00:00:43 ◼ ► And Orion, which is a WebKit based web browser, sort of a peer to Safari for Mac, iPhone, iPad.
00:00:58 ◼ ► All right, now we have to go back to the pronunciation of Kaji because I have been a Mac user for a very long time.
00:01:15 ◼ ► But in the 90s, when the beginning of the 90s, the only way to distribute software effectively, I mean, there was technically an Internet,
00:01:26 ◼ ► but primarily it was all on disks and commercial software was something you would go to a store, buy in a box.
00:01:34 ◼ ► And in the early 90s, again, it was before CDs, you'd get floppy disks and shareware would be distributed by like word of mouth.
00:01:53 ◼ ► There were users groups all over the country, but it was sort of, it was like a meat space thing.
00:02:04 ◼ ► You would show up at the office in the student center where the Mac user group had an office with a couple, as many floppy, blank floppies as I could bring.
00:02:13 ◼ ► And then they had a machine set up where I could select popular recent shareware apps to download to my floppy disk, take back to my dorm and install.
00:02:21 ◼ ► Now, by the middle of the 90s, the Internet had blown up and shareware had gotten more, a step towards what we now consider just normal,
00:02:37 ◼ ► And in the early days of shareware, you'd go to the about box and it would say, you know, you can use and distribute this app, but if you like it, please send $10 to me here.
00:02:47 ◼ ► And then they would give you your postal address. That's, I don't know if you remember, if you were, maybe you were using other computers at the time, but that's how shareware used to work.
00:02:59 ◼ ► But then there was a company and it was one of the first ones at K-A-G-I dot com that would let shareware developers collect money from users.
00:03:28 ◼ ► And it also had some brand recognition with Apple users, exactly, and two reasons you just described.
00:03:36 ◼ ► And to make it funnier, I still to this day receive emails from users who purchased stuff from Kagi like 20 years ago or 10 years ago and something broke and they, they want support or to redownload a piece of software they downloaded like 15 years ago.
00:03:54 ◼ ► And to me, just extraordinary in terms of like timelessness of the platform and just the longevity of the Apple ecosystem from a software standpoint.
00:04:13 ◼ ► So even though you weren't, you had nothing to do, nothing that you or your team at Kagi.com today have to do with that.
00:04:22 ◼ ► The name though, it was a great service and a company that everybody who is around remembers with nothing but fondness.
00:04:30 ◼ ► And I guess, and again, this is the problem I have with the pronunciation and with my history of mispronouncing things that I've only ever read.
00:04:55 ◼ ► But now that I remember, I knew it was a Japanese word and I do not speak Japanese, but I at least kind of know what Japanese sounds like.
00:05:04 ◼ ► And now that I think about it for one second, Kagi sounds much more like a Japanese word than Kaji.
00:05:15 ◼ ► Like what would be the correct pronunciation, but it was interesting to me because the word means key in Japanese and the idea of Kagi was to sort of unlock the internet or this friendly version of internet for people.
00:05:31 ◼ ► So a little bit of, because I'm sure almost everybody, if not everybody who listens to the talk show has at least heard of Kagi.com.
00:05:47 ◼ ► I'll just say this as my primary search engine since September 2022, I believe is when I started because it correlates directly to when I started paying for a subscription.
00:06:01 ◼ ► But I think the reason I've been so looking forward to having you as my guest on the show.
00:06:08 ◼ ► Is that while people have heard of it and they think, oh yeah, it's like one of those small private search engines, a sort of satellite to the mega death star that is Google.
00:06:20 ◼ ► You know, like DuckDuckGo, except Kagi is paid and DuckDuckGo isn't paid and it's filed away in their head.
00:06:30 ◼ ► And it's, I think people have heard of Kagi, but I don't think enough people have really looked at it and seen just how remarkable it is.
00:06:40 ◼ ► And I'd really like to explore that, but before we do, what, what is, how did you get into this?
00:07:01 ◼ ► We have a half a million websites that we out, updates, backup security, things like that.
00:07:07 ◼ ► So I have pretty good exposure to web and web technologies for the previous 15 years or so.
00:07:19 ◼ ► But I have this itch and early 2018, this is where it hit me that Google is getting worse and that the way to words is kind of insulting my intelligence.
00:07:37 ◼ ► I often, one of my axioms for life personally and, and as advice to entrepreneurs, or just what I'm looking for in the companies and products I support is the idea that it's not enough to just have a short list of priorities.
00:08:09 ◼ ► And in the early years of Google search, it was very obvious that their number one priority was delivering the best search results as fast as possible.
00:08:29 ◼ ► And whatever anybody thinks of Google search today, and even in my audience, where I think people are most receptive to alternatives to Google and are sort of technically minded and understand, just even understand what it means to use a non default search engine.
00:08:53 ◼ ► But I think it's also undeniable, even amongst people who still use Google, that their priorities as a company have obviously shifted.
00:09:02 ◼ ► And that might, at some level, they still want to deliver good search results and make people who use them happy and satisfied.
00:09:14 ◼ ► It is very clear to me as a user of Kagi, like I said, full time for over two years now, and somebody who was interested for a while before it, that that is the number one priority at Kagi.
00:09:37 ◼ ► Oh, I think Google's priorities have been the same for the last 20 years, but the the power they got changed.
00:09:53 ◼ ► And the moment Google embraced an ad based business model early 2000s is when this shift started.
00:10:01 ◼ ► And if you remember what the end of 90s looked like when Google showed up, we were using AltaVista and Vycos and Dogpile, Ask James.
00:10:15 ◼ ► And Google really came where the technology that was breakthrough and the quality of results was great.
00:10:24 ◼ ► And even Larry and Sergey, the founders of Google recognized back then that all the user wants is the relevant searches also the query.
00:10:34 ◼ ► And they have criticized very heavily the ad based business model that all these search engines have embraced already.
00:10:41 ◼ ► And they said putting ads in front of the user is a clear conflict of interest and it will lead to the search becoming bad and things like that.
00:10:50 ◼ ► And still three or four years later, that's exactly the business model the company embraces.
00:11:09 ◼ ► Clearly marketing whatnot like the ad blockers didn't even exist, but we were not bothered.
00:11:15 ◼ ► But over the years as you optimize for revenue and from Google, the thing that is clear, the user is not the customer.
00:11:27 ◼ ► So when you optimize for advertisers for 20 years and every company optimizes for the customer, you get into the situation where we are now.
00:11:45 ◼ ► But even I hear from people that find about Kagi and they have to go back to Google on a friend's device to describe this as nightmare.
00:11:57 ◼ ► So yeah, that's what 20 years of pursuing a certain idea about the product and then the customer of the product.
00:12:05 ◼ ► And it really puts restraints to what you can do for the user and the product when the user and customer are not the same.
00:12:22 ◼ ► But if you could look at screenshots of Google search results for just pick like an evergreen query, best backpack.
00:12:32 ◼ ► Something somebody, you know, that somebody would have been a common search 20 years ago and would be a common search today.
00:12:40 ◼ ► And if you looked results every six months from 20 years ago, 40 results, just every six months, a snapshot of what they look like.
00:12:50 ◼ ► It's I'm sure because I used it for most of those years and I still look at it once in a while.
00:12:58 ◼ ► It's the example of that slow boiling frog analogy where there was never any pinpoint where you'd say, oh, it was September 2009 when they did this and that's where Google went bad.
00:13:14 ◼ ► It was eased at the slowest rise of shifting of priorities each step of the way where no ads at all to very clearly marked ads in a box that had a different background color.
00:13:36 ◼ ► So if you said best backpacks, it would be like two ads for backpacks marked sponsored links and here's two backpack brands at the top of your search results.
00:13:55 ◼ ► And you get to heat you get to today and for a lot of searches like that, that above the fold.
00:14:02 ◼ ► In other words, the first screen full of results might all be content from Google and from Google shopping and blah blah blah and you have an old stage floor.
00:14:14 ◼ ► And it's just very and it's so thinking about like when Google first came out in 1999 or 1998 when it when it was like a student project and.
00:14:27 ◼ ► How much better it was than all previous search because it had gotten rid of all the other clutter and it wasn't even ads in the alternatives.
00:14:36 ◼ ► It was just like that the alternatives that the mania at the time was called portals right?
00:14:46 ◼ ► Oh, you'd go to excite.com or whatever and in addition to having a box where you could search the web for other things they were trying to get you to sign up to do your webmail there.
00:14:59 ◼ ► They had their own news articles there they this that the other thing and it was oh let's use a table layout and have six columns of stuff.
00:15:12 ◼ ► You got on their homepage it's and it's such a part of their brand that they have kept this.
00:15:19 ◼ ► Where the google.com homepage is relatively simple but back then it was just a text box and two buttons search and I feel lucky that's it.
00:15:30 ◼ ► It's they've kind of kept that and you have to the thing that people may not remember is that back before google that homepage on yahoo.com or excite.com or askjeeves.com had all sorts of other crap on it.
00:15:45 ◼ ► But now the results page is so different and shows so many different pulled priorities in different directions.
00:15:53 ◼ ► And like you said it and again it can get overuse that analogy of if you're not paying.
00:16:01 ◼ ► You're not the customer you're the product but I think at google search it is it's undeniably true.
00:16:14 ◼ ► Well we're removing all the bad things and what what is making google bad and that is fundamentally the business model.
00:16:21 ◼ ► So basically google turned users into product which is garbage and we are making users customers again.
00:16:30 ◼ ► Which is how it should work and by the way this is not just unique to google most of the tech.
00:16:36 ◼ ► The developers from from 2000 so on where also embrace this model so whether that's social media or browsers for that matter.
00:16:44 ◼ ► We talk about that probably later but all of them fundamentally had this contract with users where it was it was free on the surface.
00:16:55 ◼ ► Like in in monetary terms the users were paying for the media attention and data and the all acceptable contract.
00:17:07 ◼ ► The point is that because of this relentless pursuit of profits it long ago passed this point where you could tolerate it.
00:17:18 ◼ ► Or you should be tolerating it because these are some of the most important products that shape our lives and and thinking and and behavior right.
00:17:28 ◼ ► Browser search social media I'm that's such a big part of our lives and for me I think even in vividly remember the moment when I woke up from the matrix.
00:17:39 ◼ ► Which is how I describe this experience of realizing this was gone to hell and it should be different.
00:17:52 ◼ ► And it showed me the the paid result for target on top and then the target organic result below it and they were completely the same.
00:18:00 ◼ ► And it was this thing like the brands have have to pay for their keyword or somebody else will like a competitor.
00:18:08 ◼ ► Will will buy that and then it's lows or or whatever the first result for target so target has to pay that sale.
00:18:16 ◼ ► But I completely understood like why is Google doing that and while brands have to play by that I just thought I as the user and being served.
00:18:25 ◼ ► Two identical results here like I'm an idiot and this is the moment where I realized that.
00:18:32 ◼ ► As the user my intelligence is being insulted I don't care about their motivation as a company or the advertiser motivation there I just cared about my experience and it was pretty crappy.
00:18:46 ◼ ► So that was the moment when I opened my eyes and said if if if Google didn't exist in 2018 and you put like 100 largest people like engineers, product managers, user, UX people, designers in a room and said design a search engine.
00:19:04 ◼ ► And somebody suggested a search engine look like what I saw it would be those two duplicate results and everything else I think they would throw them out of the room and say you were insane.
00:19:17 ◼ ► And yeah so so that was clear that the fundamental business model just misaligned incentives there and instead of having a direct relationship with your with your search engine.
00:19:29 ◼ ► There was an intermediary between that between you and the search engine and that has to go in order to create a product that's really user centric and search engine is one of those products that we it's so much important for our lives to use it all the time.
00:19:47 ◼ ► It was just insane to me that it had an intermediary and that prevented the innovation in search as well like in 20 years we barely had any innovation happening if you look at Google, Bing or any other search engines basically they haven't changed.
00:20:05 ◼ ► So I thought if we change that equation and remove the intermediary and talk directly to users and have them pay and make that experience insanely good that they they will pay.
00:20:19 ◼ ► And then you can innovate so much in terms of the search experience itself not just the quality of searches out but also add stuff to search as a product and start treating search as a product like you treat any other product.
