► just looked at it just the same way a person would.

00:31:14 ◼   ► - Well see, and that I think, that's why

00:31:17 ◼   ► I'm fairly hesitant to restrict the training side of things

00:31:20 ◼   ► because in my opinion, I would rather use the systems

00:31:24 ◼   ► we already have with copyright and fair use

00:31:27 ◼   ► and trademark also, which is a little bit relevant

00:31:30 ◼   ► in some of these things, but anyway,

00:31:31 ◼   ► I'd rather use the systems we already have

00:31:34 ◼   ► around copyright and fair use to manage

00:31:37 ◼   ► when things go wrong and when things are misused,

00:31:40 ◼   ► rather than say this new kind of technology

00:31:43 ◼   ► that's really exciting and has a lot of good valid uses,

00:31:45 ◼   ► we're gonna make it way harder

00:31:47 ◼   ► and way less practical to create.

00:31:49 ◼   ► Because if somebody uses an AI model

00:31:51 ◼   ► that's been trained on whatever they could find,

00:31:53 ◼   ► if they use that to generate something

00:31:55 ◼   ► that is too close to an existing copyrighted work

00:31:58 ◼   ► or that is clearly a derivative work

00:32:00 ◼   ► of a protected character or franchise or whatever,

00:32:02 ◼   ► then we already have infrastructure in place

00:32:05 ◼   ► to issue takedown notices, to sue for something being

00:32:09 ◼   ► or not being fair use, we already have all those systems.

00:32:12 ◼   ► And I think we are better off relying

00:32:14 ◼   ► on that human judgment side rather than

00:32:17 ◼   ► kneecapping this technology in its early days

00:32:20 ◼   ► in ways that are extremely impractical.

00:32:22 ◼   ► I mean, if you think about, look at a similar problem

00:32:25 ◼   ► that we faced when making web 2.0 and everything else.

00:32:29 ◼   ► We had the rise of user-generated content sites,

00:32:33 ◼   ► places like YouTube and Flickr and Tumblr

00:32:36 ◼   ► and all these places.

00:32:37 ◼   ► And you could say, legally speaking,

00:32:40 ◼   ► you could say it should be impossible

00:32:44 ◼   ► for anyone to ever upload a copy of my work

00:32:47 ◼   ► to these services.

00:32:48 ◼   ► In practice, that's really impractical and limiting.

00:32:51 ◼   ► And so, kind of what you said a minute ago, John,

00:32:54 ◼   ► so really the practice became,

00:32:56 ◼   ► all right, well we're gonna have the structure in place

00:32:57 ◼   ► to manage when there are violations,

00:32:59 ◼   ► but it's not gonna be perfect,

00:33:01 ◼   ► and it's not gonna be the intellectual property ideal

00:33:05 ◼   ► of having it be impossible to break the law,

00:33:07 ◼   ► but we're gonna just make a process in place

00:33:10 ◼   ► so that we can deal with when people do things

00:33:12 ◼   ► that are over the line.

00:33:13 ◼   ► - But their process is they have a program

00:33:15 ◼   ► called Content ID that will immediately detect

00:33:17 ◼   ► if you have music playing in the background

00:33:19 ◼   ► that's copyrighted, even if it's like

00:33:21 ◼   ► this next house over, right?

00:33:22 ◼   ► It is so incredibly fast and automated

00:33:25 ◼   ► and defaults to rejecting.

00:33:26 ◼   ► I mean, the YouTube video I wanted to see about Fusion

00:33:31 ◼   ► that I think I might have mentioned on a past show

00:33:33 ◼   ► was from Real Engineering, I think.

00:33:36 ◼   ► It was rejected by copyright ID from YouTube or whatever,

00:33:40 ◼   ► because the music used in the background was copyrighted.

00:33:44 ◼   ► And the music used in the background

00:33:45 ◼   ► was from a set of licensable music

00:33:47 ◼   ► that the channel had licensed.

00:33:49 ◼   ► So the copyright ID thing is so automated and so strict

00:33:53 ◼   ► and so defaulting to rejection that it says,

00:33:55 ◼   ► oh, this is copyright.

00:33:55 ◼   ► It's like, yeah, and I paid for it.

00:33:57 ◼   ► I paid for it to use on my channel.

00:33:59 ◼   ► But it's like, no, program says no.

00:34:00 ◼   ► So if it becomes possible to programmatically detect if your AI training model was trained

00:34:06 ◼   ► with any copyrighted data, no one will ever be able to upload any AI trained stuff to

00:34:11 ◼   ► YouTube, because YouTube will immediately detect it and reject it, even if you've paid

00:34:14 ◼   ► for the rights, because that's the side they err on because big corporations don't like

00:34:19 ◼   ► it.

00:34:20 ◼   ► It's more like, you know, okay, so that's YouTube, but then what about the wider internet,

00:34:23 ◼   ► like that it'll be everywhere.

00:34:24 ◼   ► And again, I think AI models are still useful and can be useful, particularly for companies

00:34:30 ◼   ► that want to train it on their own intellectual property.

00:34:32 ◼   ► Like it is useful to train things on stuff that you own.

00:34:34 ◼   ► And even an individual might not own enough to train a model,

00:34:37 ◼   ► but there are licensable image sets that you can train on.

00:34:40 ◼   ► That's also legally valid, right?

00:34:42 ◼   ► I don't think, to your point about kneecapping disorder,

00:34:44 ◼   ► I don't think we need any new laws.

00:34:46 ◼   ► My interpretation of the existing laws is,

00:34:48 ◼   ► this is already illegal, but the existing laws

00:34:50 ◼   ► also provide lots of ways that it's legal.

00:34:53 ◼   ► You know, Getty Images can sell image training sets.

00:34:56 ◼   ► There are free public domain image training sets.

00:34:59 ◼   ► if you would like to license an image training set from Disney, maybe they'll sell you one

00:35:03 ◼   ► and then you can train your model on that.

00:35:05 ◼   ► And then we have to have all the automated systems saying, "Okay, I can tell this was

00:35:08 ◼   ► trained on X, Y, and Z images, and X, Y, and Z images are from these licensable sets,"

00:35:13 ◼   ► and then you have to show that you licensed it just the same way the real engineer had

00:35:16 ◼   ► to say, "No, no, I licensed this music thing from this company.

00:35:20 ◼   ► I'm allowed to use it in my video.

00:35:21 ◼   ► This is how the system is supposed to work, YouTube.

00:35:24 ◼   ► Stop rejecting my videos."

00:35:25 ◼   ► Right?

00:35:26 ◼   ► and annoying, but if anything, if there is a programmatic way to enforce existing laws on fair use and intellectual property,

00:35:34 ◼   ► it will be worse for creativity and innovation than humans enforcing it, because computers are really, really good at doing stuff like detecting if there's copyrighted music.

00:35:43 ◼   ► And I presume eventually computers will be pretty good at figuring out if any copyrighted works from some set are part of a training set.

00:35:51 ◼   ► training set, because that's, you know, have I been trained or whatever, and maybe they

00:35:54 ◼   ► just have a database, I don't know how it works or whatever, but once it's programmatically

00:35:57 ◼   ► possible to answer this question, it swings farther in the direction of Disney stopping

00:36:02 ◼   ► you from ever training any image set on Star Wars.

00:36:05 ◼   ► Well, I mean, and again, I think using Content ID on YouTube as an example is, that's kind

00:36:10 ◼   ► of an extreme case, even though it is a highly relevant one, just because Content ID is so

00:36:13 ◼   ► overly aggressive, but you know, that's not, and if you think about also, like, some other

00:36:17 ◼   ► angle of this, you know, if Disney starts using this technology to like, you know, scan

00:36:22 ◼   ► everything that they can find on YouTube and be like, "Oh wait, that's Elsa. Oh, stop,

00:36:26 ◼   ► wait." And like, they can use the AI to detect derivative works even. Disney could start

00:36:31 ◼   ► scanning like DeviantArt and Flickr and stuff for like fan-drawn versions of their characters.

00:36:38 ◼   ► Well, but under existing laws, they have to show it was taken from like one of their actual

00:36:43 ◼   ► Well, drawing's not a picture of Elsa, right?

00:36:45 ◼   ► Although if you do want to put Elsa on a t-shirt and sell it,

00:36:48 ◼   ► you can't say, well, I drew the Elsa myself, so it's fine.

00:36:50 ◼   ► Actually, I don't know where the law is on that,

00:36:52 ◼   ► depending on how derivative the work is.

00:36:54 ◼   ► But anyway, what you would expect for the training thing

00:36:56 ◼   ► and how it would be trained is not just, hey,

00:36:58 ◼   ► you drew a picture of an apple, and there's

00:36:59 ◼   ► a picture of an apple in the training set.

00:37:00 ◼   ► It's like, is it literally your picture of the apple?

00:37:03 ◼   ► And that's where it gets tricky.

00:37:05 ◼   ► Again, I don't know how this have I been trained thing works.

00:37:08 ◼   ► Maybe it is a byte provider-- they

00:37:09 ◼   ► know the checksums of every image that

00:37:12 ◼   ► was part of the data set and you can upload one of your images

00:37:15 ◼   ► and if the checksum matches, then it's the same.

00:37:17 ◼   ► Is it literally byte for byte and if you resize your picture

00:37:20 ◼   ► or change one pixel, it doesn't match?

00:37:22 ◼   ► Or is it fuzzier than that?

00:37:23 ◼   ► This is kind of similar to the CSAM system

00:37:26 ◼   ► of perceptual hashing or whatever.

00:37:28 ◼   ► Because that gets into what you're saying.

00:37:30 ◼   ► It's like, OK, well, it's not from a Disney picture of Elsa.

00:37:33 ◼   ► It's a picture that I drew of Elsa.

00:37:35 ◼   ► And that's different.

00:37:37 ◼   ► And again, with Disney being overly aggressive, saying,

00:37:39 ◼   ► oh, great, we have a way to tell if anyone never

00:37:41 ◼   ► training set on any picture of Elsa.

00:37:44 ◼   ► I'm not sure where that falls under existing laws.

00:37:46 ◼   ► But is that even possible?

00:37:48 ◼   ► Is it possible for them to say,

00:37:50 ◼   ► we can tell that you trained your model on these images?

00:37:53 ◼   ► - I don't know if you can reverse it,

00:37:54 ◼   ► but I assume the people who train the models

00:37:57 ◼   ► know literally every single image that's in the image set.

00:37:59 ◼   ► It's a long list, but they probably,

00:38:01 ◼   ► they have the data somewhere,

00:38:02 ◼   ► and they could build a database based on that data

00:38:05 ◼   ► with either real content hashes, perceptual hashes.

00:38:08 ◼   ► That's how I assume have I been trained is working.

