(Preview) All About Agents: Dots Arrive, Google and Apple Questions, Muse Follow-Ups, Super Intelligence, and the Mandate of Heaven

Welcome and a tech terminology lesson

0:00Hello, and welcome to a free preview of Sharp Tech.

0:09Hello, and welcome back to another episode of Sharp Tech. I'm Andrew Sharp, and on the other line, Ben Thompson. Ben, how you doing? I just had an amusing moment. You had to drop a couple of hellos. One of them was my fault. I wasn't quite ready. The second one was a little too enthusiastic. It was more a greatest of all talks hello, which, by the way, I always laugh because my kids would always imitate it whenever I'd say your name. They would just drop into a long extended hello. And come to find out, all your listeners do this. You've got multiple emails of kids imitating the Andrew hello.

0:47Anyhow, you said after you did it too enthusiastically that you had too much vim. Too much vim, yes. I had to crank it down a little bit. Did you know that this is very funny because this is reestablishing our normie versus nerd credentials. Did you know that there is a vim? Vim has a tech reference? Oh, no. What does vim stand for? I'm still a little bit ashamed of the VPN, VPS mistake on the last episode. Oh, I think that's fine. That was also very funny. No, vim is a text editor. And there's like famous wars that are used like Unix type systems. And there's like vim people and there's like Emacs people. Emacs is like the ultimate nerd out sort of you can configure literally everything.

1:31And people have these massive configuration files that they carry around their whole life and they like can't function without it. So the nerd chads, they're all Emacs people. Is that right? I think Emacs is going to be more on the nerd side of the spectrum. Vim a little less. I don't actually use any of these. If I have to edit text in a terminal, I use nano, which is like just the bare bones sort of thing. So I'm actually out of my depth as well, but it did amuse me. Vim, there you go. That's your word of the day. The more you know, you know, the Emacs chads versus the vim virgins out there in technology.

2:03No, look, if you know what vim is, you're already off the reservation. So it doesn't matter. Everyone is a chad nerd in that scenario.

OpenAI announcements and domain name drama

2:10Well, Ben, we are going to be talking agents today, and I want to start with a note that I saw earlier this week. OpenAI famously announced Dots on Tuesday. And at some point, either before the announcement or immediately afterward, SpaceX AI bought Dot.com. And Dot.com now redirects straight to the GrokBot download page. And I highlight that sequence first because I find it delightful.

2:41And second, just to step back and celebrate the Wild West nature of the ecosystem right now, it's the best. That was going to be my response to this. I saw the rundown. This sort of stuff used to happen all the time, like just like companies doing practical jokes on each other. There's a whole thing about I think Salesforce was actually like they would do all these crazy stunts at their keynotes. I should have like come up with some examples before of just like making fun of their competitors. That was like way over the top and arguably over the line, actually, almost certainly.

3:14And that used to be the way tech operated all the time. And you're right. It's good to have it back. Dot.com goes to a different spot. I feel like we completely missed that generation of technology. I certainly did. Yeah, there was no podcast period back then, which is a real shame because there's. But then again, maybe that's why it all happened, because there wasn't enough documentation going on. And, you know, that's right. You can get away with mind stepping here and there. Well, this is all a thousand times more fun than dealing with a bunch of monopolies who have their own little fiefdom.

3:45So thank you to all of our participants in the AI wars. That's right. Frontier wars. That's right. As for dots and OpenAI Dev Day, I will say at the top, my feelings after consuming the content from the OpenAI event are close to the exact opposite of my feelings after watching Meta Connect last week. So I'm curious for your thoughts. We'll start with an email from Steven. Ben and Andrew, I can't wait for Ben's rant on how OpenAI has 90 percent of consumer mindshare when it comes to AI, but somehow completely fumbled the dots launch to feature SF coded techno parents launching their startups and side hustle companies as the primary use case.

4:32Rather than going hard into family scheduling, kids activities, catching up with friends, taking care of parents, groceries, et cetera. It even came with dystopian fake house aesthetics that feel like a Black Mirror episode, presumably to ensure we really get the point that the real world and real people's lives are an afterthought. And all that matters is the techno grind lifestyle. What do you think of those sentiments from Steven?

Consumer strategy and the state of tech

4:59It's interesting because I think Steven made me realize something that's sort of just been in the back of my head, which is I've given up on OpenAI in the consumer market ages ago. Like, it's just I just I think at some point I think it was when the Gemini looked like they were, you know, Gemini three looked amazing. And I'm just like, they don't have an advertising model. Even if they watch it now, they're going to like they're swimming upstream. It's going to be tough. And I think I just sort of gave up at that point. It is what it is. They missed.