00:20:40 ◼ ► It's not a God-given right that like we have search like we are born and we have the search for free.
00:20:48 ◼ ► It's not one of those fundamental freedoms in right it's a product made by a private company that's also making billions of dollars of it.
00:21:01 ◼ ► I do think and again I'm going to show my age and I know you're of the age where you remember the pre-internet world vividly and and we're old enough now where there are some people listening who don't really remember the either don't remember it at all because they really were born after during or after the 90s or they were so young that they don't.
00:21:24 ◼ ► But before the internet the only real network we as a society had was the telephone network and we had the phone book and the phone book had white pages which were people and that would and they literally they were called this again I know if you're old enough you're like why are you explaining this John and if you're young enough it's I'm telling you this is how it was half of the book was the white pages or if you lived in a big town you'd get two different books.
00:21:54 ◼ ► One book was the white pages printed on white paper and then the yellow pages were printed on yellow paper and that was commercial business interest it listings and they were just text listings in alphabetical order in the yellow pages.
00:22:09 ◼ ► And you could pay for display ads so if you were selling heating and air conditioning service you could go in the yellow pages to heating and it would say heating and air conditioning and then it would list them all in text.
00:22:27 ◼ ► Small print and because it was listed alphabetically there was game it just like now is like it was like the yellow pages SEO was giving your company a name that started with a and then people would give their company double aid heating and air conditioning.
00:22:44 ◼ ► AAA heating and air conditioning quadruple aid heating and air conditioning or whatever but it was sort of a racket where okay these are just text listings but then there's these big graphic ads where you know just like ads in a magazine or newspaper that also were on those pages.
00:23:03 ◼ ► And you had to pay and you paid the phone company and the phone company made a lot of money from the yellow pages because it was the only way everybody could find anything.
00:23:17 ◼ ► And I think in the same way that you're saying that people don't really think of Google as they really don't as big as Google is they just think it's it's the box at the top of the Internet right like the browser is the Internet.
00:23:30 ◼ ► And there's a box at the type top where you type stuff and you go there but I don't think people really thought of the phone company as a company even though it was and it was a monopoly that was eventually broken up when I was around 10 years old.
00:23:45 ◼ ► But it was like a shakedown racket really if you wanted to be a successful business in your local community you needed to pay for a better ad in the yellow pages because nobody would find your just plain text listing if your competitors had all paid much more money to get a big picture at.
00:24:15 ◼ ► And Google is basically a tax on using the web paid by the least tax savvy. At least the tax savvy use ad blockers and can find their way around it but who's left is the least tax savvy and sort of most vulnerable people and it's also both sad and kind of annoying.
00:24:38 ◼ ► I remember even FDI has published a memo I think in 2022 for the official recommendation of FDI because of so many cases of these ads being to malicious websites that steal user information credit cards whatnot.
00:25:02 ◼ ► So if you think about that if every US citizen did what the FBI asked them to do or recommended to do Google would be out of business.
00:25:18 ◼ ► And just to be clear I'm not against ads. I think you talked about the newspapers. I remember when I was younger in the 90s at least the newspaper where I grew up the ads were sort of in the middle of the paper and I could just skip that portion when I was not interested.
00:25:41 ◼ ► I was looking for a plumber or an apartment or I would go straight to that portion so it had very high sort of usability.
00:25:50 ◼ ► Right and it was not in my face all the time. It was contextual sort of right. You go there when you have intent to go there.
00:25:58 ◼ ► The problem of inserting ads into your information provider is that you're getting served information that you're not looking for whether you like it or not.
00:26:11 ◼ ► That can influence your outcomes in terms of information being consumed but also the entire purpose of ads is of course to change your behavior.
00:26:20 ◼ ► That's what advertising is paying for. So you're constantly influenced by someone's want to change your behavior which is what makes it dangerous in the context of information consumption.
00:26:32 ◼ ► And this is why I believe that title "business model does not belong in something as important as information consumption".
00:26:42 ◼ ► And we haven't as a society been I think very respectful of the value of information that equals you.
00:26:52 ◼ ► Just to give an analogy, like 40 years ago nobody cared about organic food for example.
00:26:58 ◼ ► There were a few people that cared about why I grew up on junk food and not intentionally by the my parents certainly didn't want it but everybody like ate crappy food.
00:27:12 ◼ ► And those who were propagating organic food were like deemed like crazy like why would you pay so much for this when this is just fine.
00:27:21 ◼ ► And I think we are starting as a society sort of to understand that what we put in our brains, the information we put in our brains is equally important as the food we put in our bodies.
00:27:39 ◼ ► I like that and I think part of the weakness of humanity is or a weakness of human nature is that we are captivated by novelty and the new.
00:27:57 ◼ ► And the food world when I was a kid, I was born in 1973 so I'm a kid of the 70s and the early 80s and the space race era was still, the heyday was the 60s.
00:28:18 ◼ ► But it was still kind of disseminated coming into the 70s and it felt the future was, you just think of 60s imagery of the science fiction future of tomorrow and part of it was stuff like it seemed cool to eat food like the astronauts did.
00:28:39 ◼ ► That was prepackaged and you could just unseal it and it would have a shelf life for months or years.
00:28:53 ◼ ► Prepackaged food that was in a box, you know, and I still eat chips and junk and crackers and stuff like that.
00:29:02 ◼ ► But we just ate more of that when I was a kid because it just seemed, well, that was futuristic and just eating a raw apple felt like old fashioned.
00:29:12 ◼ ► And in hindsight, that just, well, that seems like nonsense and that was my child mind of the 70s.
00:29:20 ◼ ► But I kind of feel like that happened with the internet and our mental diet too, where the whole thing was so new, right? In the early 90s, there was no internet for the public.
00:29:33 ◼ ► And so if you wanted to find stuff, you had to go to the library and read books and libraries are great.
00:29:44 ◼ ► And so if you just had like a trivia question in your head right now in your living room or at a bar or just hanging out with friends on a car trip in the middle of a car, you weren't going to take your car off the highway, find the local library and ask the librarian to get the answer to your trivia question.
00:30:13 ◼ ► And if somebody is wondering what the name of the actress was in the Buck Rogers movie in 1980, you can find it in 30 seconds. Just type it out with your thumbs.
00:30:31 ◼ ► But I feel like there's a sense that nobody, it took until recently for everybody or a lot more people to start looking skeptically at the mental nutrition of that presentation.
00:30:47 ◼ ► And could this be better? Could we still get the answers, but do it in a way where we're not simultaneously being bombarded with distractions from what we were asking for?
00:31:00 ◼ ► Oh, not only we can, but we can get better answers. And I think what you describe is exactly with that.
00:31:16 ◼ ► And Google helped us find this amazing world online, connect with all these people and thoughts and ideas and seeing the browser, seeing the social media,
00:31:32 ◼ ► which is why we were, I think, willing to accept this contract with tech for the first 10 years or so.
00:31:40 ◼ ► But the problem is it went overboard and it went so overboard that it became dangerous.
00:31:49 ◼ ► It became outright bad for us and our health. I mean, just look at what social media is doing to young minds and then misinformation and that curve for the functioning of a healthy society.
00:32:03 ◼ ► So the problem is nobody put a nutritional label on it. The food industry at least had that order. At some point, we put those nutritional labels and said, here's what's in it.
00:32:21 ◼ ► But I remember there was a time where I think, I don't even know if it was the law that they had to print the ingredients, but it was at least tradition that there would be a list of ingredients, but I'm not sure there was any enforcement of the accuracy of it.
00:32:37 ◼ ► And then at some point in the 80s, it became the law in the United States, probably elsewhere around the world around the same time, where not only did you have to print the nutrition, you had to comply with some pretty good regulations on the presentation.
00:32:56 ◼ ► And you had to list the following things. You had to say, what is your, you the maker of the food, what is your definition of a serving size? How many calories are in that serving?
00:33:10 ◼ ► In that serving, how much fat, how much whatever, how much iron? Here's the things you need to list.
00:33:16 ◼ ► And with some kind of penalties in place for lying or just being wrong about those ingredients.
00:33:23 ◼ ► Yeah, but that at least kept the food industry in check to a certain extent. I mean, those ladles are not even perfect today. They are also kind of all over the place. There's no standardized measure.
00:33:35 ◼ ► Like I need the serving size because it doesn't help me compare apples to apple. I have to calculate like in my head when I'm comparing two products and things like that.
00:33:44 ◼ ► But at least there are some checks or regulations and that didn't happen for information consumption. And you could argue some companies are trying to change that.
00:33:55 ◼ ► Apple did with the privacy label for apps. I think that's the step in the right direction.
00:34:05 ◼ ► So, we are now at the beginning stages of this sort of nutritional label for information consumption. We are still probably years ago before we actually put any weight on what that nutrition label says.
00:34:20 ◼ ► Like when the nutrition label for food came out, you didn't care that much how much sugar was in it because you weren't necessarily educated what sugar is doing to your body or trans fats or whatever.
00:34:32 ◼ ► So, that matters. So, we're in those early stages when we are starting to bring awareness to the value of information we put on our hands and our well-being.
00:34:44 ◼ ► And I think there's going to be some time until we start putting weight to those values.
00:34:51 ◼ ► This is why we've caught these sort of deals with a very long-term perspective in mind.
00:34:57 ◼ ► We didn't chase huge growth and whatnot, but rather focused on the long-term goal of achieving what we are trying to achieve.
00:35:09 ◼ ► And truly understanding that these habits that we are going against are some of the most entrenched habits in society and you cannot change that overnight.
00:35:18 ◼ ► It's gone a few years probably and once we adopt them, I think we will be looking at the past and asking ourselves how did we let this happen in the first place.
00:35:38 ◼ ► I think that today if I went back and just somehow there was a video camera of what my family and my parents were not, we didn't eat like fast food all the time.
00:35:48 ◼ ► But if I just looked at what they fed me as like a five-year-old kid, I would have been appalled compared to how we fed my son 15 years ago when he was five years old.
00:36:06 ◼ ► Alright, speaking of advertising and me being a fan of it, I need to take a break here and do a podcast sponsorship.
00:36:18 ◼ ► Memberful offers best in class membership software to creators and it is used by some of the web's biggest creators, publishers, and media companies.
00:36:57 ◼ ► A) I don't know why anybody would sign up to use such a service without a promise like that.
00:37:01 ◼ ► But B) from Memberful's perspective, that's the confidence they have that if you start with them, you're going to want to stay with them.
00:37:10 ◼ ► You can set up and manage your membership program with Memberful's intuitive platform and create multiple tiers at different payment levels, different payment options that cater to different audience segments and maximize your revenue potential.
00:37:23 ◼ ► You can ask them for help because they've helped so many creators get their membership successful, and Memberful only succeeds when you succeed.
00:37:33 ◼ ► They can say, "Here's what's worked for other people in a similar position," and guide you to what type of pricing or what sort of tiers you should offer.
00:37:42 ◼ ► You get a direct connection with your audience, and you can build a closer relationship because Memberful integrates with all sorts of other products like WordPress, MailChimp, Google Analytics, Discord.
00:37:55 ◼ ► So, like with Discord, you can set up areas of a Discord server for member-only access gated by their Memberful membership.
00:38:08 ◼ ► And again, dedicated, top-notch customer support from Memberships' team of real human beings, not chatbots, not AI, ensuring that you get the help you need to succeed in building and maintaining your membership program.
00:38:26 ◼ ► It's all just easy breezy, nothing to worry about, and it helps the people who I want to support be supported.
00:39:07 ◼ ► That's a very direct user-centric business model, like you pay for your iPhone, you pay for your search engine.
00:39:16 ◼ ► Allows us to focus entirely on the user because we give the user this magical power to walk away with their wallet if they don't like something.
00:39:32 ◼ ► Right, so in other words, if somebody says, "I'm going to give Kagi a try," number one, you do offer a free trial.
00:39:41 ◼ ► Yeah, we give 100 searches for free so you can get started, see the quality of results, see all the features we built, and decide if that's for you.
00:39:51 ◼ ► Like I said, I've been a user for over two years, and I just, before we started recording, looked up my recent personal stats.
00:40:10 ◼ ► You get 300 searches per month, and at $10 per month for professional, and me, I'm a professional internet user.