00:38:10 ◼   ► Like they know, if you train it on a data set

00:38:13 ◼   ► and then you delete the data set

00:38:14 ◼   ► and then you can be like, oh well Shrug,

00:38:16 ◼   ► I don't even know what it was trained on.

00:38:17 ◼   ► I think it's difficult to like,

00:38:19 ◼   ► difficult or impossible to reverse it to say,

00:38:21 ◼   ► oh, I can tell by looking at the output

00:38:22 ◼   ► what it was trained on.

00:38:23 ◼   ► I don't think that's possible.

00:38:24 ◼   ► But hopefully you keep around.

00:38:27 ◼   ► This is where you might need new laws.

00:38:28 ◼   ► Like, hey, if you wanna use an AI model,

00:38:30 ◼   ► you have to be able to prove something

00:38:32 ◼   ► about your training set.

00:38:33 ◼   ► And if you have an AI model and you say,

00:38:34 ◼   ► oh, we deleted our training set.

00:38:35 ◼   ► Oh, well, the law would say,

00:38:37 ◼   ► yeah, you can't use any of that.

00:38:38 ◼   ► because part of whatever our laws,

00:38:40 ◼   ► new laws that come out will say,

00:38:42 ◼   ► "Okay, here are the situations

00:38:44 ◼   ► "under which the output of an AI model is copyrightable,

00:38:48 ◼   ► "can be used successfully."

00:38:50 ◼   ► You probably have to be able to prove something

00:38:53 ◼   ► about the training set,

00:38:53 ◼   ► and ignorance would not be a defense of the law of the law.

00:38:56 ◼   ► They're like, "We don't even know

00:38:57 ◼   ► "what the training set was, so everything's fine, right?"

00:38:59 ◼   ► Yeah, probably not.

00:39:00 ◼   ► - I don't know, I'm still on the side of

00:39:02 ◼   ► this is addressing it at the wrong stage of the pipeline.

00:39:04 ◼   ► I mean, 'cause it's gonna be more complicated.

00:39:07 ◼   ► many new models that come out are not based on nothing

00:39:10 ◼   ► and trained from scratch, they're based on initial models

00:39:13 ◼   ► that were pre-trained by somebody else.

00:39:15 ◼   ► (laughing)

00:39:16 ◼   ► - That's part of what's in favor of the chaos

00:39:18 ◼   ► of the internet, it's like now we can't even tell,

00:39:19 ◼   ► what are you gonna do now?

00:39:20 ◼   ► - Right, you know, I think there's gonna be,

00:39:23 ◼   ► I think it's gonna be quickly very difficult

00:39:25 ◼   ► to ever tell how something was trained in completion.

00:39:29 ◼   ► And I think, again, I think this is as unenforceable

00:39:33 ◼   ► as it should be impossible to ever upload

00:39:35 ◼   ► copyrighted material to your platform.

00:39:36 ◼   ► Like that should, it's just impossible.

00:39:39 ◼   ► - But I mean, but that's more of like a practicality.

00:39:42 ◼   ► The legal thing is like, oh, you can't actually,

00:39:44 ◼   ► you don't actually own this copyrighted material.

00:39:46 ◼   ► The law is clear, it's just the practicalities

00:39:48 ◼   ► of the enforcement of the law are not clear.

00:39:49 ◼   ► And here I think the law is not clear yet.

00:39:52 ◼   ► - Right, yeah, because I still think it's definitely

00:39:55 ◼   ► a very open question about like whether training

00:39:58 ◼   ► an AI model on a copyrighted work is itself

00:40:02 ◼   ► a violation of that copyright or not.

00:40:04 ◼   ► Again, 'cause I still don't think it is.

00:40:06 ◼   ► Well, we'll see what the courts say,

00:40:08 ◼   ► but I'm pretty strongly on one side of this.

00:40:10 ◼   ► But, and again, this doesn't mean I'm against AI models.

00:40:12 ◼   ► I think they're super useful,

00:40:13 ◼   ► and I think there's lots of places where they can be used.

00:40:16 ◼   ► It's just that I think there should be some discipline

00:40:19 ◼   ► about the training, that's all.

00:40:20 ◼   ► We'll see.

00:40:21 ◼   ► - We are brought to you this week by Blaze.

00:40:25 ◼   ► Have you heard of blaze.tech?

00:40:27 ◼   ► Blaze.tech lets you build software and web apps

00:40:30 ◼   ► without writing code.

00:40:31 ◼   ► So really, that's it.

00:40:33 ◼   ► If you've been waiting around for an engineer

00:40:34 ◼   ► to help you build something,

00:40:36 ◼   ► You don't need to wait anymore.

00:40:37 ◼   ► Just use Blaze.tech and you can build software

00:40:40 ◼   ► without any coding knowledge at all.

00:40:43 ◼   ► Anyone with an entrepreneurial spirit

00:40:45 ◼   ► can build software with no code with Blaze.tech.

00:40:47 ◼   ► So if you want, you can also ask Blaze.tech's

00:40:50 ◼   ► world-class team to fully set up and launch your app.

00:40:52 ◼   ► That way you can focus on your business

00:40:54 ◼   ► and Blaze.tech can take care of your app

00:40:56 ◼   ► and it's super easy to use.

00:40:58 ◼   ► They have a drag and drop builder,

00:41:00 ◼   ► integration with every popular API you can think of

00:41:02 ◼   ► so you can create and launch apps in minutes.

00:41:05 ◼   ► People have used Blaze to build things like healthcare apps, inventory management systems,

00:41:10 ◼   ► customer portals, HR and recruiting apps, automated workflows, and tons of other software,

00:41:16 ◼   ► all without coding or needing an engineer to help them.

00:41:19 ◼   ► They are super secure also, so healthcare, legal, and finance organizations love them.

00:41:24 ◼   ► So check out Blaze.tech.

00:41:26 ◼   ► Both startups and Fortune 500s are using Blaze.tech as their secret weapon to quickly create and

00:41:32 ◼   ► launch new software.

00:41:33 ◼   ► Sign up for Blaze.tech at blaze.tech/ATP for 10% off.

00:41:38 ◼   ► Once again, blaze.tech/ATP for 10% off.

00:41:42 ◼   ► Pricing starts at just 25 bucks per user per month.

00:41:46 ◼   ► Thank you so much to Blaze.tech for sponsoring our show.

00:41:50 ◼   ► - Ted Shop writes that we would probably get a kick

00:41:57 ◼   ► out of a new Twitter to Mastodon web app,

00:42:00 ◼   ► and this is @mastedon-flock.vercel.app.

00:42:04 ◼   ► I didn't try it past the first screen or two,

00:42:07 ◼   ► but as a former Windows user, oh, the nostalgia.

00:42:11 ◼   ► Oh my word.

00:42:13 ◼   ► This made me laugh quite a bit.

00:42:14 ◼   ► And it looks like a Windows 95 installation wizard.

00:42:17 ◼   ► Big, Nsys, like Nolsoft install system energy.

00:42:21 ◼   ► I remember I used to play with that,

00:42:23 ◼   ► probably around the time that Marco and I--

00:42:24 ◼   ► - No, Nsys was what replaced this.

00:42:26 ◼   ► That was much better.

00:42:27 ◼   ► Wasn't Nsys not full screen?

00:42:29 ◼   ► - That's true, I think you're right.

00:42:30 ◼   ► This is that full screen blue gradient that was used a lot around that time.

00:42:35 ◼   ► And now this is, you could tell that they installed the driver for their GPU before they ran this wizard

00:42:40 ◼   ► because the background gradient is not super dithered in 16 color mode.

00:42:45 ◼   ► I thought this was really interesting because, well, two levels.

00:42:48 ◼   ► One, it is just like Move to Dawn in that it gives you follow on follow links.

00:42:52 ◼   ► So ignore the whole Windows 95 look or whatever.

00:42:54 ◼   ► It is another tool that is like Move to Dawn.

00:42:56 ◼   ► So Move to Dawn never goes away and you want to try something similar,

00:42:59 ◼   ► this does the same thing, it authenticates with all your things, then it gives you follow

00:43:03 ◼   ► buttons for all the people who have the info.

00:43:06 ◼   ► I don't know if it does a better job of scraping or whatever.

00:43:08 ◼   ► But the second thing is, part of the gag of it looking like Windows 95, it's not as far

00:43:15 ◼   ► as I can tell, maybe you can tell me if you're familiar with Windows 95, it's not a pixel

00:43:20 ◼   ► for pixel reproduction, it's almost like it's vector.

00:43:23 ◼   ► They have the Windows 95 look, but they've done it with vector pixels, because it looks

00:43:28 ◼   ► way bigger, it's scaled up, but it is pixel accurate if you imagine the pixels being these

00:43:34 ◼   ► smooth vector-y things.

00:43:36 ◼   ► Well, if you zoom in on the next and cancel buttons, on the corners, you see that they

00:43:41 ◼   ► have like a beveled edge.

00:43:44 ◼   ► Now in actual Windows 95, that would have been probably a one pixel thick line.

00:43:48 ◼   ► And so here they have like a beveled edge in the corners, so obviously it's like a CSS

00:43:51 ◼   ► border trick or something.

00:43:52 ◼   ► Yeah, and I'm on the final page, and I'm looking at the headers of the table, and the headers

00:43:58 ◼   ► have like a 3D gradient on them and you can see that there is like a you know a

00:44:02 ◼   ► dark line on the bottom and lighter line on the left and top and that line is

00:44:06 ◼   ► made of a chunky like would have been like a Windows 95 size pixel but then

00:44:11 ◼   ► the bevel is made with pixels that are 1/8 of that size that would not exist in

00:44:15 ◼   ► Windows 95 right because they're on an angle so I think it's very clever cut

00:44:19 ◼   ► it's a little bit of an uncanny valley thing I'm like this looks like Windows

00:44:22 ◼   ► 95 but Windows 95 didn't never look like this down to the pixel anyway I thought

00:44:26 ◼   ► it was a little extra fun sauce on top of a useful tool.

00:44:31 ◼   ► And you can just in case the Move

00:44:33 ◼   ► to Don goes away.

00:44:34 ◼   ► But it's one of the show notes if you want to try another one

00:44:35 ◼   ► or get a little bit of weird uncanny value

00:44:37 ◼   ► Windows 95 nostalgia.

00:44:39 ◼   ► - And while we're on the subject,

00:44:39 ◼   ► Massodon continues to be great, by the way.

00:44:42 ◼   ► It just continues to be awesome.

00:44:44 ◼   ► And it's way better than Twitter right now,

00:44:47 ◼   ► in my view and for my needs and for the people I follow,

00:44:50 ◼   ► it's way better than Twitter.

00:44:52 ◼   ► So I'm happy there.

00:44:54 ◼   ► - Briefly going back to this web app,

00:44:55 ◼   ► Did you notice that if you click the X on the installer,

00:44:58 ◼   ► it actually drops you down to Windows desktop

00:44:59 ◼   ► and it has a privacy policy link.