5:30They had the chance. They missed it. Now, it sounds like their advertising product is doing OK by all accounts. I think this relationship with Amazon is very interesting. There's a bit where it's very much in Amazon's interest that OpenAI wins because OpenAI is a surface for Amazon ads. That is a way for Amazon to deliver on sort of the external agentic. Yes, you can have a great experience. Look at OpenAI accesses or products. No problem. They also show our ads. That's the trade. Doesn't kill our ads business. That's right. Sure.

6:01And so so maybe so maybe there's maybe I'm like I've jumped to conclusions too quickly. And so this didn't really this didn't really bother me because it's what I expect. And I think like the big change, if anything, is the extent to which Meta and Muse have walked to do the opportunity and Google has. Right. I feel like Meta has highlighted a consumer opportunity that other and including on this podcast, we haven't really talked about the consumer space for like nine months on Sharp Tech.

6:31Well, the other two was the sort of the chat GPT app reboot, which made it unbelievably more complicated in, I think, an attempt to introduce people to what is possible and capable. Yeah. But along those lines, it was still kind of jarring for them to introduce dots by basically mocking the use cases. Oh, of course, you could use it to check your school email, but you could also do this. And and it's funny because I actually think there is a lot to mock about the use cases being put forward for consumer agents.

7:08To be clear, everyone is terrible at this, but my prescription is not that you highlight how you can embrace the grind lifestyle. My prescription is actually you have to go the exact opposite direction and market these as the modern day Tylenol, the modern day Advil. It's pain relief. There's all this stuff in your life that is painful that agents can make easier and better. And it helps that I actually think that is true. Like the thing I didn't really put in my Monday article, I mentioned it in sort of a follow up on Tuesday, I think, is the reason I'm optimistic about agents is not because I'm a nerd who thinks agents are fun and cool.

7:50That part is all true, to be clear. The reason I'm optimistic is because I think tech sucks. It sucks for normal people. It's so hard to use. It's apps on an iPhone were way better than what came before. But there is a mindset that sort of looks at the world as apps were so much better than before. Therefore, apps are the best experience. They're not. And actually, if anything, apps have gotten way harder to use.

8:20I was going to say, there was probably a point in time where they were a much better experience. At this point, I think you mentioned you have like 700 apps on your phone. That's true for everybody. They have so many apps. They don't even know where they are. And you just search and find one. And they're all kind of a pain in the ass to use these days. Well, especially things like security, like just keeping track of logins and then two factor authentication and stuff doesn't stay logged in. And you have to do it all over again, like migrating to like the one of the reasons I bailed on the max.

8:52I didn't like I was like, I should try it for at least a week to see if I can get used to the size. And I got halfway through migrating. And it's like, this is so painful that if I go through with this and I don't like it, I'm going to have to do the same thing all over again in the opposite direction. And I know I'm done. I don't want to grind through it. Yeah. It's funny because I've been like with a Mac for a long time. Like I, I knew the big price X were coming and I was telling everyone, if you need a computer, buy a computer.

9:22I did not buy a computer because I'm like, mine's fine. Don't put me through that. I don't absolutely have to go through that. Fair enough. Well, and as far as dots is concerned, it's interesting because I had the experience of introducing my wife to Muse last weekend because I thought it would be more useful for her than it has been for me. And her first response is, why do I need that? I use ChatGPT for all of my personal organization and AI work.

9:52And I said, just try it. And she really liked it the first couple of days. So she's now a Muse user. And I understand from OpenAI's perspective, the best market for them is still enterprise works. And so pointing dots in the productivity direction actually makes sense. Like the consumer facing agent category is still mostly theoretical. So this could be a reflection of continued strategic coherence on OpenAI's part. However, watching the event, like the visual language around dots and, of course, the lowercase d for dots on OpenAI trademark.

10:31Really? Very groan-worthy stuff from them as ever. But like the cuddly little creatures that are just a straight ripoff of the Muse character. I think that's unfair. I mean, these are, I'm assuming these were designed a while ago. But maybe, look, all I'm saying is visually it is completely incoherent to me because it looks like a product that's designed exclusively for mass market consumers. But that's not really what the product is. And so that was part of what was confusing to me this week.

11:03I'm not sure what the product is.

Managing projects and evaluating Dots

11:05Like the way I, you know, I'm going to have to like re-update an update. So there is a clear need in code, or at least that I've found. This is probably one of those things where I was over-extrapolating from my sort of experience. But from what I find, I like the chat, the codex sort of organization where you basically have projects that align to files on your computer. You can see it's like it grew up, like connected to the local computer, which is a big part of all the problem of why they're smashing all these different products together.