00:40:39 ◼ ► And when I first signed up, I think that it might have been slightly different two years ago, but it was around 300 as a, hey, do I want the lower plan or higher plan?
00:40:59 ◼ ► Well, it can't be 100 because that would be three a day times 30 days, roughly 100 a month.
00:41:18 ◼ ► I think an awful lot of people out there who might be listening who, like me, don't know how many times you search your current search engine, 300 might be more than enough.
00:41:27 ◼ ► I'm a little over that, but I'm also happy to support you, and I don't want to worry about limits.
00:41:34 ◼ ► I don't want to pry into, if I ask you any question, you don't want to reveal the answer to, but how typical is that for users of Kagi?
00:42:00 ◼ ► But also one thing that happens with people coming over to Kagi from other search engines is that they generally find things faster, and they don't have to do as many searches.
00:42:14 ◼ ► We have anecdotal evidence, people that change their search history and compare, before coming to Kagi, they would have like 1,000 searches a month, and then would have to apply to 500.
00:42:25 ◼ ► Because Google is getting so bad, and nowadays it's kind of incentivizing you to search more every time you search, the opportunity to serve more ads.
00:42:37 ◼ ► Where for Kagi, we really optimize for the thing you're searching is in the first three results every single time.
00:42:45 ◼ ► Because that's the promise to our customers, right? And if we don't follow up on that promise, you cancel.
00:42:52 ◼ ► So we have a pretty strong positive feedback loop going on there, and we are trying to do everything we can to make sure that whatever you're searching for is not hidden anywhere.
00:43:04 ◼ ► And we try to surface sites that other search engines don't, like the small web and non-commercial content is sort of forgotten.
00:43:14 ◼ ► So we all lead an effort to really help the user do what they came to search engine for, which is find information they're looking for and be gone.
00:43:25 ◼ ► So I'm happiest when people spend the least possible time on Kagi, which if you were a product manager at Google and said that, your crowd would be fired.
00:43:38 ◼ ► That is sort of the fundamental truth. And the difference between old Google and new Google is one way of defining it is old Google was optimized for you spending as little time at Google.com as possible.
00:43:53 ◼ ► You go there, it loads fast, you type what you want, hit return. Your result is as near the top as possible, often first.
00:44:02 ◼ ► You click it and then you just go, you leave, and they just trust that you'll come back.
00:44:15 ◼ ► That is what allows you to do that. And it's why this is fundamentally and radically different than any other search experience.
00:44:25 ◼ ► I do think it is such an, I think there's a couple obstacles. I know you're aware of them, but I'm talking to the listeners of the show.
00:44:37 ◼ ► But the obstacles that I think people have to overcome to think about this are, it's so ingrained in not just individuals, but just culture, society, that internet search is free.
00:44:57 ◼ ► And even if you don't use Google, you go to Bing.com or DuckDuckGo, who is in, I think, probably in a certain sense, similar to Kagi and having good ideals.
00:45:14 ◼ ► But also therefore your rival in a way, or a big rival, not because they're a big company, but because the idea of an independent small company focused on privacy and quality is a rival to Kagi.
00:45:28 ◼ ► But DuckDuckGo, it's free. If you would like to switch to DuckDuckGo for your internet search, you can just go to DuckDuckGo on your own and search as many times as you want, or change your default search in your browser to DuckDuckGo.
00:45:43 ◼ ► And you don't pay DuckDuckGo anything. And Kagi, that is just so different. Before we talk about quality or whatever, just the nature that the basic idea is you can try it for 100 searches for free, just to get a taste of it.
00:46:01 ◼ ► But then you become a $5 a month paid customer is such an obstacle for people to overcome. But people pay $5 a month for so many different things. People always use the analogy to a cup of coffee.
00:46:14 ◼ ► But it's a good one because lots of people buy a cup of coffee every day. And $5 a month is not that much money for something that you use hundreds of times a month.
00:46:25 ◼ ► But it's an obstacle for people to overcome. And I'm curious how you've approached trying to get people to just overcome the hurdle of paying something, no matter how much, for something that so many people think is just supposed to be free.
00:46:45 ◼ ► Yeah, it's something I think about a lot. I think there are not a lot of examples in the history of tech where a company, a new company had to compete with a business model of free and also win.
00:47:02 ◼ ► There are really not many. And you're right, it is very hard, I think, from a perspective of just inertia to overcome that. I think there are a couple of things that are changing in the tech today, namely the rise of AI.
00:47:21 ◼ ► And suddenly there are millions of people paying for this sort of information consumption, which in some ways is competing or completing search engines, depending on how you want to see it.
00:47:38 ◼ ► But so I think that's definitely a positive thing. And the main thing there is once you're paying for the information being consumed, you're aligning incentives with information providers.
00:47:49 ◼ ► So I think there will be many more people in the future who will value that sort of a contract. What Kagi is really doing is rewriting this contract with tech and saying you're not paying with your attention and data anymore, you're paying with your wallet.
00:48:06 ◼ ► And so that's one thing. The other one is we are really, really trying to make Verge really, really good. Our goal is Kagi to be 10 times better at all times for any query.
00:48:22 ◼ ► We are not there for all queries yet, and there are some things we need to improve on, like maps and local searches and shopping, for example. But for the most part, when you're looking for information, just to take your example on Best Backpack, Kagi will surface discussions from the web.
00:48:41 ◼ ► And Kagi will group all the lithicals together, those annoying top 10 Best Backpacks 2024, which are all the same, and many of them didn't even review the backpack and just slammed a few ladling there.
00:48:55 ◼ ► We will group all them together, and then we would surface content from these personal blogs and websites that somebody reviewed a backpack recently. So we try to compete on quality, really, and this is the main reason people pay for Kagi.
00:49:11 ◼ ► It's not just that this contract is putting you in the center of our universe, but also the quality of product is sort of like what Whole Foods did to grocery stores in a way. You pay a premium, but there is associated quality with it, and you know what you're paying for.
00:49:30 ◼ ► So I always thought it's not for everybody, and not everybody can afford that in a seal free, but that's fine. I think it will take some time, and there are definitely ways that I think it's a pretty deep topic.
00:49:46 ◼ ► But how should we consume information as a society? I think you cannot expect everybody to be able to pay, but I would expect that the government would step in and potentially subsidize or have a search engine on their own for their own citizens.
00:50:04 ◼ ► We had that with libraries for hundreds of years. The libraries were our window to consuming information that was for all citizens, and we sort of don't have that for the last 30 years. So that is how I expect or would want things to develop in the future, where we would have basically three types of search engines.
00:50:26 ◼ ► One is government-based subsidize, like you have subsidized transportation, for example. The other one is premium quality paid for by colleagues.
00:50:37 ◼ ► And the third one is the advertising base, which really I think in the future will start to carry a label for entertainment purposes only, because the quality of the information you find there and the incentives at play make it at best for entertainment purposes.
00:51:00 ◼ ► I do think, and we could, I could delve into it in entire hours-long podcast on the politics of it, but there is a massive generational, it's not like even a recent, politics has changed in recent years, and it's changed in recent decades.
00:51:20 ◼ ► But on a generational level, something has truly changed, certainly in America, but I think throughout the entire Western world about the view of what the government, or the public, not the government.
00:51:36 ◼ ► The government has become seen as I guess sort of to bureaucracy and they're annoying and the less they are involved, but that we as a society, as a municipality, provide services to each other and we still do certain things like twice a week I put garbage, I just take bags of garbage out, put them in front of my house, and in the morning, the city of Philadelphia has paid employees who come by and pick it up and take it away.
00:52:02 ◼ ► I get water service to the house and electricity and it's all regulated by the government.
00:52:12 ◼ ► In the United States, we have the Library of Congress, which has been part of the country since the founding.
00:52:18 ◼ ► That's why it's called the Library of Congress, and it still serves as like the library of record for what books even exist.
00:52:27 ◼ ► But it stopped in the printed world, right? And I think if you went back in time to 1950, after World War II, and at the beginning of this boom of the post-World War II era of enthusiasm and optimism for what the Western world could achieve.
00:52:50 ◼ ► From an era when we hadn't even put a man in space yet, to saying we're going to put not only put people in space, we're going to put people on the surface of the moon by the end of the decade.
00:53:06 ◼ ► If you had explained to someone in that era what the internet was like, today's internet, you could explain it in terms they would get.
00:53:15 ◼ ► You have a keyboard and you type stuff and it appears on a screen in front of you and you can ask any question you want and get answers.
00:53:21 ◼ ► They would think that, I think most people would guess the Library of Congress would provide that index to the entire web that you would search for.
00:53:32 ◼ ► And that sure, a commercial company would be free to offer an alternative in a way that bookstores have always been allowed to sell books that you could go to the library for.
00:53:42 ◼ ► But that there'd be a central trusted-by-everybody service paid for collectively by our taxes.
00:53:50 ◼ ► And that it would be trusted and deserve to be trusted and could serve as a canonical resource.
00:54:00 ◼ ► And maybe sort of started going wrong, just to go back to my Yellow Pages and White Pages directory, that those were never really provided by the government or by society but were always products of the Ma Bell telephone system.
00:54:16 ◼ ► A commercial entity that sort of didn't pretend or lie to be the government but just sort of acted that it was part of the infrastructure.
00:54:37 ◼ ► Yeah, it is this thing that we submit taxes to and it's providing infrastructure for our basic needs.
00:54:46 ◼ ► And whether that's education, transportation, police, it seems to me that access to information is one of those fundamental things, if not rights.
00:55:03 ◼ ► And basically what happened for the last 20 years is we outsource information access to the world's largest advertising company.
00:55:18 ◼ ► And so that is the main problem I see with that and I've been in Library of Congress just a month ago, you know, and I thought about these things and we basically put the brakes on that.
00:55:31 ◼ ► We obsessed information 30 or 40 years ago when libraries became obsolete and there was just nothing that replaced them.
00:55:39 ◼ ► Apart from this company that became a child company and a private company and that put an intermediary between us and the information consumption.
00:55:56 ◼ ► But I think the right question to ask is why would you tolerate ads in your search results?
00:56:02 ◼ ► And it does require a fundamental thinking and I really cannot imagine our society, whether Kage is successful at sort of achieving this for many people or not.
00:56:21 ◼ ► This business model will stay the main way we consume information, that we will tolerate having an intermediary between us and information just because information is becoming so important for, I mean, it always has been, but the speed of it, we can obtain it.
00:56:42 ◼ ► And in the importance it has on our scientific, political thinking, productivity at work, everything is so huge that just by the nature of things that I think healthy competitive markets, we should get to the stage where those services who are able to provide the highest quality information will sort of surface to the top.
00:57:05 ◼ ► And people will naturally want to pay to have the search results faster and information you're looking for faster.
00:57:12 ◼ ► Obviously, AI is playing a role in this now and kind of changing the paradigm for the user interface, how we consume the information.
00:57:22 ◼ ► I do not think that advertising will be a business model at the end of the day and will continue thriving into this AI sort of age.
00:57:47 ◼ ► I would say the two things I would like to most encourage everybody listening to this conversation is one, to overcome any resistance you might have to thinking, yes, this all sounds good, but it just seems so weird to pay for search engines.
00:58:08 ◼ ► But then the second thing I think people and I personally want to vouch for it is the quality of the search results.
00:58:16 ◼ ► And before I switched to Kagi full time a little over two years ago, I had spent years, I forget how many, but at least five years maybe where my default search engine was DuckDuckGo.
00:58:41 ◼ ► And when I switched from Google to DuckDuckGo, let's say, I don't know, I'll just say seven years ago, so around 2017 or so.
00:59:00 ◼ ► I have not, for example, canceled my Twitter account because I disagree with Elon Musk.
00:59:06 ◼ ► I use Twitter a lot less than I used to, and it's so good that Mastodon and now Blue Sky especially, and even Threads as an alternative are thriving.
00:59:17 ◼ ► But I don't do that. If there's an actor or an author who's involved in a personal scandal, I don't not read their books or watch their movies.
00:59:29 ◼ ► If I think it's a good movie, I watch the movie, and if I think it's a good book, I read the book, whether the person had a personal scandal or something.
00:59:38 ◼ ► And I didn't switch from Google to DuckDuckGo as a stunt or a statement about my growing antipathy towards Google and where they were heading.