00:45:02 ◼   ► - Oh my God. - An about link

00:45:03 ◼   ► that opens Windows Explorer,

00:45:06 ◼   ► that opens Internet Explorer.

00:45:07 ◼   ► And then if you click on the home page,

00:45:08 ◼   ► or no, not the home page, what was it?

00:45:10 ◼   ► It was one of these icons.

00:45:11 ◼   ► Oh, I guess it was the bookmarks icon at the top.

00:45:14 ◼   ► Then you go to what appears to be

00:45:16 ◼   ► the author's like old 1999 website,

00:45:20 ◼   ► which is pretty delightful.

00:45:22 ◼   ► Big like GeoCity's energy on this one.

00:45:24 ◼   ► There's a lot more here than I thought.

00:45:25 ◼   ► This is really great.

00:45:26 ◼   ► Oh, that's so cool.

00:45:28 ◼   ► Yeah, isn't that neat?

00:45:30 ◼   ► The other way I would have imagined

00:45:31 ◼   ► doing this that would have been fun

00:45:33 ◼   ► is all the ones that run older operating systems in Web

00:45:37 ◼   ► Assembly in your browser.

00:45:39 ◼   ► There's one for all the versions of Mac OS.

00:45:41 ◼   ► It's like macos-9.app, system7.app, macos-8.app.

00:45:44 ◼   ► Try those URLs.

00:45:46 ◼   ► And those are basically running a virtual machine in JavaScript

00:45:48 ◼   ► that's running the real version of those operating systems.

00:45:52 ◼   ► If you had done that and then wrote

00:45:54 ◼   ► moved it on in Windows 95 and ran it inside the Windows 95

00:45:58 ◼   ► that's running in your browser,

00:45:59 ◼   ► then you wouldn't have to fake any of this stuff.

00:46:01 ◼   ► You could literally just make everything exactly

00:46:05 ◼   ► in Windows 95, although that might be more difficult

00:46:07 ◼   ► trying to authenticate with modern SSL stuff

00:46:10 ◼   ► from Windows 95.

00:46:12 ◼   ► - Yeah, I was gonna say, can you imagine trying to make

00:46:14 ◼   ► just any network request that would work

00:46:16 ◼   ► in the modern era from that?

00:46:18 ◼   ► - I think some, I saw someone who did a 68K Motorola 68K

00:46:22 ◼   ► classic Mac OS Mastodon client, and I think somehow they compiled a version of OpenSSL

00:46:28 ◼   ► or something. They basically found a way to make a TLS connection from classic Mac OS

00:46:33 ◼   ► just to support their idea of doing this funny classic Mac OS client from Mastodon. That's

00:46:38 ◼   ► amazing.

00:46:39 ◼   ► That's pretty good stuff. Alright, and then let's see what else we have. Oh, that's right.

00:46:45 ◼   ► I wanted to make note, this news is actually a little bit old. We spoke about rewind.ai

00:46:50 ◼   ► I don't know, a couple months ago, this is the livestreams thing that John apparently

00:46:54 ◼   ► has always wanted.

00:46:56 ◼   ► It is available to anyone at this point, I believe.

00:46:59 ◼   ► It is 30 days free trial, but then it's $20 a month, which I don't mean to begrudge.

00:47:05 ◼   ► They've written software, they deserve to be compensated for it, but I don't know.

00:47:09 ◼   ► I don't love the idea of this 30 days free, and then you have to come back and remember

00:47:14 ◼   ► to cancel.

00:47:15 ◼   ► So I have not personally tried it, especially since this is not an itch I feel like I need

00:47:18 ◼   ► But for those that are interested,

00:47:21 ◼   ► it's available if you wanted to give it a shot.

00:47:23 ◼   ► - It ain't begrudging anything.

00:47:24 ◼   ► It's just the question of, you know,

00:47:25 ◼   ► do you think the price is worth the value for you?

00:47:28 ◼   ► And I was into this thing.

00:47:30 ◼   ► I'm like, oh, I'm gonna try it out of technical curiosity.

00:47:32 ◼   ► And I actually got in early and they sent the email

00:47:34 ◼   ► and said, hey, it's available for you to try.

00:47:35 ◼   ► I'm like, great, I'll try it.

00:47:36 ◼   ► And then I saw it was one of those free trial

00:47:37 ◼   ► and then we'll start charging you.

00:47:39 ◼   ► And I knew myself well enough to know,

00:47:41 ◼   ► I'm probably gonna forget and then I'm gonna get charged

00:47:43 ◼   ► and then I'm gonna feel bad.

00:47:44 ◼   ► In fact, I did that with the Lenza app, speaking of AI.

00:47:47 ◼   ► make money off other people's stuff.

00:47:49 ◼   ► The Lenza thing, like you signed up for a free trial

00:47:52 ◼   ► of the $50 a month subscription.

00:47:54 ◼   ► - Oh my goodness.

00:47:55 ◼   ► (laughing)

00:47:56 ◼   ► - And then, or it was $50 a month or $50 a year, I forget.

00:48:00 ◼   ► But anyway, you sign up for the free trial,

00:48:01 ◼   ► and then once you sign up for the free trial,

00:48:03 ◼   ► it gives you a discount on the in-app purchases.

00:48:05 ◼   ► And so I made like several $6 in-app purchases

00:48:09 ◼   ► to play with it, right?

00:48:10 ◼   ► But then I'm like, oh hey, I just gotta remember

00:48:11 ◼   ► to cancel this before the $50 thing.

00:48:13 ◼   ► And sure enough, I forgot,

00:48:15 ◼   ► 'cause I didn't make a reminder for myself.

00:48:16 ◼   ► And I got a refund for it because I asked for a refund the second it kicked in and Apple

00:48:20 ◼   ► was nice enough to give me a refund.

00:48:22 ◼   ► But anyway, that's why I haven't tried to rewind.

00:48:24 ◼   ► I am curious about it.

00:48:25 ◼   ► I'm like, "I don't want to have to sign up for it and set myself a reminder, remember

00:48:30 ◼   ► to cancel it."

00:48:31 ◼   ► It's just like, for me, that price and that situation was enough of a barrier to not want

00:48:38 ◼   ► to deal with it.

00:48:39 ◼   ► And honestly, since we've talked about it, I'm like, "Oh, I'll try for technical curiosity."

00:48:42 ◼   ► I'm still kind of skeeved out by it.

00:48:44 ◼   ► I'm like, "Well, I'll try it on my laptop.

00:48:45 ◼   ► It's more isolated."

00:48:46 ◼   ► But not that I think they're doing anything nasty.

00:48:48 ◼   ► It's just, I don't know.

00:48:51 ◼   ► I probably will try it eventually as well.

00:48:53 ◼   ► In terms of $20 a month, if I got tons of value from it, I would pay $20 a month if

00:48:57 ◼   ► I felt like I was getting $20 a month worth of value out of it.

00:48:59 ◼   ► That's the question of software.

00:49:01 ◼   ► How do you price it such that enough people will think that price is acceptable?

00:49:05 ◼   ► Casey was also scared away by that price.

00:49:08 ◼   ► But I think none of us are in the situation where we think we would possibly...

00:49:12 ◼   ► The only reason we'd be looking at it is a little bit of curiosity, right?

00:49:14 ◼   ► And so, maybe we'll try it.

00:49:17 ◼   ► I can see this, like I said, this company eventually being spun off into like an enterprise

00:49:21 ◼   ► software because I think lots of companies might like something like this so they can

00:49:25 ◼   ► spy on their employees.

00:49:26 ◼   ► This is a rich and burgeoning market that already has existing players that are probably

00:49:30 ◼   ► worse than Rewind.ai because the technology is newer and better.

00:49:33 ◼   ► But so far, their sales pitch does not seem to be to corporations, it seems to be to individual

00:49:40 ◼   ► users.

00:49:41 ◼   ► If you're an individual user that's already tried it,

00:49:43 ◼   ► write in and tell us what you think.

00:49:45 ◼   ► John, real-time follow-up, you cannot run this on your Mac Pro

00:49:48 ◼   ► because it is not sufficiently powerful.

00:49:50 ◼   ► I have ARM Macs in the house.

00:49:52 ◼   ► OK, well, you said you would run it on your laptop.

00:49:54 ◼   ► I thought you were trying to imply that you would have

00:49:56 ◼   ► otherwise run it on the Mac Pro.

00:49:57 ◼   ► And that is--

00:49:58 ◼   ► The Mac Studio is another choice.

00:50:00 ◼   ► You know, I have a couple of ARM-- yeah, some.

00:50:02 ◼   ► Well, because your Mac Pro is just too old and busted,

00:50:04 ◼   ► doesn't have Apple Silicon, which

00:50:05 ◼   ► is required for rewind.ai.

00:50:07 ◼   ► It's got an immunity to the spyware.

00:50:10 ◼   ► (laughing)

00:50:12 ◼   ► - All right, John, tell me about your iPad mystery.

00:50:17 ◼   ► - My son got a new iPad for Christmas.

00:50:18 ◼   ► He was long overdue for one.

00:50:19 ◼   ► - Which one?

00:50:20 ◼   ► - He was using, well, that's part of the thing.

00:50:21 ◼   ► He was using a really old one.

00:50:23 ◼   ► He'd been using it for years.

00:50:24 ◼   ► I actually, he wanted to draw with the Apple Pencil,

00:50:27 ◼   ► so I had an old, like the original 9.7-inch iPad Pro,

00:50:31 ◼   ► like the very first one that,

00:50:32 ◼   ► I think it was the first one that had pencil support.

00:50:33 ◼   ► - Well, the 12.9, that was the first small one.

00:50:36 ◼   ► Yeah, that's right.

00:50:38 ◼   ► So anyway, I said, here, you can use this,

00:50:40 ◼   ► and use my old, roly Apple pencil,

00:50:41 ◼   ► and laugh at how it gets charged and everything.

00:50:44 ◼   ► And I'm like, well, you should just use that as your new iPad,

00:50:46 ◼   ► because it's better than--

00:50:48 ◼   ► his iPad was so old, it was older than the 9.7-inch iPad

00:50:51 ◼   ► Pro.

00:50:52 ◼   ► But eventually, he said, no, there's

00:50:54 ◼   ► something haunted on your iPad, which I kind of agreed with him

00:50:57 ◼   ► that 9.7 started getting weird in its later years.

00:50:59 ◼   ► I don't know if it's a hardware thing,

00:51:00 ◼   ► or it just doesn't agree with the modern OSes.

00:51:02 ◼   ► But he kept using super-ancient iPads.

00:51:04 ◼   ► So he was due for a new one and I got him a new one.

00:51:07 ◼   ► And I bought it a while ago.