00:59:57 ◼ ► But the fact is DuckDuckGo was pretty good in terms, and for most of my searches, was giving me as good or better results than Google presented in as good or definitely better way than Google.
01:00:13 ◼ ► Less distractions, less, no ambiguity between what might be a paid search result or not, and DuckDuckGo supported a thing they call them bangs, where if you search for best backpack and you don't like the results you see in DuckDuckGo, you could redo the same search and put space G exclamation mark.
01:00:36 ◼ ► G and then the exclamation mark. That's why it's called a bang. And DuckDuckGo would redirect you to Google to perform the same search.
01:00:46 ◼ ► They make it as easy as possible if you would like to try the same thing in Google and see what you get.
01:00:55 ◼ ► Koggy supports a lot of the same shortcuts. Maybe all of them? I don't know, you have a lot of the same type shortcuts.
01:01:01 ◼ ► I used that from the get go very little, and that's why DuckDuckGo stuck as my default for years.
01:01:10 ◼ ► Because if I used it all the time, I would have switched back to Google. I would have. The quality had to be good enough. I was not making a statement.
01:01:19 ◼ ► In the two years since I switched from DuckDuckGo to Koggy, I used that shortcut so much less than I used to. Almost never.
01:01:32 ◼ ► I mean, in fact, I wish before we recorded, I'd looked at my Koggy history and see the last time I did it.
01:01:38 ◼ ► I almost never, almost never use it. And often when I do, if I'm searching for something that seems, I know in advance, this might be really hard to find in any search engine.
01:01:51 ◼ ► And I search in Koggy, and I look at the ten results, maybe I scroll a little further, I don't see anything that looks like the answer.
01:01:59 ◼ ► Let me try it in Google. G-bang on my query. At least nine times out of ten that I do that, which is already very rare, Google doesn't give it to me either.
01:02:11 ◼ ► I'm searching for something that's hard to find. I almost never go to Google search anymore.
01:02:16 ◼ ► That is how good Koggy search results are for me, John Gruber, personally. Almost never.
01:02:22 ◼ ► And it really shows to me that you and your team are relentlessly focused on that aspect, because it wouldn't be true otherwise.
01:02:33 ◼ ► I find I like the people at DuckDuckGo. They're actually sort of low, their headquarters is very local to Philadelphia.
01:02:41 ◼ ► I wouldn't say I'm personal friends with them, but I've met the founder and I know a couple people who work there.
01:02:47 ◼ ► I like them, I like their ideals, but just in terms of pure quality, Koggy is not just better than DuckDuckGo, I think it's better than Google.
01:02:57 ◼ ► Just quality. Quality alone with no "Oh, I don't like ads" or "I've got an ad blocker." Like, that might be a common thing for listeners on my show.
01:03:06 ◼ ► I don't really worry about, somebody might say, I don't worry about Google's ads because I have an ad blocker that blocks them.
01:03:17 ◼ ► Well, thank you so much for that, John. And first, let me say that we did not pay John to say this.
01:03:30 ◼ ► But since you are so generous, let me also be generous and share four ingredients of Koggy's secret. So I'm laying down our gems, the business secrets that anyone can use to copy Koggy tomorrow and create a competing service.
01:03:55 ◼ ► Number one, the crazy, stupid thing, but it works. We downrank websites that have a lot of ads and trackers on them.
01:04:11 ◼ ► Well, it turns out that the more ads and trackers the website has, it usually correlates with lower quality content.
01:04:19 ◼ ► Think spam, think affiliate crap. So the more ads and trackers there are on a page, you know that the purpose of that page was not to inform and educate you.
01:04:41 ◼ ► Number two, we focus our crawling on the non-commercial web. Again, those are websites that do not have a lot of ads and trackers and they are like personal blogs and websites.
01:04:53 ◼ ► We maintain a project called Koggy's small web that anyone can submit their website to.
01:04:59 ◼ ► And it's basically our local index of high quality stuff. And you will see these personal voices surfaced in our search results and our users love them because this is something they cannot find anywhere else.
01:05:14 ◼ ► No other search engines is basically interested in surfacing personal content anywhere.
01:05:24 ◼ ► Very high quality, especially when people write it because of their passion. There are no ads on that site. They're not trying to monetize you.
01:05:32 ◼ ► You can have a Nobel Prize winner blog post that wouldn't surface in other search engines. We would find it and surface it.
01:05:41 ◼ ► And I think that that's really unique and what gives Koggy this sort of humane fear and our mission is to humanize the web.
01:05:51 ◼ ► And this is one of the reasons we are doing this. Trying to really connect people to other people.
01:05:57 ◼ ► And so in Koggy, we would surface results you will not find anywhere else, even from like a web archive blog posts that are down for 10 years and do not exist on the web.
01:06:10 ◼ ► We still think they should be surfaced if they are relevant. And we would even surface things like that.
01:06:16 ◼ ► So there's a unique content that is high quality and that you cannot find anywhere else.
01:06:21 ◼ ► Number three is personalization. In Koggy, you can block a website from showing your searches off.
01:06:29 ◼ ► So even if we sometimes surface something that's bad, bad results, you're two clicks away from saying, make this website dead to me.
01:06:41 ◼ ► And I don't want to see it in my results ever again. And you can do the opposite as well. If there is a site you like, you can promote it.
01:06:48 ◼ ► So here we are really trying not to impose our biases on what you want to read and what you prefer, but rather give you the user the power to do that.
01:07:00 ◼ ► And number four are just all the other search experience features that we built, like lenses.
01:07:06 ◼ ► For example, if you have favorite websites you like, like hobby sites or recipes, you can just create a lens and they would search only within the subset of those websites.
01:07:17 ◼ ► And you have some pre-made lenses like academic research or programming or things like that.
01:07:24 ◼ ► So these four things are what our users tell us are the main reason they're paying for Koggy and why they're anyone just to take those and copy them and create our own version of Koggy and make an alternative to the way we are consuming information.
01:07:49 ◼ ► Koggy is this mobile-strapped company. We are not VC funded and I really welcome more companies entering the space and helping educate the market about the value of information.
01:08:02 ◼ ► That's exactly what I mean about the order of somebody's priorities because I feel the same way and I've done this at times where I've given talks at conferences where I talk about how I write Daring Fireball, how I turn Daring Fireball into a sustainable business for me and my family.
01:08:24 ◼ ► And I would encourage as many other people who might want to do what I do to learn any lessons that I could teach them with.
01:08:33 ◼ ► I can't think of anything that I would hold back as secret sauce, even though it might welcome people who would write about the exact same things I write about.
01:08:43 ◼ ► If there were more sites like mine or Jason Snell and Dan Morin at Six Colors, Federico Vitici and his team at Mac Stories, I wish there were more sites like that so that I could read them.
01:08:57 ◼ ► It's not a zero-sum game and I believe it from your perspective and I think sharing that is indicative of it.
01:09:08 ◼ ► Let's talk about how you build the index at Kagi because I think I know a little bit of this but I actually don't know a lot.
01:09:17 ◼ ► But to me it's like the biggest curious question is, you talked about the small web index that you do the crawling for, but it seems too good to be true to think that Kagi's entire index is your own crawling because the internet is too big.
01:09:36 ◼ ► It is too good to be true, that's not the case. You guys have a multifaceted backend for the crawling. Can you explain it?
01:09:45 ◼ ► Sure. So we use basically every other search engine out there to provide basically as a source and then we have algorithms to decide which ones we actually use and including our own index.
01:09:59 ◼ ► Our own index is relatively small to all these other big players out there and as I said we are focusing on just a very specific portion of the web which is still manageable to crawl and index efficiently which is the non-commercial web.
01:10:14 ◼ ► And it's relatively small to everything else but at this stage we are not really interested in crawling the entire web nor I think that's something you can successfully do today. I think that sharing left the station basically early 2000s and just the complexity of the problem of crawling and ranking the entire web is just so huge.
01:10:39 ◼ ► I like to compare it to analogy of rebuilding the entire railroad infrastructure of the US in order to put one train on it. Nobody would do that and you would instead use something that somebody already built and there's no reason to reinvent the wheel.
01:11:00 ◼ ► No, it's possible and Microsoft tried that with Bing and they spent 20 years, I think 100 billion is what I read and had thousands of the smartest people working on it and we all know what Bing results are like and I'm not saying they're necessarily that bad but they are not as good as Google and that's a billion dollar company trying that.
01:11:21 ◼ ► So it's really impossible for a small startup to compete with that nor I think we should. I think we should instead focus on providing the different business model. I think we should focus on innovation in search and there's also a lot of regulatory pressure now with the DOJ trial with Google.
01:11:41 ◼ ► You're finding what to do. The search index is one of those things that the DOJ suggested in their proposal to be open basically and just to prevent what I just described and also help proliferate startups that will offer different experiences.
01:12:00 ◼ ► There's no reason why there wouldn't be a search engine just local for some time that somebody could build on this vast index and there should be no reason why our governments should and I mean government in a positive sense here should not provide a free search engine for its citizens.
01:12:19 ◼ ► Search.gov should exist and Google.com should exist and Kavi.com should exist and they should all share this one index that is really good and has been built over the last 20 years and then they would compete with different business models which would create different product outcomes and user experience and all that.
01:12:45 ◼ ► So that's how we approach that. So we're trying to leverage what our strengths are which is product thinking, trying to innovate in search, not necessarily trying to solve a problem that's almost impossible to solve and we don't want to rebuild the railroads.
01:13:03 ◼ ► We want to put a train on the railroads which you can pay for. You can pay that. So imagine Google as a train that is free but you have to watch ads nonstop.
01:13:20 ◼ ► So that's the situation we have currently basically search. All the trains are free everywhere you go. When you enter the train you get advotes and you're being played ads and you're watching ads nonstop and you're saying I want to pay for my train. I don't want that crap. Can I do that?
01:13:36 ◼ ► But that's the Kagi train and I think government should also have their train which is maybe free or subsidized like train. So that's the healthy transportation landscape that we are trying to create basically.
01:13:51 ◼ ► And so do some of your results or a portion or a large portion come from Google's index? When you search at Kagi you get results powered by even Google's index?
01:14:02 ◼ ► We use five or six major search engines so everything that exists in the world and another advantage of doing that although it costs more money is basically we ensure that if it doesn't surface on Kagi it probably doesn't exist anywhere which is what you describe.
01:14:21 ◼ ► You don't find something on Kagi you go to Google it doesn't exist there but since we use every possible search engine that we can then we kind of ensure across the board access to the information and also remove some of the biases as you mix all these in.
01:14:39 ◼ ► You sort of remove some of the biases that the search engines have and then when you put on top of that personalization like the users can then further tweak their search feed and block websites or promote a website they like you get this sort of final outcome of Kagi results being super high quality.
01:15:00 ◼ ► I've done some research to do this interview and I've like I said I've been a full time Kagi user for years now so I kind of knew that but I think that's very surprising to most people listening to us that Google has offers that you're not scraping Google results and just stealing it it's something you pay a lot of money to do.
01:15:30 ◼ ► So Google has an offer that is basically a franchise and you have ability to get Google searches out but you also have to get Google ads so it's one package and we have been trying for years to sort of license Google results in a way that we don't get the ads so what we currently do is there are services out there that
01:15:54 ◼ ► basically resell Google results that we use because we cannot directly retry but Google is not ready to do that and this is what the DOJ trial I think is very important for.
01:16:07 ◼ ► So some of the search engines we license that we can some of the are getting to these resellers so it's a mixed bag but all in all there are five or six search engines that we use.
01:16:19 ◼ ► Right. I don't want to sidetrack it because it's a whole other topic but I think, I think much of what the DOJ is asking for after winning their monopoly lawsuit against Google is ill considered, in my opinion.
01:16:35 ◼ ► But the idea of a requirement because of their monopoly stature that they would have to, I forget the exact terms that the DOJ is using but that they would have to offer a sort of direct API for the search results that would be paid on what's the industry term, reasonable terms.
01:16:59 ◼ ► I think it might make it very difficult for a free search engine to use that to compete with Google because the free search engine would have because it's free would have to use ads and wouldn't be as good at selling ads as Google and therefore wouldn't be better than Google.
01:17:17 ◼ ► But that a paid service like Kagi where you have recurring real money subscription revenue coming in per user could be used to foot what would be a hefty bill.
01:17:30 ◼ ► And that it would be a virtuous circle where it actually would work out, where it wouldn't be the government saying that Google must give away its search results to would be competitors free of charge but that they would be compensated reasonably for the, for offering that.
01:17:51 ◼ ► In the same way that the phone monopoly was broken up here in the US in the 80s, it wasn't just give it away, it was there should be other competitors that can offer long distance phone service over these wires that are already the physical infrastructure connecting the telephones in houses and businesses around the country.
01:18:16 ◼ ► Yeah, and I mean it's a pretty deep topic but since you started it, it's just, it is a very deep topic.
01:18:25 ◼ ► The Google was found a monopoly and it has used this position over the years to accumulate this massively valuable resource which is the index of the web.
01:18:47 ◼ ► And it's uniquely irreproducible, similar to the railroads that I'm in. So in our view, it's a uniquely valuable resource that prevents anyone from building anything similar and evidence is Microsoft did Bing.
01:19:10 ◼ ► And two, if you want to have a competitive marketplace, which is the whole point of the Sherman Act, then you really need to have a mechanism for other search startups to be created.
01:19:23 ◼ ► And if the requirement for them is to replicate Google index to exist, well, obviously that's not going to happen.
01:19:33 ◼ ► In the past, there's something called the essential hostilities doctrine and a number of times it was involved for monopoly situations.
01:19:42 ◼ ► And I'm not a lawyer, so I don't want to go too deep into that, but people can find, read more about it.
01:19:47 ◼ ► And the first time it was used actually on railroads and in '89 railroads were built by these private companies and then the government one day realized, hey, this is a massively useful resource and infrastructure that should be public good.
01:20:04 ◼ ► And then through, you know, serious acts of Congress and so forth, they made railroads basically, you could access them at a fair established price and things like that.
01:20:17 ◼ ► And it has gone on over the decades with power lines, with telecommunications, even news and things like that.
01:20:30 ◼ ► Now, it hasn't never been done in digital space and a lot of people still sort of have a difficulty understanding this search index as this valuable natural resource in a way.
01:20:44 ◼ ► But I think really in order to have a healthy competitive market, and by the way, I don't think that all DOJ suggestions are really productive.
01:20:54 ◼ ► Like, I don't think spitting out Chrome or Android or any of that is because the question I'm asking, what problem are we trying to solve here?
01:21:10 ◼ ► And I argue that it's not the lack of choice, which is one of the primary arguments for this.
01:21:23 ◼ ► I don't think it's lack of choice. I think we underestimate consumer intelligence here by doing that.
01:21:29 ◼ ► It's really the lack of access to different business models and products and competitors.
01:21:42 ◼ ► So forcing somebody to choose Bingo or Google on a choice screen or even removing Google like some of these, I don't think it's productive.
01:22:01 ◼ ► And to me, the essential problem is that there is an intermediary in the way we access information as a society that is unhealthy for a society.
01:22:13 ◼ ► This business model has led to that. We need to solve by creating a healthy marketplace of search engines where the consumers will have other options other than ad based search.
01:22:31 ◼ ► And our proposal was to focus completely on the essential facilities doctrine and treating the search index as this unique resource that would be open then for companies to use, private companies, Google themselves to use and also pay into that government to use if they want to create search.gov.
01:23:05 ◼ ► And I think this is how Mike monetizes it better than Google because Mike can find a local coffee shop and whatnot and create his own search engines.
01:23:15 ◼ ► And that way you kind of paved the way for having thousands of search engines, not just three.
01:23:21 ◼ ► Everyone could create a search engine and then whether they make a successful business or not is a different question.
01:23:40 ◼ ► I again, that was a wonderful answer, very thoughtful and largely aligned with my thoughts and in particular, I completely agree and will not further go down that path because it could take forever.
01:23:54 ◼ ► But completely agree with you that the whole part about spinning off Chrome and Android is a distraction from the core problems.
01:24:07 ◼ ► And even if there's even if we can have a variety of answers to what is the question, what is the problem that is trying to be solved?
01:24:18 ◼ ► I think all of them, all the only possible questions that we are trying to solve, what are the answers to what are the problems should at least start with.
01:24:45 ◼ ► And for example, forcing Google to divest itself of Chrome in and of itself not only wouldn't help Chrome users, it would be detrimental to Chrome users.
01:24:57 ◼ ► Part of the whole reason they use Chrome, I have a whole piece I hopefully will write before the end of the year at Daring Fireball that's half in draft.
01:25:30 ◼ ► And it doesn't have any integration with Google's products or services, or even have a default search engine plugged into the thing.
01:25:43 ◼ ► So why would another company want to spend $40 billion to buy Chrome from Google when they can do it today?
01:26:00 ◼ ► Brave search engine and Brave the browser is just a version of Chromium that they've added Brave branding and integration to.
01:26:11 ◼ ► So forcing Google to get rid of Chrome is only going to hurt the Chrome users who right now count on the integration with stuff like Gmail and whatever.
01:26:24 ◼ ► And I also think one way, and again, I'm not a lawyer either, and I'm certainly not an antitrust specialist.
01:26:30 ◼ ► But one thing that I think armchair monopoly critics get confused about is the difference between a very successful company with a large, perhaps even over 50% market share, who's just very good at competing in the market they're in, versus a company who has carved out a space where competition is no longer possible.
01:27:21 ◼ ► And I think even Microsoft would agree, failed to achieve what Bing set out to achieve.
01:27:29 ◼ ► And even whatever Bing's market share is, it is not as big as they would have thought was the minimal standard for, okay, Bing was a success.
01:27:41 ◼ ► It's not just that Microsoft's a big company with a lot of money, they're a big company with a lot of money whose entire company history is that they are very, very good, almost too good, and got themselves into antitrust trouble at when they see a market they want to move in on, that they are very good at making a competing product that comes in and takes over.
01:28:25 ◼ ► And with search, they were unable to do it, even when they threw the full power of the company behind it.
01:28:32 ◼ ► And I think it's as close as we'll ever come to proof that it's impossible at this point to build a competitor to Google's index.
01:28:42 ◼ ► Yeah, and not just that, but Microsoft made another mistake, which is to go head on with Google with the same business model.
01:28:52 ◼ ► You know, just have the hurdle of the index where Google obviously had early mover advantage, but also you're trying to compete on the business model.
01:29:07 ◼ ► Because if you're trying to offer similar search results, but also have ads and all this, there's very little incentive for a user to switch.
01:29:17 ◼ ► So in order to have this healthy competitive marketplace, you really need to have products with different business models.
01:29:25 ◼ ► And in order to do that, really, as you said as well, you have to identify the consumer harm.
01:29:34 ◼ ► Because that is the point of the healthy or a definition of healthy marketplace is one where you reduce the consumer harm to minimal.
01:29:51 ◼ ► I invite people to go to blog.caghi.com and you'll find it there where we basically explore this in very much detail.
01:30:01 ◼ ► All right, let me take a second break here and thank our second and final sponsor of this episode.
01:30:15 ◼ ► Squarespace. You can go to squarespace.com/talkshow and by using that URL, they'll know you came from the show.
01:30:22 ◼ ► I talk about Squarespace all the time. They're the longest running sponsor on this show.
01:30:27 ◼ ► But the whole spirit of the company really does fit into everything that we've been talking about on this episode.
01:30:33 ◼ ► Which is that it's a platform where you can build your own small web presence on the open Internet.
01:30:41 ◼ ► And you register your domain with Squarespace, pick your template, tweak your template, start with a blank slate and design your own entire look and feel from scratch using their built-in design tools.
01:30:55 ◼ ► Which are super easy to use, super visual, don't require any knowledge of HTML, CSS and JavaScript.
01:31:03 ◼ ► Or to go all the way to the other extreme, support your being able to use your expertise in coding with CSS and JavaScript and HTML to build your own website.
01:31:14 ◼ ► Whichever way, wherever you land on that spectrum of totally visual, want to use a visual design tool, or want to roll your sleeves up and get into the code.
01:31:27 ◼ ► And they've got so many built-in features and functionality from building a store where you can sell stuff and they take care of all the commerce and the credit cards and the integration with stuff like Apple Pay and Pay It Later services and stuff like that.
01:31:42 ◼ ► All sorts of other things that you might be interested in, email and sending out newsletters and stuff like that.
01:31:49 ◼ ► Absolutely top-notch human tech support that can help you with any sort of problem you have along the way.
01:32:00 ◼ ► But if you run a business or service that requires invoicing and a collection of payments from the invoicing, Squarespace has that built in.
01:32:09 ◼ ► You can send beautiful invoices to your customers, to your clients, and they can pay you directly through Squarespace and it goes right to your bank account.
01:32:19 ◼ ► Built-in analytics, you name it. Soup to nuts. Everything you need to have your entire presence on the web, Squarespace supports it.
01:32:37 ◼ ► Use it for a full month. Build your website. See how you like it before you try anything else.
01:32:46 ◼ ► Again, 30 days. And squarespace.com/talkshow and they'll know you came from here and you'll save 10% off your first purchase of a website or domain, including paying for an entire year at once.
01:33:02 ◼ ► We should talk about AI before we move on to Orion. And ChatGPT or OpenAI, whatever you want to credit, has opened up ChatGPT web search this week as a sort of big money.
01:33:24 ◼ ► Maybe the highest, sort of the anti-Kagi, although they do collect, OpenAI does collect money from paying users.
01:33:32 ◼ ► But Kagi has, it's not just a search engine. And there are, you know, we could keep going on and you talked about lenses.
01:33:50 ◼ ► If you end a query to Kagi with a question mark, you get AI powered answer at the top of your search results because you use the question mark to ask for it.
01:34:06 ◼ ► So I'm curious where you see this playing into Kagi's core business of bringing humanity back to the web.
01:34:20 ◼ ► That's a great question. Well, I think the way you do it is follow how science fiction movies have done it.
01:34:36 ◼ ► This is Sigourney Weaver in Alien or Star Trek. Like these AIs are sort of around, but there is still pretty much human interaction and we are in charge, so to speak.
01:34:49 ◼ ► So we treat them as a tool is the key here. And the last few years since GPT-3 came out, I think a lot of tech world jumped on the train and calling this
01:35:03 ◼ ► Euro for cancer basically equivalent in tech. And it wasn't clear to me that that was the case.
01:35:17 ◼ ► And we try to think about this as just another tool that you try to mindfully integrate with existing products, whether those are search, browser, email, whatever.
01:35:28 ◼ ► And it seems to me like even today, everybody is just trying to slap chatbots on everything.
01:35:36 ◼ ► Like, you said Word and every single product has chatbots of some sort and AIs and every single billboard in every single airport.
01:35:51 ◼ ► Yeah. And I have to say, although we have an AI product, we remain a little bit of a skeptic of this approach.
01:36:14 ◼ ► And this technology is fundamentally flawed, like the very way it's generated, just predicting X token without any sort of model of the world is forever going to be prone to mistakes like that.
01:36:38 ◼ ► And even Apple had, if you remember from '87, the Knowledge Navigator vision that John Sculley created with George Lucas.
01:36:48 ◼ ► And it's a super, it feels so natural, that vision of you talking to your AI assistant.
01:36:58 ◼ ► But even today, I think we still struggle to completely deliver on that vision. Among other things, I think one of the issues is you need to control the product end-to-end.
01:37:12 ◼ ► So from hardware to software, you cannot just build a map on the phone that will act as your personal system because it needs to access your files, your calendar, your email, all these things.
01:37:24 ◼ ► So I think there are a couple of companies that are uniquely positioned to execute on this correctly, Apple being one of them.
01:37:34 ◼ ► We certainly like to give it a try, but it's a very expensive thing. Once you get into hardware, it's a whole different ballgate.
01:37:41 ◼ ► But I could imagine a cadmium version of smart speakers powered by our searches belt, having a service you can trust and pay with your money and all that.
01:37:51 ◼ ► So I think we are at very interesting moments where AI is trying to proliferate its way into our lives.
01:38:00 ◼ ► And we are definitely trying to be very mindful about how we use it because the last thing I wanted to become is a replacement for my mind.