00:51:08 ◼   ► And it was just sitting somewhere and then wrapped it and gave it to him.

00:51:11 ◼   ► And then I was setting it up for him, because that's what dads do.

00:51:14 ◼   ► I was trying to give it a name.

00:51:17 ◼   ► And I wrote general about it and wrote the name of the thing.

00:51:21 ◼   ► I remembered to change the name so it's easy to find.

00:51:24 ◼   ► And the device list.

00:51:25 ◼   ► And I started saying, "Oh, wait a second.

00:51:28 ◼   ► Which iPad?"

00:51:29 ◼   ► Because I like to give it the name of the thing.

00:51:33 ◼   ► My wife says like, you know,

00:51:34 ◼   ► Tina's Apple Watch Series 8.

00:51:36 ◼   ► So I know it's the Series 8 one.

00:51:37 ◼   ► Like, people ask that as an AskATD question.

00:51:39 ◼   ► How do you name your devices?

00:51:40 ◼   ► I name it John's iPad 14 Pro.

00:51:43 ◼   ► Like, 'cause otherwise, or iPhone 14 Pro.

00:51:45 ◼   ► 'Cause otherwise I'll never friggin' remember

00:51:47 ◼   ► what the names are and I can't just reuse the same name.

00:51:49 ◼   ► I give them the name that is named with the device.

00:51:51 ◼   ► - Yeah, it's nice too, like when you down the road

00:51:53 ◼   ► when you upgrade that device,

00:51:55 ◼   ► it's nice to be able to like keep the backups straight

00:51:57 ◼   ► and like do the migration correctly and everything.

00:51:59 ◼   ► It makes it very nice.

00:52:00 ◼   ► - Yeah, and I do rename them to be like,

00:52:02 ◼   ► okay now it's Alex's iPhone 14 Pro, right?

00:52:05 ◼   ► But it's still iPhone 14 Pro,

00:52:06 ◼   ► so I know which device we're talking about.

00:52:07 ◼   ► And luckily in the house we tend not to buy

00:52:11 ◼   ► multiple devices of the same type,

00:52:13 ◼   ► just because people end up getting things

00:52:14 ◼   ► at different years, but if we had like two iPhone 12 Pros,

00:52:16 ◼   ► that might be a little bit confusing.

00:52:18 ◼   ► Maybe I would put the color in then.

00:52:20 ◼   ► Anyway, I wanted to give this thing a name of what it was.

00:52:24 ◼   ► So I named it Alex's M2 iPad Pro.

00:52:27 ◼   ► 'Cause it was like, I got him the fancy one,

00:52:30 ◼   ► He deserves it.

00:52:31 ◼   ► He's been using an ancient iPad.

00:52:33 ◼   ► I think he was using an original iPad Air maybe.

00:52:35 ◼   ► It was super old.

00:52:37 ◼   ► No face ID, very slow, very old.

00:52:39 ◼   ► And he uses it all the time.

00:52:41 ◼   ► It's a very important machine for him.

00:52:43 ◼   ► So I'm a big fancy one.

00:52:45 ◼   ► But then I went to General About and it said, "iPad Pro 4th Generation."

00:52:51 ◼   ► I'm like, "Is that the M2?

00:52:54 ◼   ► Or did I change my mind and say the M2 is too expensive and get him a cheaper one?"

00:52:58 ◼   ► So I do what I normally do and type in a Google query for whatever, and I usually go to the

00:53:03 ◼   ► Wikipedia page because it's usually straightforward and I know the sidebar is going to have the

00:53:07 ◼   ► name of the SOC in it, and there's some basic information.

00:53:11 ◼   ► It's kind of like why I don't go to IMDB anymore, I just go to Wikipedia, because I want to

00:53:14 ◼   ► just know what's the cast of this thing.

00:53:16 ◼   ► I can find it on the Wikipedia page and IMDB is just a cluster and I can't find anything

00:53:20 ◼   ► on that friggin' site anymore and I hate it.

00:53:22 ◼   ► So bad.

00:53:23 ◼   ► So I go to Wikipedia and I go to iPad Pro 4th generation Wikipedia page and I scroll

00:53:28 ◼   ► down the thing it says system on a chip a12z bionic did I get him something with

00:53:34 ◼   ► an a12z because I would originally looking for a you a refurb m1 iPad Pro

00:53:38 ◼   ► could not find one Apple didn't have them you could find them used at other

00:53:41 ◼   ► places but I wanted to get an app official Apple refurb and Apple was not

00:53:44 ◼   ► selling refurb m1 iPad pros for whatever reason they just didn't have them all

00:53:49 ◼   ► right and so I thought I had decided to get an mm - it's I'm holding the device

00:53:53 ◼   ► in my hand I'm like how annoying is it that I can't tell if this thing has an

00:53:56 ◼   ► M2 in it. Just like, because Apple's thing just says fourth generation and when I

00:54:00 ◼   ► google fourth generation it says A12Z. So I downloaded Geekbench. I'm like look,

00:54:04 ◼   ► I know what the Geekbench numbers should be for an M2 and it's gonna be very

00:54:09 ◼   ► different from the A12Z. First of all Geekbench immediately says in the

00:54:11 ◼   ► sidebar this has an M2. So I'm like, I got the right one. But there was like 20

00:54:16 ◼   ► minutes where I was like did I buy the wrong thing from, I was looking at the

00:54:19 ◼   ► orders like maybe I bought the right thing because the amount looked like an

00:54:22 ◼   ► an M2 amount, it looked like an M2 iPad Pro amount,

00:54:25 ◼   ► but like maybe they charged me for the M2

00:54:27 ◼   ► and shipped me a different one.

00:54:28 ◼   ► So this is just a PSA to let everybody know.

00:54:31 ◼   ► On the Wikipedia pages for iPad Pro,

00:54:33 ◼   ► they have a list of generations

00:54:34 ◼   ► that go up to sixth generation.

00:54:36 ◼   ► The iPad Pro sixth generation on the Wikipedia page

00:54:41 ◼   ► is what Apple calls the iPad Pro 11 inch fourth generation.

00:54:45 ◼   ► So do not be confused,

00:54:46 ◼   ► and I think Wikipedia should really change this.

00:54:48 ◼   ► - 'Cause it's the fourth generation of the 11 inch.

00:54:51 ◼   ► - Exactly right.

00:54:52 ◼   ► And they're called like the main articles,

00:54:55 ◼   ► iPad Pro parentheses fourth generation,

00:54:57 ◼   ► which is exactly what you will see in general

00:54:59 ◼   ► about on your iPad, but it is not right.

00:55:02 ◼   ► So anyway, when in doubt, get Geekbench.

00:55:04 ◼   ► And by the way, I actually ran Geekbench

00:55:05 ◼   ► to make sure the number that came out was the M2 number.

00:55:07 ◼   ► So he has an M2 iPad Pro.

00:55:09 ◼   ► Oh, and the other thing is when I was doubting him,

00:55:11 ◼   ► I'm like, oh, this is an easy way to tell

00:55:12 ◼   ► 'cause I got him an Apple Pencil with it

00:55:13 ◼   ► 'cause he does a lot of drawing on his iPad.

00:55:15 ◼   ► I tried to do the hover thing.

00:55:17 ◼   ► I'm like, well, I have Procreate, let me launch Procreate.

00:55:19 ◼   ► I know Procreate has the hover

00:55:20 ◼   ► and nothing would work on the hover.

00:55:22 ◼   ► And so I looked at a YouTube video

00:55:23 ◼   ► and I saw people hovering over the icons on the home screen

00:55:27 ◼   ► and they would make, you know, with the pencil

00:55:28 ◼   ► and they would make the icons bigger

00:55:29 ◼   ► and that didn't work either.

00:55:30 ◼   ► And I'm like, is there something I have to enable for hover

00:55:33 ◼   ► and is there some preference of hover in Procreate

00:55:35 ◼   ► but why doesn't it work on the home screen?

00:55:37 ◼   ► This is why I was like, I think I got the wrong iPad.

00:55:39 ◼   ► It doesn't hover, it can't possibly be an M2.

00:55:41 ◼   ► Did I get the wrong pencil?

00:55:42 ◼   ► I googled for like, is there a second,

00:55:45 ◼   ► second generation Apple pencil that does the hover?

00:55:47 ◼   ► I'm like, no, it's all the ones with the flat side.

00:55:50 ◼   ► he's got the one with the flat side,

00:55:51 ◼   ► it magnetically connects, this should be doing hover.

00:55:53 ◼   ► The answer to that was, you gotta update to 16.2.

00:55:56 ◼   ► - Oh, gosh.

00:55:57 ◼   ► - It should, oh, it shipped with the wrong OS.

00:55:59 ◼   ► So it was a confusing morning where I was like,

00:56:01 ◼   ► what the hell did I buy, and don't put it past me,

00:56:04 ◼   ► I've bought the wrong thing before.

00:56:06 ◼   ► When I ordered his MacBook Air, I ordered the wrong color

00:56:08 ◼   ► and I had to cancel it and did a big thing,

00:56:09 ◼   ► or at least I caught that before it shipped,

00:56:11 ◼   ► but lots of confusion on Christmas morning,

00:56:13 ◼   ► but he does indeed have an M2 iPad Air,

00:56:16 ◼   ► which he very richly deserves, with an Apple pencil on.

00:56:18 ◼   ► Once updated to 16.2, Hover started working.

00:56:21 ◼   ► - All right, so that brings up something

00:56:22 ◼   ► I've been wondering about for a long time,

00:56:24 ◼   ► and I keep forgetting to bring it up.

00:56:26 ◼   ► So I got my first and only set of AirPods Pro

00:56:28 ◼   ► a year ago for Christmas.

00:56:29 ◼   ► I love them, I do want the new ones,

00:56:31 ◼   ► but I'm too cheap because the ones I've got

00:56:32 ◼   ► work really well, the battery's fine,

00:56:34 ◼   ► no rattling, et cetera, et cetera.

00:56:35 ◼   ► The thing that's driving me batty

00:56:37 ◼   ► is if you go into Find My, and you go into the Devices tab,

00:56:40 ◼   ► it shows my AirPods Pro, and it shows them

00:56:44 ◼   ► as outside of the case, which I'm looking at them.

00:56:46 ◼   ► I'm holding them in my hand right now.

00:56:48 ◼   ► I can assure you they are not outside the case.

00:56:49 ◼   ► And it shows them last having been seen

00:56:52 ◼   ► on September 27th, 2022.

00:56:55 ◼   ► And I vaguely remember, because again,

00:56:58 ◼   ► my memory is garbage, that when I think it was 16,

00:57:01 ◼   ► I was 16 first came out, or I don't,

00:57:03 ◼   ► or maybe it was a late beta, I forget exactly what it was.

00:57:05 ◼   ► But my AirPods Pro could do the homing beacon,

00:57:10 ◼   ► W1 or whatever it is where you wave your phone around

00:57:14 ◼   ► and it shows you where the AirPods are, et cetera, et cetera.

00:57:16 ◼   ► And I think that that only worked for a minute,

00:57:19 ◼   ► and then they pulled that for the original AirPods Pro,

00:57:22 ◼   ► and now it only works on the AirPods Pro 2.

00:57:24 ◼   ► But the stupid Find My app is stuck on September 27th,

00:57:28 ◼   ► and it's like it never got a clue that this feature,

00:57:31 ◼   ► the homing beacon feature, whatever it's called,

00:57:33 ◼   ► doesn't work.

00:57:34 ◼   ► And so according to this, my left AirPod,

00:57:37 ◼   ► or my left AirPod, that's a great movie,

00:57:39 ◼   ► my left AirPod Pro was last seen September 27th, 2022,

00:57:43 ◼   ► at 2.38 p.m.

00:57:45 ◼   ► And it is now, well, according to Marco, it's December 24th,

00:57:47 ◼   ► but in most of the world, it's December 26th now.

00:57:50 ◼   ► What do I do to fix this?

00:57:51 ◼   ► Like, I don't wanna unpair the things,

00:57:52 ◼   ► which I guess I could.

00:57:54 ◼   ► I really do not wanna sign out of iCloud,

00:57:56 ◼   ► 'cause that's a nightmare.

00:57:57 ◼   ► - No.

00:57:58 ◼   ► - Is there a thing I can do to fix this?

00:57:59 ◼   ► I'm asking you two, and I'm asking the audience.

00:58:01 ◼   ► Like, what do I do about that?

00:58:04 ◼   ► - Hmm, I don't know any answer to that.

00:58:06 ◼   ► - Right? (laughs)

00:58:07 ◼   ► - I would unpair, like, 'cause unpairing

00:58:08 ◼   ► is not that big a deal,

00:58:09 ◼   ► and it's definitely the first thing you try.

00:58:11 ◼   ► Like, restart it, do it over, unpair.

00:58:13 ◼   ► Yeah, that's the right move.

00:58:15 ◼   ► I'll take that as a homework assignment for myself.

00:58:18 ◼   ► People in the chat are saying, oh, I'm

00:58:19 ◼   ► seeing some of this as well.

00:58:20 ◼   ► So I'm sorry if I've just made you aware of this.

00:58:22 ◼   ► But I don't know.

00:58:23 ◼   ► It was really cool for the 10 minutes

00:58:25 ◼   ► that it worked with the little homing beacon thing.

00:58:27 ◼   ► And I really liked that.

00:58:28 ◼   ► But then Apple took it away because they're meanies.

00:58:30 ◼   ► And then people are saying, when you unpair it,

00:58:32 ◼   ► then remove it from Find My when it's unpaired.

00:58:35 ◼   ► And then start fresh.

00:58:37 ◼   ► Fair enough.

00:58:38 ◼   ► I'll take that as homework assignment.

00:58:39 ◼   ► And do what Margo said.

00:58:40 ◼   ► Then you have to reset all your preferences on all

00:58:42 ◼   ► your devices to--

00:58:43 ◼   ► Well, if you want auto pairing,

00:58:44 ◼   ► you have to do the defaults are in your favor there.

00:58:46 ◼   ► - Yeah, yeah.

00:58:47 ◼   ► All right, iOS 16.2 came out a little while ago,

00:58:49 ◼   ► but it has a feature that we wanted to call

00:58:51 ◼   ► to everyone's attention.

00:58:52 ◼   ► And it was called to our attention by Brandon Butch,

00:58:54 ◼   ► who writes, "You can now stop your wallpaper

00:58:56 ◼   ► and/or notifications from appearing on the iPhones

00:58:58 ◼   ► always on display."

00:59:00 ◼   ► This was as of Beta 3, which obviously now it's true

00:59:03 ◼   ► for the shipping version.

00:59:05 ◼   ► But yeah, there's a switch in all,

00:59:06 ◼   ► or there's several switches and always on display.

00:59:07 ◼   ► There's just on/off.

00:59:09 ◼   ► Then there's show wallpaper and show notifications.

00:59:12 ◼   ► So you can turn the wallpaper on or off

00:59:15 ◼   ► and the notifications on or off individually.

00:59:17 ◼   ► - Yeah, so remember when the 14,

00:59:20 ◼   ► first of all, right before the 14 Pro came out

00:59:23 ◼   ► and we were speculating about the rumors

00:59:24 ◼   ► about the always-on display,

00:59:26 ◼   ► I had said, wouldn't it be cool if they basically

00:59:29 ◼   ► had this whole new aesthetic on the display

00:59:32 ◼   ► where most of the screen is black

00:59:34 ◼   ► and you have maybe just monochrome icons or widgets

00:59:39 ◼   ► and the clock and have those things glow

00:59:41 ◼   ► and have the rest of the screen be black,

00:59:42 ◼   ► which people quickly pointed out that that's basically

00:59:44 ◼   ► how Android phones with always on displays have always been.

00:59:47 ◼   ► And that this was not a new idea in the Android world.

00:59:50 ◼   ► So thank you for that.

00:59:52 ◼   ► They all probably used Opera as well.

00:59:53 ◼   ► So when the 14 came out though,

00:59:57 ◼   ► the only option you really had through iOS was

01:00:01 ◼   ► just take whatever your home screen is,

01:00:03 ◼   ► just make a dim version.

01:00:04 ◼   ► That's it, like that's what it did.

01:00:05 ◼   ► And there was no customization.

01:00:07 ◼   ► And this led to I think a couple of shortcomings.

01:00:11 ◼   ► one of the biggest ones being obviously

01:00:12 ◼   ► it wasn't very power conservative.

01:00:14 ◼   ► It wasn't as power conservative as it could be

01:00:17 ◼   ► if most of the screen was black

01:00:18 ◼   ► and only some pixels were being lit

01:00:20 ◼   ► if you had any kind of colored wallpaper.

01:00:22 ◼   ► But also it became harder to tell

01:00:25 ◼   ► whether your phone was awake or asleep.

01:00:28 ◼   ► And that was a huge, I think stumbling point

01:00:31 ◼   ► for people when they first got the 14 Pro

01:00:34 ◼   ► 'cause iPhone users were never accustomed to having

01:00:37 ◼   ► always on anything.

01:00:38 ◼   ► And so you'd see your phone kind of like out of the corner

01:00:41 ◼   ► of your eye and you kind of freak out, oh, it's awake.

01:00:43 ◼   ► Like you'd think there was notification on it or something

01:00:44 ◼   ► or you think it was ringing or whatever

01:00:46 ◼   ► because you were conditioned to believe that a phone

01:00:49 ◼   ► with the screen on is awake and trying to tell you something.

01:00:52 ◼   ► And so what they added with 16.2 is an option

01:00:56 ◼   ► to turn off the wallpaper when it's in, you know,

01:01:00 ◼   ► sleeping always on mode.

01:01:02 ◼   ► And you could also optionally turn off notifications

01:01:05 ◼   ► which I believe Do Not Disturb always did

01:01:08 ◼   ► on its lock screen, but it was kind of custom on other ones.

01:01:11 ◼   ► Anyway, the big thing for me was the wallpaper toggle.

01:01:14 ◼   ► So if you turn off Show Wallpaper

01:01:17 ◼   ► under the Always On Display settings,

01:01:18 ◼   ► you now have in sleep mode a black background

01:01:22 ◼   ► that then only shows the clock and your widgets

01:01:26 ◼   ► and whatever notifications you might have there.

01:01:29 ◼   ► And you can turn off the notifications

01:01:30 ◼   ► if you want there as well.

01:01:31 ◼   ► So I've been using this since the 16.2 betas,

01:01:34 ◼   ► which were, it's been a couple of months now, I think,

01:01:38 ◼   ► and it has radically improved my interaction

01:01:41 ◼   ► with the 14 Pro's always on screen.

01:01:43 ◼   ► I can strongly recommend if you are a little bit put off

01:01:47 ◼   ► by like, you know, it not looking too different

01:01:49 ◼   ► or if you keep still thinking that your phone's awake

01:01:51 ◼   ► when it's not, try this option.

01:01:54 ◼   ► Turn off show wallpaper.

01:01:55 ◼   ► And it is, I think it's massively better.

01:01:59 ◼   ► I think it looks cooler and it functions better

01:02:01 ◼   ► in the sense that it allows me to know

01:02:04 ◼   ► my phone is not awake right now.

01:02:06 ◼   ► - I definitely believe you,

01:02:07 ◼   ► but I did turn off the always-on display ages ago,

01:02:10 ◼   ► mostly because, for what you said,

01:02:12 ◼   ► it seems like my phone was always on.

01:02:14 ◼   ► But now that I've gone back to the old way,

01:02:17 ◼   ► I'm not anxious to sacrifice battery life.

01:02:21 ◼   ► Because what I did learn after using it for a while

01:02:24 ◼   ► with the always-on display on

01:02:25 ◼   ► is that I didn't derive a lot of utility from it.

01:02:28 ◼   ► There was the downside, which is like,

01:02:29 ◼   ► "Oh, I think my phone is on."

01:02:30 ◼   ► But setting that aside, setting aside the downside,

01:02:32 ◼   ► Did I find occasions where it was useful for me to have the screen always on?

01:02:36 ◼   ► And I think the answer to that is no.

01:02:38 ◼   ► So even though this is, I agree, a much better way of doing it, at least for my purposes,

01:02:42 ◼   ► so I'd be able to tell that that's the always on display because the background is black,

01:02:47 ◼   ► I don't think I get any value from it the way I use my phone, so I'm not going to sacrifice

01:02:53 ◼   ► the battery life for this.

01:02:55 ◼   ► But I agree this is a great feature and it might even be useful to be the default, but

01:02:59 ◼   ► that could just be old habits, right?

01:03:02 ◼   ► Maybe new users who their first iPhone has

01:03:04 ◼   ► the always on screen will just be used to it

01:03:05 ◼   ► and not develop the habits that we all have

01:03:07 ◼   ► to expect a phone with our wallpaper being displayed

01:03:10 ◼   ► is on or trying to tell us something.

01:03:12 ◼   ► And I think related to wallpapers,

01:03:13 ◼   ► didn't 16.2 also add the thing where they separated

01:03:17 ◼   ► the screen that appears under your icons on springboard

01:03:21 ◼   ► from the lock screen?

01:03:23 ◼   ► Like it used to be whenever you made a custom one,

01:03:24 ◼   ► it made you make both of them as a match set, as a pair.

01:03:29 ◼   ► and I think they split that out, is that correct?

01:03:31 ◼   ► - I thought that was the case.