01:38:18 ◼ ► And I think the trick will be to do that in a way that where you still achieve productivity benefits, but you are not completely outsourcing your thinking, so to speak.
01:38:31 ◼ ► I would not want my daughter to be an addict of a chatbot for all sorts of needs and purposes.
01:38:40 ◼ ► So the feature that you mentioned, adding a question mark to your queries are a sort of attempt.
01:39:18 ◼ ► I have written recently, I'm trying to force myself to use ChatGPT more just to learn more about it.
01:39:31 ◼ ► And I have written, and this is true, that as a basic fork in the road, when I'm at my keyboard, at my Mac,
01:40:09 ◼ ► And then I type my query, hit return, and then boom, I've got instant results in a new Safari tab.
01:40:22 ◼ ► Yesterday I was, I sold t-shirts this last week at Daring Fireball and I wanted to send some to Ben Thompson,
01:40:32 ◼ ► my friend and colleague at Dithering, but he's over in Taiwan and I wanted to make sure I was formatting his address right for a package to be sent to Taiwan.
01:40:50 ◼ ► But in preparation for this episode of the show, an interview and talking with you, I've thought more about this question mark shortcut at Kagi.
01:41:00 ◼ ► And it turns out a lot of those questions that I've been going to ChatGPT for, if I think of it as a question and then do the most natural thing and just put a question mark at the end of my query,
01:41:12 ◼ ► Kagi's integration with AI does the same thing that I like about ChatGPT. It gives me an answer right there at the top of the results.
01:41:21 ◼ ► It's, and so I'm now I'm back to, I'm not quite sure why I would need ChatGPT other than for something that I would anticipate having a,
01:41:32 ◼ ► not just one query get the answer done, but an actual back and forth query chat, an actual chat, which is something that Kagi doesn't offer and it doesn't seem to want to offer.
01:41:45 ◼ ► Well, we do actually offer something called Kagi Assistant, which is a window to all the top models out there, including the one subscription.
01:41:54 ◼ ► But yeah, I think there is a difference between the thought of this information consumption scenario when you're looking for certain information and then other things that LLMs are good at,
01:42:08 ◼ ► like expanding an idea or compressing content or translating words to code or so there are different use cases, which I think are pretty good.
01:42:22 ◼ ► And I have been writing code like crazy with the help of LLMs in recent months. Like I never was that productive.
01:42:32 ◼ ► So there are definitely use cases where this technology is super powerful. Also voice to text, text to speech, like God, like miles better.
01:42:43 ◼ ► We recently launched Kagi Translate, which has beats Google Translate completely. It just uses all these advanced models. It's so much better.
01:42:59 ◼ ► In the context of search, you're also experimenting. How do you, what is the next step? Because people, I think, are naturally converging toward the UI of natural language improvement.
01:43:11 ◼ ► It's much better than trying to frame your question as keywords that you type in the search engine historically.
01:43:18 ◼ ► So that's definitely going to be the outcome. And all the science fiction movies also tell us that's the end of the road.
01:43:25 ◼ ► We are talking to an AI. So we are investing in that through the products we're making.
01:43:31 ◼ ► Right now, I think the main problem with what maybe TransGPT is knowing and others who are adding only this is that it is AI who's deciding what sources are being used to sort of condense the information.
01:43:49 ◼ ► So the AI is making the choice, not you. And the second one is that no matter how good the model is, the answer will still depend on the quality of search results.
01:44:01 ◼ ► And this is why I think we will have a natural competitive edge there because we are so much focused on search and have been for the last six years.
01:44:10 ◼ ► And we will be able to provide superior search results to these models. And actually what I think will happen in the future is that the consumer of search results will become less of a user and more of an LN.
01:44:29 ◼ ► So the search will become, it may be even more important, but you will talk to your agent and you say, you know, what are the news for today when you wake up? And it will go to a search engine.
01:44:41 ◼ ► Now the LLM is consuming a search engine. It's not you anymore. And then the LLM is doing the work for you.
01:44:47 ◼ ► And I think that LLM will be subscription-based because I cannot imagine that LLM having an ad where you're 7 a.m. like here's your news, John, and here's for Omaha.
01:45:02 ◼ ► The LLM is not going to fall for engagement-based time consumption, right? Like it's not going to get sucked into spending three hours watching TikTok.
01:45:13 ◼ ► Yeah, but it will do what it's optimized for. So you could say advertising-based LLM. It will obviously follow up with ads. And Amazon tried out with Alexa. That didn't work. I don't think it will work.
01:45:27 ◼ ► So I'm sort of bullish from that perspective, the subscription-based access to information will proliferate even more.
01:45:34 ◼ ► I said TikTok, but I guess YouTube is the better example because to me that is one of the pernicious aspects of Google's entire company.
01:45:45 ◼ ► And I don't think it is a coincidence and it is a huge source of maybe the single quality reason that I prefer Kagi search results to Google is Google's predisposition to answer questions by sending me to a YouTube video with the answer.
01:46:10 ◼ ► My toilet handle a couple years ago, the toilet never stopped running. I know almost nothing about plumbing, but it seemed like a problem I should be able to fix without calling a plumber.
01:46:25 ◼ ► And all Google wanted to show me was YouTube videos. I don't want YouTube videos. I want an answer I can read.
01:46:32 ◼ ► And the answer was trivial, but watching five minutes of a video to get to the answer was a waste of time.
01:46:42 ◼ ► And that's just one example. But once you go to YouTube, if they're successful at sending your, "Hey, I have a question. Here's a YouTube."
01:46:51 ◼ ► And then all of a sudden you're in the YouTube vortex and they're showing you these things on the side, which are remarkably tempting.
01:46:59 ◼ ► I admit it. YouTube knows what I like to watch. And so once I'm there, now I've got to fight the urge that I'm a huge David Letterman fan.
01:47:10 ◼ ► And all of a sudden, here's three of David Letterman's new videos on his YouTube channel. Well, I'm supposed to be working.
01:47:16 ◼ ► That's not going to happen when an LLM is your agent between you and the search engine. And there's no reason for them to even try it.
01:47:28 ◼ ► I completely agree. And I like, I concur with your approach where it's not so much that this LLM technology is this breakthrough and nothing else.
01:47:39 ◼ ► Like it's a new era, like the AD versus BC. This is year zero and nothing previous mattered.
01:47:49 ◼ ► And now we live in the world of LLMs and that's all that matters. It's just a new layer on top of everything that's already there.
01:47:56 ◼ ► And it is a fascinating, useful, powerful layer. But like for answering questions, it's really just the best way to think about it.
01:48:06 ◼ ► And I'm very glad to hear you say that for five, six years of foreseeable future, that's where it is.
01:48:12 ◼ ► It's just a layer that will read the search results for me and format them and be able to cite those sources. Right. Which is important.
01:48:22 ◼ ► Yeah. The knowledge navigator, the old version does call for a bit more than just that.
01:48:37 ◼ ► And right now, what makes it difficult for companies like mine to sort of deliver on this vision is the fragmentation of platforms.
01:48:55 ◼ ► Just getting the, yeah, well, getting it as your default search engine is such a hard thing.
01:49:03 ◼ ► Like Safari doesn't support it natively, so you have to use our explanation, which is not perfect because there's no good way to do this.
01:49:11 ◼ ► So when you just think about how much friction things like that cause, it's very hard to be able to completely deliver on this mission that I just described.
01:49:21 ◼ ► Very frictionless access information. So there are a couple of companies that are sort of set up to do that in a proper way.
01:49:29 ◼ ► And we can only hope to start working on those things. But you need to control your hardware to deliver that.
01:49:53 ◼ ► And I find it, I understand the business reasons, but I don't understand, I think those business reasons should not override the user centric reasons that Safari to me, as far as I know, stands alone amongst popular browsers for not allowing the user to easily add their own,
01:50:22 ◼ ► default search engine choices. And typically it's, you know, for example, to add it to my beloved launch bar tool, or which is a thing like other people use Alfred, there's a new one called Raycast on the Mac where you type command space and bring up Kagi as your search or DDG as a shortcut for DuckDuckGo if you want to search there.
01:50:47 ◼ ► And it's just a URL format. And you could say, oh, so every time I search on Kagi, here's the URL to use, and here's the part of the URL where the search terms go.
01:50:59 ◼ ► And now you could have a new one. And so, like you said, some husband and wife team in the middle of nowhere could start their own search engine.
01:51:08 ◼ ► And they might only have a dozen users, but those dozen users could add that search engine as the default for their browser by doing something very simple.
01:51:18 ◼ ► Safari doesn't support that. The only options listed in Safari for search are, it depends on which country you're in.
01:51:27 ◼ ► But like in the United States, it's Google by default, of course, Bing, Adobe, DuckDuckGo, and Yahoo, I think. You might know more than I do. I think those are the five.
01:51:39 ◼ ► And they all, everybody knows that Google pays Apple $20 billion a year in traffic acquisition costs from the users of Safari who do Google searches that go back.
01:51:53 ◼ ► But all of those companies have traffic acquisition cost arrangements with Google. That's how they get on that list in Safari.
01:52:02 ◼ ► And so, even if only one in a thousand, and it's probably, to be honest, probably less, but if one in a thousand Safari users change their default to Adobe,
01:52:14 ◼ ► that one tenth of one percent of Safari users who do their default searches in Adobe, Adobe has a traffic acquisition cost deal with Apple.
01:52:24 ◼ ► And it pales in comparison to Google's. And Apple's never fully explained that that's why, but it's the only thing that makes sense.
01:52:32 ◼ ► And that doesn't really make sense for Kagi as a search engine that isn't based on advertising to jump into.
01:52:41 ◼ ► I mean, have you had talks with Apple about this or is it like there is no contact point for this? Have you?
01:52:54 ◼ ► I think they are also, well, we are, first of all, what's happening in the market with Google and their deal with Google is likely to be stopped going forward.
01:53:07 ◼ ► But also from a perspective of what is best for the user. And the DNA of Apple has always been to be customer centric.
01:53:17 ◼ ► And I think they always had a problem with sending people to Google and increasingly over the years as it got worse and more sort of harmful to users.
01:53:28 ◼ ► So I think there's willingness in Apple to reconsider how is this done. And this is definitely one of those pain points where I think Apple is letting down their users,
01:53:40 ◼ ► strictly by the virtue of prohibiting what search engine are you able to use in your browser.
01:53:51 ◼ ► And having this thing be locked or the user is sort of taking away that freedom and power that I think Apple has been known for.
01:54:00 ◼ ► So, yeah, I is definitely something we would like to see change that I think it would benefit Apple in a big way.
01:54:08 ◼ ► Right. And at a certain layperson cynical perspective, you could say, well, no, why would Apple ever agree to that?
01:54:17 ◼ ► Of course, they're just going to do the thing that keeps the traffic acquisition costs money flowing from Google as best they can.
01:54:25 ◼ ► But I think it's almost inarguable if you'd give someone it doesn't fit into one sentence.
01:54:34 ◼ ► But a slightly longer explanation is that the reason Apple is a successful company is traces back to 1977 when it was founded, which is a focus on their users and providing the best computing experience all in one.
01:54:52 ◼ ► The hardware is it a hardware company that makes software or software company that makes hardware?
01:54:57 ◼ ► It's both and anything that gets them away from delivering the best hardware, software services combined experience for their users is in the long run against their financial interest.
01:55:12 ◼ ► So even from the perspective of the CFO at Apple, who's all who says I'm only concerned with the financial well-being of the company.
01:55:33 ◼ ► And they're no longer in a position where one product's next quarter raises or lowers the company's financials.
01:55:42 ◼ ► They have got so many different products across so many different segments that they can take the long run, even if the long run is just five years.
01:55:50 ◼ ► But have that mindset would so Orion is Kagi's browser and it is built on WebKit instead of Blink.
01:56:07 ◼ ► Almost all competing Firefox and Mozilla are obviously an exception on their own custom entire engine.
01:56:18 ◼ ► Almost nobody. I don't I don't even know of any other browsers that use Mozilla's engine Gecko.
01:56:30 ◼ ► Talk to me about why you made Orion, why you chose WebKit and why you're focused on the Apple devices as a platform.
01:56:45 ◼ ► So going into this recording, you told me that I would have to use the Chromium browser because this.