01:03:33 ◼   ► I haven't tried it either, but I thought that was the case,

01:03:35 ◼   ► and it drove me nuts.

01:03:36 ◼   ► It drove me nuts in the original version 16.

01:03:39 ◼   ► - And they should be split out,

01:03:40 ◼   ► because I, especially when I was first messing with it,

01:03:43 ◼   ► I wanted to make lots of different fun lock screens.

01:03:46 ◼   ► I did not want to make lots of different thing

01:03:48 ◼   ► that goes behind my icons,

01:03:50 ◼   ► 'cause the thing behind my icons is 100% black

01:03:52 ◼   ► and always has been, and that doesn't change for me,

01:03:55 ◼   ► and so every time I had to,

01:03:57 ◼   ► wanted to make a new lock screen, it's like,

01:03:58 ◼   ► "Oh, and don't forget, you also need to make a..."

01:04:00 ◼   ► And I go, "Ugh, just separate these,

01:04:01 ◼   ► "it's two different things."

01:04:03 ◼   ► So if you wanna make a million different,

01:04:05 ◼   ► I guess they would call, what would you call it?

01:04:06 ◼   ► You would call it the, well, they call wallpaper

01:04:09 ◼   ► the image that's behind stuff,

01:04:10 ◼   ► but I think there should be a different name

01:04:12 ◼   ► for the image that appears behind your icons on springboard,

01:04:15 ◼   ► and then the image that appears on the lock screen.

01:04:17 ◼   ► And they call that image wallpaper in this setting,

01:04:22 ◼   ► because the option is, on the always on display,

01:04:25 ◼   ► do you want to show your wallpaper?

01:04:27 ◼   ► they mean do you want to show the image that shows behind the icons on your, on Springboard.

01:04:32 ◼   ► Anyway, confusing nomenclature but I'm glad they're working this feature out to be more

01:04:36 ◼   ► flexible.

01:04:37 ◼   ► Yeah, because one issue I still have with the Always On Display is that when it's in

01:04:42 ◼   ► sleep mode, whatever is on screen is not tappable.

01:04:46 ◼   ► It looks like it's tappable, but it's not.

01:04:48 ◼   ► The first tap that you do will put it into awake screen mode and then you have to tap

01:04:52 ◼   ► again or swipe or whatever you're doing to interact with something on that screen.

01:04:57 ◼   ► And so having, like part of the reason why I like this option of turning off show wallpaper

01:05:02 ◼   ► is that it makes it more clear to me, not only is the phone not awake, but it is not

01:05:05 ◼   ► interactive right now.

01:05:07 ◼   ► Like if I want this thing to be interactive, I have to tap it so that it is no longer mostly

01:05:11 ◼   ► black.

01:05:12 ◼   ► And then I know, oh it's lighting up with my, I still like the weather background, it's

01:05:16 ◼   ► lighting up with the rain that's outside, great, now I can interact with the notification

01:05:20 ◼   ► or whatever.

01:05:21 ◼   ► And that's, I hope, I really, really hope,

01:05:25 ◼   ► in future hardware iterations,

01:05:26 ◼   ► I hope they are at some point able

01:05:28 ◼   ► to make the sleep state more interactive.

01:05:31 ◼   ► I know there's a lot of challenges with that

01:05:33 ◼   ► and a lot of kind of practical concerns of like,

01:05:36 ◼   ► well, what if it's in your pocket

01:05:37 ◼   ► or what if you're handling it, you know, whatever it is.

01:05:39 ◼   ► And I think there are certain risks

01:05:41 ◼   ► and maybe it's not possible to do that well,

01:05:44 ◼   ► but I would love for them to at least try to do that well

01:05:46 ◼   ► because, or at least make it more obvious

01:05:49 ◼   ► that something's not interactive in the design.

01:05:52 ◼   ► And then ideally also, hopefully, make it wake up faster.

01:05:56 ◼   ► That would also help a lot.

01:05:58 ◼   ► Like if it can go from sleep state to awake state

01:06:01 ◼   ► in half the time it does now,

01:06:04 ◼   ► I'm sure some of that is a choice

01:06:06 ◼   ► in terms of animation speed,

01:06:07 ◼   ► but also I think some of that is giving the OS time

01:06:10 ◼   ► to wake up and using the animation

01:06:12 ◼   ► as kind of cover for that time.

01:06:14 ◼   ► So whatever they can do to make it more interactive

01:06:17 ◼   ► when you want to interact with it,

01:06:18 ◼   ► that would feel very good.

01:06:20 ◼   ► And right now, it still feels a little bit clunky

01:06:22 ◼   ► in that way.

01:06:23 ◼   ► - They could do kind of,

01:06:24 ◼   ► the only safe way I think they should try doing this

01:06:26 ◼   ► is kind of like they do for continuity camera,

01:06:28 ◼   ► which is the thing where you can take your phone

01:06:30 ◼   ► and use it as like the quote unquote webcam for your laptop.

01:06:34 ◼   ► And continuity camera works by detecting

01:06:36 ◼   ► that you have taken your camera

01:06:38 ◼   ► and put it in a particular orientation

01:06:40 ◼   ► that is more or less stationary

01:06:41 ◼   ► when you put it on those little clips or stands

01:06:43 ◼   ► to be behind your display, right?

01:06:45 ◼   ► So the feature you want,

01:06:47 ◼   ► It would be extremely, it would have lots of accidental input if it was just like that

01:06:51 ◼   ► all the time.

01:06:52 ◼   ► Oh, the lock screen, but it's interactive because people, like I said, people would

01:06:54 ◼   ► put it in their pocket, it would activate.

01:06:56 ◼   ► My parents are constantly accidentally calling me because, just because the delay between

01:07:00 ◼   ► when they lock their phone and when they put it into their pocket is too long and they

01:07:06 ◼   ► end up like touching something on their phone when they put it into their pocket and they

01:07:08 ◼   ► butt dial me, right?

01:07:10 ◼   ► But it could detect, for example, when it is laying on its back on a table for a certain

01:07:15 ◼   ► period of time and say, now I've detected that I'm on a table,

01:07:19 ◼   ► and I am going to actually register your taps, whatever

01:07:23 ◼   ► you tap on.

01:07:23 ◼   ► But as soon as you pick it up and start moving,

01:07:25 ◼   ► it's like, nope, nope, I'm locked.

01:07:26 ◼   ► No input means anything to me, right?

01:07:28 ◼   ► So that you-- it's tricky to do, right?

01:07:30 ◼   ► But I think if you just make it active all the time,

01:07:32 ◼   ► people are going to dial things like crazy.

01:07:34 ◼   ► Like it's just-- it's impossible to have.

01:07:37 ◼   ► Because again, when you lock your phone

01:07:39 ◼   ► and you put it into any of your pockets,

01:07:42 ◼   ► the expectation is there's nothing

01:07:43 ◼   ► can happen in that pocket that's going to do stuff on my phone, right? Even if I accidentally

01:07:48 ◼   ► hit the power button, which is a physical button for now anyway, it's still not going to unlock

01:07:53 ◼   ► my phone because Face ID will fail, Touch ID will fail, it's inside my pocket, right? Whereas if that

01:07:57 ◼   ► tap would actually activate the notification, as we know from stupid iOS, once you look at

01:08:01 ◼   ► a notification it's gone forever and you can never see it again and I hate that so much. I wish there

01:08:06 ◼   ► was like, here are the last 100 notifications that you dismissed, just show me them, keep them around.

01:08:11 ◼   ► So anyway, tricky to add feature, but I get where you're coming from, and I do think the

01:08:16 ◼   ► 16.2 things are all improvements.

01:08:19 ◼   ► I actually left my Always On Display in all the default settings.

01:08:23 ◼   ► I like it.

01:08:25 ◼   ► I wouldn't say it's been an earth-shattering change for me.

01:08:28 ◼   ► I will say, kind of tangentially related, that battery life, when I first got the new

01:08:32 ◼   ► phone and it was chugging on photos and all that and all the machine learning stuff, like

01:08:36 ◼   ► the battery life was garbage as expected.

01:08:38 ◼   ► Then it got really really good for a while and then I don't know if it was 16.2 or something

01:08:43 ◼   ► But sometime in last month or two both errands and my phones

01:08:46 ◼   ► Our battery life has gone through the crapper recently and I haven't dug much to figure out what's going on there

01:08:52 ◼   ► There was nothing obvious when I went, you know spelunking in the battery like history and all that in settings

01:08:58 ◼   ► But it's it struck me as a little odd that both of us were having similar problems

01:09:02 ◼   ► So, I don't know if that's just a list family thing or if that's broader or whatever

01:09:06 ◼   ► But it's been a little bit of a bit of a bummer and so I might be turning off the always-on display if it continues

01:09:12 ◼   ► To be a problem, but all told I do like it

01:09:14 ◼   ► I have it rotating between pictures of the kids and pictures of Erin

01:09:17 ◼   ► And I like that every time I pick up my phone or well every time it gets locked

01:09:22 ◼   ► There's a new picture there and then the next time I lock the phone

01:09:25 ◼   ► There's a different picture and I can see it dimly on the display while it while it's just sitting and I dig that I think

01:09:30 ◼   ► That's fun cute

01:09:31 ◼   ► But I totally understand if it's not for everyone and by the way even without the always-on display like part of my problem of being

01:09:37 ◼   ► Distracted by it is that I would think like someone is calling because there I see their face on my phone

01:09:43 ◼   ► But I also have a rotation of my family right so every time I pick up my phone

01:09:46 ◼   ► It's the picture of someone in my family, but if my phone starts ringing

01:09:50 ◼   ► And it's like a spam call from just you know random spam number telling you my car warranty repair or whatever

01:09:56 ◼   ► But my phone screen will light up and I'll see a picture of my son's face on it

01:10:01 ◼   ► I'm like, oh I'm getting a call from my son. But yeah, I mean like that was the problem with the always on display

01:10:06 ◼   ► I always thought that was being contacted by somebody even without the always on display

01:10:09 ◼   ► that's still that's just a side effect of putting people's face as the you know, as your

01:10:14 ◼   ► Wallpaper or lock screen or whatever because you have to actually then say ignore the face

01:10:18 ◼   ► Look up at that

01:10:20 ◼   ► You know at the little caller ID thing that shows when you're getting a call to see what number it's coming from

01:10:24 ◼   ► from.

01:10:25 ◼   ► And if my son does call, I think it still shows his face, but it's a different picture,

01:10:29 ◼   ► but it's still him.

01:10:30 ◼   ► So anyway, possible confusion.

01:10:32 ◼   ► I mean, I guess I could just do landscapes or, you know, this is the first, whatever

01:10:35 ◼   ► it is, the iPhone 14 Pro is the first phone and iOS 16 that I did not have a picture of

01:10:42 ◼   ► one of my past dogs as my lock screen.