01:57:11 ◼ ► And actually there appears to be no no actual feature that WebKit is missing that just the developers were either lazy or.
01:57:35 ◼ ► I really appreciated the nativeness, the performance, the battery life, everything that WebKit brings to Apple ecosystem being a native rendering engine.
01:57:58 ◼ ► And I'm also curious to hear your thoughts because you were there when Safari launched and what was the job and the idea with that.
01:58:25 ◼ ► Old browsers had many white items to show a color picker in 2005, six, seven, eight, all of them removed it.
01:58:51 ◼ ► It has to follow Apple's HIG and there were all these things that annoy me that Safari was missing and it just felt it's never going to be there.
01:59:03 ◼ ► So following all this, like the heritage of Safari and all the Mac browsers through the years, OmniWeb was a big inspiration, Shira as well.
01:59:14 ◼ ► I have an iMac G4 on my desk here actually, which runs like the early versions of all these browsers, which had some fantastic ideas.
01:59:23 ◼ ► And I really wanted a native fast browser, but also one we can extend and add things like a Web extension support.
01:59:32 ◼ ► So one of the things you're doing for Orion is building support for both Chrome and Firefox extensions.
01:59:44 ◼ ► So it's sort of an open platform and you can load an extension build for any browser out there.
01:59:51 ◼ ► And there were things like that that I wanted to do that were hard to do, but I think worthwhile.
01:59:59 ◼ ► And I really want Orion to become that general purpose browser that's good for you as a power user, but also for your grandma.
02:00:10 ◼ ► It kind of gets out of your way. You have full control and you can extend it and you can change it.
02:00:29 ◼ ► It was sort of selfish. I wanted a browser that I can use as well, although Windows is much bigger market, obviously.
02:00:35 ◼ ► But also, since we were launching a paid search engine, I thought that the overlap with Apple ecosystem customers made more sense as people using Apple products are usually more receptive to paying for premium products, which definitely could be is one.
02:01:03 ◼ ► We're just resource constraints. We are trying to get this right on the Apple ecosystem.
02:01:12 ◼ ► There was so much to do. One thing that differentiates building on top of WebKit to Blink is that for Blink, there is Chromium, which is the web browser app framework.
02:01:27 ◼ ► You can just change the name and you have a browser for WebKit. There is no Chromium equivalent.
02:01:34 ◼ ► You have to create every menu, every button, every everything, which is why it took us six years to get where we are.
02:01:43 ◼ ► And on top of that, we also decide to port Web extensions, to port API natively to WebKit.
02:02:00 ◼ ► Well, yes, it takes time to do this properly, but we are determined to do that properly.
02:02:06 ◼ ► And of course, it also has the native ad blocker included and all these good things that a browser should have.
02:02:23 ◼ ► This is sort of a loaded question, but would Orion, it's a massive undertaking, even if, and again, you said you have aspirations to go to other platforms eventually,
02:02:33 ◼ ► but just biting off Mac plus iPhone plus iPad is quite a thing for a small company in and of itself.
02:02:43 ◼ ► Even building on, not building your own engine from scratch, but building on top of WebKit.
02:02:49 ◼ ► But would you have, would Orion even exist if Safari made it easier to just integrate with third party search engines?
02:02:59 ◼ ► Or is there more reasons than just having a natural home for Kagi to be the built in search engine and therefore you have to build your own browser?
02:03:11 ◼ ► There is more to it, and it's also about creating an ecosystem of tools that are aligned philosophically.
02:03:20 ◼ ► And even if Safari was great, at the end of the day, it's still an ad driven browser, although indirectly, by the way Safari is monetized is.
02:03:38 ◼ ► So philosophically, even if Safari had everything that I needed as a user, I still think we would make a case to build a browser that would inherit all the good things and remove the bad one.
02:04:05 ◼ ► You open it in the morning, that's the last thing you close in the evening, and it knows everything about you.
02:04:15 ◼ ► And just the thought that you're letting an intermediary pay for that is chilling for me.
02:04:24 ◼ ► And that we let as a society have browsers that are free, but have so much control and so much knowledge of what we do.
02:04:38 ◼ ► So philosophically, I would still create Orion, but Safari is far from a perfect browser for all these little things that I just described.
02:05:12 ◼ ► And the version of WebKit that Orion runs on Mojave is newer than what Safari runs on SQL.
02:05:28 ◼ ► Which is built into the operating system, whereas Kagi, or not Kagi, Orion has the entire WebKit framework in the app bundle.
02:05:37 ◼ ► And therefore, you can put that entire app bundle self-contained, so if you decide you don't like it, you just drag Orion.app to the trash, and it's gone.
02:05:56 ◼ ► I mean, it's the Mac style of installing software where no hooks, but now it gives you that.
02:06:06 ◼ ► And in some sense, Apple doing it that way is sort of eating its own dog food, living on the version of WebKit that every other third-party developer that uses the system version uses.
02:06:18 ◼ ► But on the other hand, it means older devices that no longer are getting operating system updates are no longer getting browser updates either.
02:06:27 ◼ ► And I've seen it with my parents' iMac, where they don't need a new iMac very frequently, and they don't...
02:06:35 ◼ ► You could say, "Well, John, you're a bad son, you should just buy them one." They don't want me to buy them a new iMac.
02:06:41 ◼ ► But if the iMac stopped getting... I think it's Catalina or something, 10.5, I don't know, 10.15...
02:06:49 ◼ ► But now they have to use Chrome to go to their bank, because the bank's security certificates are more modern than the version of WebKit on their Mac supports.
02:06:59 ◼ ► And try explaining to an 87-year-old, now you have to use this browser for that, but go back to Safari for everything else.
02:07:10 ◼ ► It's complicated. And it does touch on... I've been sloppy about it in my own writing over the years.
02:07:17 ◼ ► And by preparing to interview you and thinking about Orion recently, I realized that I've been sloppy about this.
02:07:32 ◼ ► Blink is the engine. Blink is the equivalent to WebKit. And in fact, started as a fork from WebKit.
02:07:38 ◼ ► People forget this. You don't, I'm sure. But in the early days of Google Chrome, it was using WebKit.
02:07:46 ◼ ► And it was... Google wanted to take the engine in different ways and move it faster than Apple.
02:07:51 ◼ ► And they were like, "Okay, this isn't working out. Sharing WebKit as an open source engine, we're going to fork it."
02:07:59 ◼ ► But as different as they've grown over the years, Blink started as a fork of WebKit in the way that WebKit started as a fork of KHTML back in 2003.
02:08:13 ◼ ► But it's true that WebKit as a framework is something that's been open to build another browser on top of.
02:08:23 ◼ ► OmniWeb back in the day was its own entire application that only in the rectangle in the window of web content was using WebKit.
02:08:33 ◼ ► You could, I guess, you do that with Blink. I'm not aware of any browser that's done it because they all just do the easier thing.
02:08:42 ◼ ► I honestly think it's lazier. And start with Chromium, where you're starting with this built...
02:08:50 ◼ ► It's a whole browser. And again, you can just go to GitHub and check it out or download a recent build and you get a complete working, no company behind it, open source browser that runs on the Mac and Windows and Linux.
02:09:03 ◼ ► And Brave is a fork of Chromium. Microsoft Edge, Microsoft's own browser, is a fork of Chromium.
02:09:11 ◼ ► And that's why they all look like Chrome. They all kind of look like Chrome and there's superficial decorative differences in the branding.
02:09:20 ◼ ► But like the way they show preferences, which I think is so weird where it shows up as a tab in the browser of web content that you scroll.
02:09:28 ◼ ► They all... And it's like, "Oh, why do they all do that?" Because that's the way Chromium does it and that's what they all started with.
02:09:33 ◼ ► Whereas Orion is a fresh start as an application on all of these platforms, not like a copy and paste of Safari with your branding on it.
02:09:46 ◼ ► And has all sorts of interesting ideas. And we can parlay this into that conversation about what Apple was thinking way back when.
02:09:53 ◼ ► But I wrote a lot about OmniWeb in the 2000s because I was a huge fan of the browser and I missed so much. I get why the Omni group got away from it.
02:10:04 ◼ ► It was a lot of work and didn't make business sense, I guess. But like the way that it made thumbnails of the web pages and put them in a horizontal tab.
02:10:15 ◼ ► The closest thing I can think of in terms of the ambition of rethinking the presentation would be Arc from the browser company.
02:10:24 ◼ ► Which maybe, maybe Arc uses Blink instead of Chromium because it is so different. I don't know if you've ever looked at it.
02:10:31 ◼ ► And I love the browser company and Arc for what they're trying to do, but it didn't, the way my mind works, I tried it, I looked at it, I was like, "Ooh, this is super interesting and totally not for me."
02:10:46 ◼ ► And close it and don't update it. But Orion does a lot of stuff like that where you can have a horizontal tab or sidebar of the open tabs in the current window.
02:10:57 ◼ ► Rather than a horizontal one where you run out of space more easily. And it's hierarchical, like an outline, where you can start with a page and then if you open tabs from within it, they open, indented underneath it and you could close the widget and say,
02:11:14 ◼ ► "All of these seven tabs are related to this parent tab, but I don't want them right now. I'm just going to close them so they're not distracting me and concentrate on these other tabs below."
02:11:28 ◼ ► Yeah, and this goes back to us really wanting to, the word is not reinvent a browser, but sort of do what a great browser would be.
02:11:40 ◼ ► Like if you take all the best things from Chrome, Firefox, and Safari and get rid of all the bad ones, what would you end up with?
02:11:55 ◼ ► Before we move on, because I don't want to forget this. With Orion, the business model is you can, anybody can download, install, and use Orion for free.
02:12:05 ◼ ► So there's a difference from Kagi, the search engine, but there is Orion Plus where you can pay $5 a month or $50 a year and unlock more features, including like access to the most recent beta builds and stuff like that.
02:12:27 ◼ ► Yeah, the idea was that Orion will be funded in the same way that Kagi is, which is user's pay per app.
02:12:35 ◼ ► Which the last time that idea existed in 2008, I think OmniWeb was the last bearer of the paid browsers where you could buy and own your browser.
02:12:49 ◼ ► And that's fundamentally, I think, changing the perspective on how you treat the browser, what it can do, and just some of the things that make Orion browser different are it comes with an ad blocker built in.
02:13:02 ◼ ► You have a pop-up blocker, you should have an ad blocker, no major browser has it or at least doesn't block all ads.
02:13:10 ◼ ► Most of them are monetized by ads. The second one, Orion in zero telemetry, so it doesn't send any data about you to us at all.
02:13:20 ◼ ► So zero telemetry by default. So there are all these sorts of privacy, ad blocking, tracking blocking things you can do by default than others.
02:13:29 ◼ ► And so we decided to create the Orion Plus as a way for users to be able to pay for their browsers and use it R, which is great to see.
02:13:44 ◼ ► And now there is a lifetime license for Orion because we really don't have, there is no running costs for the browser so we can do that.
02:13:52 ◼ ► We cannot have a lifetime for Kagi because there's running costs every time you search.
02:13:56 ◼ ► But with a browser we can definitely do that. So we opened up a lifetime license and now again after what is that like 15, 20 years, you can own your browser and know that it's built with your best interest.
02:14:07 ◼ ► And there is a team out there that listens and reads every single user feedback and acts on it.
02:14:14 ◼ ► And that lifetime license is a very reasonable to me $150, which is just three years of the annual Orion Plus.
02:14:23 ◼ ► I never paid a dollar for a web browser. It might seem weird, but if you think about it and if you really, really, it all comes back to the if you're not paying, you're the product, not the customer.
02:14:36 ◼ ► And it is definitely true with Safari. I think Safari is an excellent browser. It is, it has been my default browser for many years.
02:14:46 ◼ ► The only browser that I would even on the radar that I would consider switching to is Orion because I because of WebKit because of your zero telemetry, the the nerd angle of having access to the Chrome extensions in addition to Safari's sort of rather infamously limited set of extensions.
02:15:10 ◼ ► There are a lot of reasons for that, but also the the user interface playground and playing around and this is the area to where it's not, oh, Apple stinks or Safari has a bad interface.
02:15:21 ◼ ► I think Safari is one of the best interfaces of anything Apple makes on across its platforms.