01:10:44 ◼   ► That has been my lock screen since the iPod touch in 2007, right?

01:10:48 ◼   ► I remember this.

01:10:49 ◼   ► It was Huckleberry or whatever, right?

01:10:51 ◼   ► Yep.

01:10:52 ◼   ► I decided, because I wanted to try the feature, I decided to try the rotating thing.

01:10:55 ◼   ► I had a bunch of nice pictures of my family that I made very tall and put room above their

01:11:00 ◼   ► heads for the clock so the clock wasn't over their face.

01:11:02 ◼   ► And I do like that.

01:11:03 ◼   ► I do like seeing those pictures rotate, but it continues to be a little bit confusing

01:11:06 ◼   ► when I get a call.

01:11:08 ◼   ► So a week or two ago, I think something like that, it was sometime in December, we got

01:11:13 ◼   ► word through Bloomberg, and I presume it was Germin, that the Apple Car project has "scaled

01:11:18 ◼   ► back" and is delayed. And it won't feature full self-driving capabilities. Imagine that.

01:11:23 ◼   ► But yeah, this was a news report and some highlights that were reported, re-reported on

01:11:29 ◼   ► MacRumors. Apple Inc. has scaled back ambitious self-driving plans for its future electric vehicle

01:11:33 ◼   ► and postponed the car's target launch date about a year to 2026. The car project has been in limbo

01:11:39 ◼   ► for the past several months as Apple executives grappled with the reality that its vision for a

01:11:42 ◼   ► fully autonomous vehicle without a steering wheel or pedals isn't feasible with current technology.

01:11:47 ◼   ► You don't say.

01:11:48 ◼   ► Imagine that.

01:11:49 ◼   ► The car will have an Apple-designed custom processor

01:11:51 ◼   ► to power AI functionality.

01:11:52 ◼   ► The chip is equivalent to four of the highest-end Mac chips

01:11:55 ◼   ► and is nearly production-ready.

01:11:57 ◼   ► That actually kind of brings us back to our discussion

01:11:59 ◼   ► about the Mac Pro last week,

01:12:00 ◼   ► but that's neither here nor there.

01:12:01 ◼   ► - That's a nip, though.

01:12:02 ◼   ► - Well, I mean, I would think.

01:12:04 ◼   ► Apple will use the cloud for some AI processing,

01:12:06 ◼   ► and the company is considering a remote command center

01:12:09 ◼   ► that could assist drivers and control the cars

01:12:10 ◼   ► from afar during emergencies.

01:12:12 ◼   ► That sounds safe.

01:12:14 ◼   ► I don't know if we really need to go

01:12:16 ◼   ► go into the ins and outs of the particulars of this,

01:12:18 ◼   ► but it is interesting that this poop show

01:12:21 ◼   ► is still a poop show from what it sounds.

01:12:25 ◼   ► - Why are they still doing any of this?

01:12:27 ◼   ► That's my, you know, this project has gone through

01:12:30 ◼   ► so many phases and ideas and certainly people

01:12:35 ◼   ► and certainly money, oh my God.

01:12:38 ◼   ► Like how long do they have to keep fumbling around

01:12:42 ◼   ► with this massive money waster before they're like,

01:12:45 ◼   ► you know what, this isn't for us.

01:12:46 ◼   ► Let's just crap can it and move on.

01:12:49 ◼   ► I still don't see why they even want to be in this business,

01:12:54 ◼   ► let alone why they keep plowing forward

01:12:58 ◼   ► without seemingly able to go near any kind of outcome.

01:13:01 ◼   ► - The story's interesting because it's kind of like,

01:13:06 ◼   ► the best line of the story is where it says that,

01:13:10 ◼   ► grappling with their vision, that blah, blah, blah,

01:13:12 ◼   ► isn't feasible with current technologies.

01:13:14 ◼   ► What it's saying is they don't have it.

01:13:16 ◼   ► They would like to have a product that does X,

01:13:18 ◼   ► but they don't.

01:13:19 ◼   ► It's like, I would like something,

01:13:20 ◼   ► I would like something that lets me levitate.

01:13:23 ◼   ► That would be a cool product, but we don't have that.

01:13:27 ◼   ► So how much money are you gonna spend on a project?

01:13:29 ◼   ► So is this R&D?

01:13:31 ◼   ► Like that's kind of the problem with self-driving.

01:13:33 ◼   ► It's difficult to do the R&D to figure out

01:13:36 ◼   ► if you can make a product without spending a lot of money.

01:13:39 ◼   ► And I get that.

01:13:40 ◼   ► It's not as simple as like,

01:13:40 ◼   ► oh, we'll have these people tooling around

01:13:43 ◼   ► with a multi-touch, right?

01:13:44 ◼   ► And that'll be a small team of a handful of people

01:13:49 ◼   ► and they'll tool away at it for years.

01:13:50 ◼   ► And eventually, if we get to the point where,

01:13:52 ◼   ► no, we could probably make a product out of this.

01:13:53 ◼   ► Maybe we'll make a tablet, maybe we'll make a phone.

01:13:55 ◼   ► Then you form the iPhone team and then the iPad team, right?

01:13:59 ◼   ► And that's when you staff up

01:14:00 ◼   ► and that's when you spend the money.

01:14:01 ◼   ► Although the amount of money that was used

01:14:03 ◼   ► to develop the original iPhone is comically low

01:14:05 ◼   ► and every company should feel bad

01:14:07 ◼   ► when they read those histories of like,

01:14:08 ◼   ► so how much money did Apple spend?

01:14:10 ◼   ► Even if we count like everything we can think of,

01:14:11 ◼   ► How much do they spend to make the iPhone?

01:14:14 ◼   ► The small number I think is they spend $150 million,

01:14:16 ◼   ► which is like, we spent $150 million

01:14:18 ◼   ► to change the color of the logo, right?

01:14:20 ◼   ► That's what big companies normally do.

01:14:22 ◼   ► So they spent $150 million to make the iPhone,

01:14:24 ◼   ► how much money is it made?

01:14:26 ◼   ► Anyway, there's more to it if you add to that, right?

01:14:28 ◼   ► But for self-driving cars, it's like, okay,

01:14:31 ◼   ► I can understand why they're saying,

01:14:32 ◼   ► it would be cool if we had a self-driving car.

01:14:34 ◼   ► I agree, that would be cool, right?

01:14:37 ◼   ► Can we make one?

01:14:38 ◼   ► Let's do some R&D to find out.

01:14:40 ◼   ► But it has always seemed to me

01:14:41 ◼   ► that this project at Apple was created

01:14:43 ◼   ► as if they had already done that.

01:14:44 ◼   ► It's like, oh, we know how to make a self-driving car,

01:14:46 ◼   ► let's just make a project to make one.

01:14:47 ◼   ► Like that it's already in the phase of like,

01:14:49 ◼   ► we're designing a product and we're staffing a team

01:14:51 ◼   ► to build that product and we're gonna contract manufacturers

01:14:54 ◼   ► to manufacture that product.

01:14:55 ◼   ► It's like, whoa, whoa, whoa,

01:14:56 ◼   ► what product are you making?

01:14:57 ◼   ► Our self-driving car.

01:14:59 ◼   ► Do you have a self-driving car?

01:15:00 ◼   ► He's like, no, but we'll probably be able

01:15:02 ◼   ► to figure it out, right?

01:15:03 ◼   ► Nope. (laughs)

01:15:05 ◼   ► Like you gotta be able to make the thing first

01:15:08 ◼   ► before you make a product team.

01:15:10 ◼   ► So again, no one knows the details of this,

01:15:12 ◼   ► but to the extent that anything with product Titan

01:15:15 ◼   ► has been in like let's make a product mode,

01:15:18 ◼   ► that has been a bad idea if the premise of the product

01:15:21 ◼   ► is a car without a steering wheel or pedals.

01:15:23 ◼   ► You cannot make a car, let's go with that saying,

01:15:25 ◼   ► you cannot make a car with a steering wheel or pedals

01:15:27 ◼   ► until it can drive itself.

01:15:29 ◼   ► And if it can't drive itself,

01:15:30 ◼   ► maybe get out of the product phase

01:15:32 ◼   ► and back into the R&D phase.

01:15:34 ◼   ► Don't go to the product phase until you think

01:15:36 ◼   ► you have something that works like the multi-touch.

01:15:38 ◼   ► We've got a multi-touch screen that we think works.

01:15:41 ◼   ► Maybe it's too big, maybe it's too clunky,

01:15:43 ◼   ► maybe it takes a lot of power,

01:15:44 ◼   ► maybe it's connected to like a power Mac or something,

01:15:46 ◼   ► but it works.

01:15:48 ◼   ► And the question is, can we make it smaller, cheaper,

01:15:51 ◼   ► blah, blah, blah.

01:15:52 ◼   ► They don't have anything that works.

01:15:54 ◼   ► Nobody does.

01:15:55 ◼   ► Nobody has this, right?

01:15:57 ◼   ► It would be cool if you could make it, but make it first.

01:16:00 ◼   ► I think this is just a hilarious story that the premise is,

01:16:03 ◼   ► Apple has decided they can't make a product they can't make.

01:16:05 ◼   ► Like, yeah, you gotta be able to.

01:16:07 ◼   ► And the other thing they can decide is,

01:16:08 ◼   ► we're not gonna make a car without pedals or steering wheel.

01:16:10 ◼   ► Make one with pedals and a steering wheel

01:16:12 ◼   ► that humans can drive.

01:16:13 ◼   ► Lots of people do that.

01:16:14 ◼   ► You can have lots of self-driving assistive functions

01:16:16 ◼   ► that I think are probably a bad idea in many cases,

01:16:18 ◼   ► but that's a thing that you can make.

01:16:20 ◼   ► But the idea that they keep trying to make a car

01:16:22 ◼   ► without a steering wheel, you can't do that.

01:16:25 ◼   ► And so I'm sad for them and I'm sad for the story.

01:16:28 ◼   ► I do like the idea that their AI powered chip

01:16:30 ◼   ► or whatever is, that sounds so much like the Mac chip

01:16:33 ◼   ► that they canceled, equivalent to four

01:16:34 ◼   ► of their high-end processors.

01:16:36 ◼   ► You know, I mean, you can't afford to put it in a $10,000,

01:16:40 ◼   ► you know, Mac Pro, because $10,000 is too much for a Mac Pro.

01:16:42 ◼   ► How much would this car cost?

01:16:44 ◼   ► Again, if it actually existed and worked.

01:16:46 ◼   ► My advice to Apple is if you wanna make a car,

01:16:48 ◼   ► add a steering wheel and pedals.

01:16:50 ◼   ► - Yeah, I just, I still go back and like,

01:16:53 ◼   ► you know, to me, like the question you asked,

01:16:54 ◼   ► like all right, can we make this thing work,

01:16:56 ◼   ► and then, you know, then move on from there.