02:15:27 ◼ ► But it's where Apple is sort of handcuffed by the fact that they have a billion users and kind of need to cater to the default user in the way that there's a thriving market of professional software or photograph tools like Lightroom from Adobe down to the indie products like Pixelmator, which Apple just bought, but Darkroom and all sorts of other utilities for more advanced photographers.
02:15:55 ◼ ► That Apple photos not only doesn't support but shouldn't because it would be it would make the product too complicated for the typical user, right? That Orion without getting turning into a terminal app.
02:16:08 ◼ ► And it has a very in anybody who just looks at Kagi.com or looks at the screenshots of Orion that you that you offer will see it's not it is a very beautiful app and it is designed with that indie Apple developer mindset in mind, just like OmniWeb was 20 years ago, but can also cater to a more advanced user or in a way that the default can't and that they just can't break habits, right?
02:16:44 ◼ ► Yeah, you know, we're not trying to do radical things like that, just incremental improvements to what Safari has or ad, right?
02:16:53 ◼ ► And plus some of the things that I think should be difficult in a browser like an ad locker or even like supporting Mojave with a fresh app kit. We kind of go above and beyond and make our life very hard.
02:17:08 ◼ ► But it's truly something I believe in. I want to have a fresh version of AppKit on my Mojave.
02:17:19 ◼ ► I am definitely rooting for you, and I'm inspired to throw myself into it and see how far I get right now and hope that you keep going.
02:17:29 ◼ ► So let me turn the microphone over to you and you ask me what you wanted to ask me about what I think about the early days of Safari.
02:17:36 ◼ ► Yeah, I think I read a lot of your writings from that period and a lot of those things that you notice in those browsers ended up inspiring Orion.
02:17:48 ◼ ► But since you were around when Safari launched, I wonder what were your thoughts on what was Steve Jobs intention with Safari and sort of follow up question.
02:18:01 ◼ ► What's an acceptable price point for a premium browser these days, if at all? How would you think about that paradigm?
02:18:12 ◼ ► Well, I think it's called the Tim Cook doctrine because he said it at an event, and I think Horace Dedue quoted it. I'll put it in the show notes here.
02:18:22 ◼ ► I swear to God, folks, I'm making a note. But it's something to the effect of Apple needs to own and control the core technologies necessary for its products.
02:18:37 ◼ ► And it's a way of focusing them on what they need to do as opposed to just, well, Apple should do everything.
02:18:46 ◼ ► So Apple doesn't manufacture its own RAM chips. They buy them from suppliers and build them and they don't need to. So why should they? They focus on what they need to.
02:18:55 ◼ ► Now they make their own systems on a chip though with Apple Silicon and their devices are all powered by their own CPUs and GPUs.
02:19:04 ◼ ► That's become a core technology for them that differentiates their products from competitors.
02:19:10 ◼ ► I think what Steve Jobs saw was as much as people think Jobs was all about the walled garden and locking people into Apple's stuff and Apple's platform,
02:19:26 ◼ ► and maybe he was against the open web because the open web in some sense is a competitor to, at the time, just the Mac as a computing platform.
02:19:38 ◼ ► Should somebody who's making a new email client, should they make a web client like Superhuman or should they make a Mac app that runs only on the Mac and uses Cocoa as the application framework,
02:19:51 ◼ ► which is at a technical level locks you to the Apple platforms, but as a user provides this standard interface that they're familiar with and gives you all this stuff for free.
02:20:02 ◼ ► And I think people would easily think, well, Steve Jobs would only be interested in the stuff that's exclusive to the Mac.
02:20:10 ◼ ► But I think he was so smart and I think he used the web and I think he liked the web and knew that the web was good for humanity and wasn't going anywhere.
02:20:22 ◼ ► And here we are in 2024 and as many problems as we can talk about that today's web has, it's not going anywhere.
02:20:32 ◼ ► And therefore Apple needed to own it, not just a browser, but its engine because it's a core technology that the company needed to own.
02:20:41 ◼ ► And they didn't own one at the time, had never owned one, and it was a problem for them that when they were looking to move this user base from classic Mac OS to Mac OS X, which took years,
02:20:56 ◼ ► I think they did it about as well as it could be done. It's hard to imagine how they could have done anything differently or done anything faster to make that transition happen
02:21:07 ◼ ► and keep the users that they already had, get them to move to a new platform, which was difficult and broke tons of compatibility,
02:21:19 ◼ ► which is always a recipe for losing your customers while building a new platform that would grow to more customers, which they needed to do to keep growing.
02:21:30 ◼ ► And they did. The default browser at the time was Internet Explorer, which didn't at the time share a rendering engine with Windows Internet Explorer.
02:21:42 ◼ ► It was two different, not just two different browser interfaces, but two different browser engines from Microsoft. And in some ways, the Mac version of Internet Explorer had a better engine
02:21:53 ◼ ► in that it was far more standards compliant and that the people behind it at Microsoft, it was a totally different team, were more interested in open web standards.
02:22:03 ◼ ► The whole Jeffrey Zeldman era of writing HTML that was compliant with the spec and pushing the spec forward and publishing these specs as open things that any other company in the future could build a browser for.
02:22:17 ◼ ► But Internet Explorer was fundamentally an app for the classic Mac OS, and it wasn't built on Cocoa, and it just didn't look great on Mac OS X, right?
02:22:27 ◼ ► Part of the whole appeal of Mac OS X, part of the carrot to get users to switch was look at how much cooler it looks, right?
02:22:35 ◼ ► Everything's anti-aliased, everything that could be transparent is transparent. You know, the lickability of the user interface was like, wow.
02:22:45 ◼ ► And I have a soft spot for the look and feel of classic Mac OS, but in terms of, ooh, which one looks cooler? It was Mac OS X.
02:22:55 ◼ ► But Internet Explorer, it's like sort of halfway, like sort of fakey fake there, and it wasn't fast, and it kind of lost compatibility with the leading edge of Internet Explorer for Windows.
02:23:10 ◼ ► And however much the Microsoft team that built the Internet Explorer for Mac might have wanted the best for it, Microsoft as a company at the very top levels was never going to really care about making a browser for Mac.
02:23:24 ◼ ► Making a browser for Apple. You know who would really care about making a browser for the Mac? Apple. And so they had to do it.
02:23:32 ◼ ► And I don't think it was a sure thing that it would be successful. I think they hired truly excellent people and a small team who made it happen.
02:23:42 ◼ ► And I think they obviously in hindsight made a very good decision to start with KHTML and go from there.
02:23:50 ◼ ► And I think the way it's played out, and as much as we can complain about WebKit today not supporting as many new standards as Blink does, I think that's philosophical, not technical.
02:24:03 ◼ ► Right? It's Apple's decision. It's not based on they can't do it. It's that they choose not to.
02:24:09 ◼ ► And the last point is, and I don't think Steve Jobs in 2002 or 2003 or whatever year it was, or 2000, I'm not quite sure when they, Ken Kishinda's Creative Selection book, he was on this show a couple hundred episodes ago to talk about the book.
02:24:28 ◼ ► He's most famous for having written the first software keyboard, the touch keyboard for iPhone, but he joined Apple to work on Safari before it shipped, and it was one of the first engineers on it.
02:24:43 ◼ ► But I don't think Steve Jobs had the iPhone in mind in 2002, but I think he had the idea of things like the iPhone in mind, where when the iPhone did come out and they were building what became known as iOS in 2006 to unveil the iPhone in 2007,
02:25:04 ◼ ► one of the, I mean, famously at the introduction, Steve Jobs said it's a breakthrough, it's a widescreen iPod with touch controls, a revolutionary cell phone, and a breakthrough internet communicator.
02:25:23 ◼ ► Well, the breakthrough internet communicator, he even said in the keynote, it's not the baby mobile web of 2006 that you could get on like a BlackBerry, right, which really wasn't the real web at all.
02:25:35 ◼ ► It was the real adult internet, like the real New York Times homepage in the keynote, loaded in a little three inch, three and a half inch screen on the phone.
02:25:45 ◼ ► Apple couldn't have done that if they didn't have WebKit, and if they wouldn't have had WebKit to build a real web browser that ran on a three and a half inch screen on a 2007 mobile processor in 2007, they wouldn't have had it if they hadn't made that decision in 2002.
02:26:06 ◼ ► And I don't think, like I said, I don't think Jobs thought, we're going to build a cell phone in five years, so we need to do this now, but I think he thought we need to own a rendering engine, the whole browser, every bit of it, and control it so that whatever we want to do in the future, we can do it, and nobody's limiting us but ourselves.
02:26:37 ◼ ► Boy, that's tough, because I would love to say no, but will enough people pay for it, no matter how enticing the features and how reasonable the price, will enough people support it to keep it going as a worthwhile business?
02:26:58 ◼ ► I don't know the answer, and if I had, I don't know what you've thought about it, I mean, sure, it's occurred to you, it's a very obvious idea to bundle Orion with Kagi, where you pay one monthly thing, like if it's $5 just to be a Kagi search user, $7 a month instead of five plus five to get Kagi plus Orion plus?
02:27:24 ◼ ► I don't know if there's a bundle opportunity there. I'd like to think it, and I totally get, if it's not free, then it has to be ads, and then ads start you down the whole path that we've been talking about for the last two and a half hours.
02:27:37 ◼ ► So I don't know. I'd like to think it's possible, and I think your prices are very reasonable, I think $5 a month is perfectly reasonable and should be successful if the browser offers enough functionality.
02:27:50 ◼ ► Do I have faith in the current mindset of the world that there's enough people who would be willing to even consider that? I don't know.
02:27:59 ◼ ► That's true. Because for my standpoint, it's important that you at least have a choice.
02:28:09 ◼ ► And I do think as much as we've got this whole internet era moving in the direction of everything becoming free, where it's not really free, you're paying with your attention, I do think we might be going through a tumultuous turning point where the mass society wakes up to that mantra of if you're not paying, you're the product, not the customer.
02:28:33 ◼ ► And that in the same way, investing in WebKit in 2002 just for the Mac, put Apple in a great position five years later to build a browser for the phone, that building out Orion now and doing this work and laying the groundwork of support and syncing things like your tabs and your history across your phone and your iPad and your Mac
02:28:58 ◼ ► in 2024 could be a very good place to be as society further awakened to the "Hey, I should just pay for the stuff I use and I use my browser eight hours a day."
02:29:17 ◼ ► Right, when you type everything you know about yourself into a box somewhere in the browser.
02:29:30 ◼ ► I think I've learned a lot. You've answered all of my questions and I wish you the, I mean this, this is not a, this episode is an interview, not a sponsorship, but I could not be rooting for both Kagi.com and Orion more hard than I am for any other products in the world because I think they're both excellent and I believe in the mission of both.
02:29:54 ◼ ► Thank you so much for giving me opportunity to tell our story. We don't get that awesome. And thanks for the going down the rabbit hole of technology.
02:30:08 ◼ ► I think that I think the fundamental problem is too many people only want to hear about a Google competitor whose plan is to put Google out of business and, and somebody whose plan is instead of taking on Hollywood by putting the major studios out of business, we want to put Disney and Columbia and Paramount out of business, but instead just wants to make really good small independent films.
02:30:40 ◼ ► That doesn't mean that those films shouldn't be paid attention to. They should be the ones, they're the ones that we remember from that era as the best films and that started the whole genre of independent filmmaking.
02:30:51 ◼ ► That's where I see there's room for independent search engines and independent browsers, even if their, their ultimate success is not bringing down the mammoth titans of the industry.
02:31:06 ◼ ► One of our advisors asked me once, what country would you like to be? What country would you like Kagi to be if you were to compare to other countries? And he asked, do you want to be like China? Go for world domination? Would you describe or? And I thought we want to be New Zealand.
02:31:38 ◼ ► And it's not about dominating the world. It's being rational, reasonable. It's still like the seventh autonomy in the world is pretty big, but it's still also a beautiful place where Peter Jackson goes to Rome, Lord of the Rings and all that. So we want to be that. And it's a long time consuming process and we are building one step at a time, trying to be reasonable and rational and create products.
02:32:07 ◼ ► Well, I would like to use, my family would like to use, and hopefully other people would like to use. And if it takes more time, so be it. I'm not in a rush. And that's what we have been doing the last six years.