01:16:58 ◼   ► I would step back again and say,

01:17:00 ◼   ► is this even a business that we need to be in

01:17:02 ◼   ► or want to be in?

01:17:04 ◼   ► And what would it mean if we entered this business

01:17:08 ◼   ► and actually stayed in it and maybe even succeeded in it?

01:17:11 ◼   ► What would that actually mean for the rest of our business?

01:17:15 ◼   ► Can you imagine Apple selling cars?

01:17:18 ◼   ► How would that even physically work?

01:17:20 ◼   ► Can you imagine Apple servicing cars?

01:17:22 ◼   ► Can you imagine Apple going through all the regulations?

01:17:25 ◼   ► Like, I can see that it's possible for them to do all that,

01:17:30 ◼   ► But why would they want to?

01:17:32 ◼   ► And what they would achieve in that

01:17:35 ◼   ► would be they'd be a car manufacturer.

01:17:39 ◼   ► Look at all the car manufacturers.

01:17:40 ◼   ► How happy are they?

01:17:41 ◼   ► How good of a business does that seem to be?

01:17:44 ◼   ► It doesn't seem like it's worth them entering this business

01:17:48 ◼   ► even if it was fairly easy to get into, and it's super not.

01:17:53 ◼   ► And so I just don't see why they would even want

01:17:58 ◼   ► to be in this business.

01:17:59 ◼   ► That even sets aside the question of,

01:18:02 ◼   ► could Apple be good at designing cars,

01:18:06 ◼   ► which I think is a huge question mark, honestly.

01:18:10 ◼   ► Apple's very good at designing a lot of things.

01:18:13 ◼   ► I don't know that I would want an Apple-designed car,

01:18:17 ◼   ► but again, I think that's even,

01:18:20 ◼   ► that's too far down the line of thinking of like,

01:18:22 ◼   ► why are they even doing this?

01:18:23 ◼   ► Why do they wanna be in this business?

01:18:25 ◼   ► That, to me, makes no sense.

01:18:27 ◼   ► I mean, it is different in many ways that we've discussed, but I think it is actually

01:18:31 ◼   ► a better fit with their traditional strengths than some other things.

01:18:37 ◼   ► Because it is manufacturing a product that they sell for a profit.

01:18:42 ◼   ► It is, you know, regulation stuff.

01:18:44 ◼   ► Again, Apple's not used to automotive regulation, but they are used to dealing with the FCC

01:18:48 ◼   ► and the various radio regulations and heat and thermal.

01:18:51 ◼   ► Like there are different laws in different countries and being in compliance with them

01:18:54 ◼   ► with electronic devices is a thing, right?

01:18:56 ◼   ► It is a scaled up version of what they do.

01:18:59 ◼   ► They're good at manufacturing, they don't have their own manufacturing plans, they outsource

01:19:02 ◼   ► that, and they're good at helping those manufacturers do a better job of manufacturing.

01:19:06 ◼   ► Apple is good at actually "making things" even if they're doing through third parties.

01:19:09 ◼   ► Like, I can squint and say, this looks like, again, a question of whether they'd be able

01:19:14 ◼   ► to make a good car, but they make hardware products that are software powered and they

01:19:18 ◼   ► manufacture them and comply with a bunch of laws.

01:19:20 ◼   ► So it is conceivable that they could get good at this after many years because it is like

01:19:25 ◼   ► what they do.

01:19:26 ◼   ► to me of this great quote that Gruber had back on December 21st, I put this in contrast

01:19:31 ◼   ► to something that Apple already does that I think is not a good fit for their traditional

01:19:36 ◼   ► strengths and model.

01:19:37 ◼   ► This is a quote from talking about, what was it, Gruber's not guested on a podcast and

01:19:43 ◼   ► they talked about various Apple related things.

01:19:46 ◼   ► And he put this little summary at the end of his link to the podcast he was a guest

01:19:51 ◼   ► on.

01:19:52 ◼   ► This is quoting Gruber.

01:19:53 ◼   ► The App Store's financial success is the worst thing that's happened to Apple this century.

01:19:57 ◼   ► It's a distraction at best and a profound corruption at worst.

01:20:01 ◼   ► Services revenue and the App Store do not fit with Apple's traditional strengths of

01:20:06 ◼   ► making a really good product and selling it to people for a profit because it's really

01:20:09 ◼   ► good and they like it.

01:20:10 ◼   ► It is a much more complicated model where you're selling access to them, you're rent-seeking,

01:20:16 ◼   ► you're controlling a platform and then charging other people money to get access to the people,

01:20:19 ◼   ► potentially collecting information about them and advertising to them.

01:20:23 ◼   ► That model does not fit with the Apple that we love in a way that the car does.

01:20:27 ◼   ► Because if Apple made a good car product, I believe that they could figure out how to

01:20:32 ◼   ► sell it, how to service it.

01:20:33 ◼   ► In the same way they figure out how to sell their hardware products, their phones and

01:20:37 ◼   ► everything, they figure out how to service them, they figure out how to comply with laws,

01:20:40 ◼   ► they figure out how to get them manufactured, they figure out how to sell them all over

01:20:43 ◼   ► the world with different regulations.

01:20:45 ◼   ► Obviously a car is way different than a phone.

01:20:47 ◼   ► is a much bigger scale thing. There are so many other complications they're working with,

01:20:51 ◼   ► a bunch of other players, but that fits better. The incentives are aligned better for the

01:20:56 ◼   ► Apple that I like, the Apple that makes its money by making a good product. Not the Apple

01:21:01 ◼   ► that is a distraction at best or a corruption at worst, paraphrasing Gruber, that is distracted

01:21:07 ◼   ► by the idea of like, "Yeah, but have we considered rent-seeking? Because that's where the real

01:21:11 ◼   ► money is." That does not align with the incentive. I want them to be incentivized to make good

01:21:17 ◼   ► The incentives that I like is, hey, if we make a bad keyboard,

01:21:19 ◼   ► we have to pay people a class action lawsuit.

01:21:21 ◼   ► And if we make good keyboard, people buy our stuff.

01:21:23 ◼   ► If we put ports back on our laptops,

01:21:24 ◼   ► people buy more of them.

01:21:25 ◼   ► If we make a really fast, low-power computer,

01:21:27 ◼   ► people buy it, right?

01:21:28 ◼   ► If we make an amazing touchscreen phone,

01:21:29 ◼   ► people buy it because it's a cool phone.

01:21:31 ◼   ► That model aligns better with my interest as a customer.

01:21:34 ◼   ► Make cool technology.

01:21:36 ◼   ► Make a profit when I buy it, because it costs you

01:21:39 ◼   ► less to manufacture it than it does for me to buy it from you.

01:21:42 ◼   ► That's your profit.

01:21:43 ◼   ► That's a really good business.

01:21:44 ◼   ► It's a model I understand that I'm more comfortable with.

01:21:47 ◼   ► than the App Store model, which is sell tons of devices that

01:21:51 ◼   ► are awesome and then charge people 30% of all the money

01:21:54 ◼   ► they make on them because you deserve it

01:21:56 ◼   ► because you made the platform.

01:21:58 ◼   ► The thing that-- just to go back a little bit to what Marco's

01:22:00 ◼   ► saying, I think the thing that makes me very worried--

01:22:04 ◼   ► maybe a little dramatic-- but worried about an Apple

01:22:06 ◼   ► car is that I feel like the same problems that everyone

01:22:10 ◼   ► snarks on Tesla about, like, oh, the windshield wipers don't

01:22:13 ◼   ► work because it never rains in the Bay Area, which I know

01:22:15 ◼   ► is not really true.

01:22:16 ◼   ► but just go with it here.

01:22:18 ◼   ► - They don't work that well.

01:22:19 ◼   ► - Well, fair.

01:22:20 ◼   ► And then the door handles would freeze constantly

01:22:23 ◼   ► because it legitimately really rarely does freeze

01:22:25 ◼   ► in the Bay Area.

01:22:26 ◼   ► This very myopic view of the world,

01:22:28 ◼   ► which I feel like is a very common problem

01:22:32 ◼   ► for engineers in the Bay Area,

01:22:34 ◼   ► I wonder if that's going to be an affliction

01:22:37 ◼   ► for an Apple car as well.

01:22:38 ◼   ► And where that would manifest itself is,

01:22:40 ◼   ► oh, surely $150,000 car is something anyone would want,

01:22:43 ◼   ► right?

01:22:44 ◼   ► We charge like 200% of what an average, you know, run-of-the-mill Android phone is for a new iPhone.

01:22:51 ◼   ► People pay 2x for a new iPhone. Surely they'll pay 2 to 3x for an entire car, right?

01:22:57 ◼   ► And no, people will not. Like maybe it's just me,

01:23:01 ◼   ► but I don't have $150,000 just laying around to buy an Apple car, to lease an Apple car at $4,000 or whatever.

01:23:07 ◼   ► I don't know. I don't lease cars, but whatever that lease payment would be.

01:23:09 ◼   ► I don't have that just laying around.

01:23:11 ◼   ► Like, it's wonderful if they think that them

01:23:13 ◼   ► and their cushy worlds with their $3 million,

01:23:15 ◼   ► 2,000 or 200 square foot houses and their, you know,

01:23:19 ◼   ► R8s and their Lamborghini Uruses and all that,

01:23:23 ◼   ► like yeah, $150,000 car may be great in the Bay Area,

01:23:27 ◼   ► but that ain't gonna work in the real world, my friends.

01:23:28 ◼   ► - Doesn't it fit with Apple's model?

01:23:30 ◼   ► Their hardware is always more expensive

01:23:32 ◼   ► than everybody else's.

01:23:32 ◼   ► And I know it's a different matter of scale.

01:23:34 ◼   ► It's like, oh, 20% higher is an extra 200 bucks

01:23:36 ◼   ► versus 20% higher is an extra 20,000 bucks.

01:23:40 ◼   ► But still, I think that it is not impossible

01:23:43 ◼   ► to be the car manufacturer that sells mostly cars

01:23:45 ◼   ► that are too expensive.

01:23:46 ◼   ► Look at Porsche.

01:23:47 ◼   ► - Oh, I agree, I agree.

01:23:48 ◼   ► But I just, I don't know.

01:23:49 ◼   ► I think the thing, and maybe it was Gruber,

01:23:51 ◼   ► maybe it was you, John.

01:23:52 ◼   ► Somebody had once said, like,

01:23:54 ◼   ► the great thing about Apple and about Coke is that if,

01:23:57 ◼   ► okay, so if you want the best Coke in the world,

01:23:59 ◼   ► the best Coca-Cola in the world.

01:24:01 ◼   ► - And he was paraphrasing somebody else.

01:24:03 ◼   ► - Yeah, wasn't this like Andy Warhol or somebody?