11:37It feels so weird, but you have like projects that align to like folders, and in those you have different threads. And part of it is, I think people don't realize that like in general, one massive thread is bad. Like the bigger it gets, the more context it gets, the, and this was sort of a. The more confounded the reasoning becomes. It just gets super confused because they're like, you load instructions into it at the beginning. And those are so far back in context, there was a big push early on in like the chat GPT era.

12:08Oh, we need larger and larger context. Context is the problem. Oh, it's a million token context. This is incredible. Actually, no, that sucks. The, this is where I was trying to get out with the right things down. The way they actually function well is continually writing things down. And the way they keep track of stuff that's written down is by having relatively fresh context and then loading stuff in as it's sort of appropriate. And managing this is actually really, really interesting. I mean, this has been part of the whole thing I've done with this, this geckobot thing and just getting a much more fine-tuned understanding of how this works. But by and large, doing a thread per task is better than having this massive sort of ongoing thing.

12:43Anyhow, Codex actually aligns this pretty well. And Codex has been, is really good about spinning up new threads. Like there's one, I'm like, oh, let's redo this with this direction. Let's throw this away. And instead of trying to like continue and remember all the, all the stuff it did that was wrong, which I talked about, like, this is an issue here. It just starts a new thread. Like it can do it itself. Anyhow, this is all great. I, if I know exactly what I'm working on, the Codex organization makes much more sense to me than Claude, which is, it's, it's much less intuitive to me as far as organizing a product.

13:17I don't want to work on a project that I'm, that I'm doing. Claude, on the other hand, manages context much better. So there's a bit where like one long quad thread just keeps track of itself. I feel better than sort of Codex does. And the way this worked out for me is I actually, because I have my agents write everything down because I want to be able to use any agent. And I don't want stuff inside, like inside their weird folders, dot Codex or dot Quad or whatever it might be. I want it out. I don't want to be locked in. I'm just allergic to lock in the way I've always operated away. I'll use text files.

13:47And I just use, I basically use Claude as sort of a chief of staff. It kind of knows everything that I'm working on. I have like this server that just came in and I also have a bunch of cameras that I'm trying to set up. And like all these different things, I can just sort of bounce around and I use this sort of quad thread to sort of keep track of it. And it's writing it all down. And then when I'm actually working specifically on a project, I find it easier working Codex. So Claude is the barrel and Codex is the ammo as far as executing on your side. I mean, I have, that's a great analogy.

14:19I think that that is sort of the case. And so I see dots come out like that's, yes, that is so needed in Codex. An overarching sort of agent that can be my liaison to all my other agents that are working. So conceptually, this makes a ton of sense. So even starting there, that was my assumption around dots. That is so heavily nerd coded, right? It's like, let me manage all my AI building projects. How big is the market really for that sort of thing?

14:49Sure. So here's the problem. Dots kind of sucks. And the reason it sucks for me in this use case, it has, it seems to not know anything about what I'm working on. It immediately like started looking for my calendar. Like, so I have a plugin to pull my calendar in for Gecko. So it found that. And it's like, I noticed you have a conflict on October 14th. Do you want me to help you resolve that? Number one, I'm like, how did you know that? That was weird. And of course I figured out how it did. But number two, like this is so far, I don't know. That's not what I need help with. That's not what I'm signing up for. I need help managing all these projects that I have.

15:20And, and so, and then I'm like, oh, and it starts asking me for permissions to everything. I'm like, I set, I made my own custom permissions file for codex. So it knows when it has to prompt me to approve something. And when it doesn't, cause I want a little bit, I want fairly loose. Cause I have a dedicated computer just for this. I don't want total full access. So I actually made my whole thing. It didn't know where that was. It didn't know about it. And I had to like direct it to the file and then it found it. I'm like, oh, I'm working on this and this file. It's like, I don't know what you're talking about. And it's like, it's weirdly, it's weirdly neutered and like unknowledgeable about stuff.

15:56And so, and then, but again, like I said, it latched onto this calendar thing. It's weird. It's just a, it's, I see a hole here in the codex offering, but then to your point, everywhere in the media, everyone's looking at, it's like, oh, they're Muse competitor. And it's like, I guess so, but it's just, it, it's not really that it's not as good as Muse. It doesn't sound like it's solving your problems either. Yeah. It's just sort of a confused and confusing product for the time being.

16:26And of course now there's dots, spaces, pages, local slash cloud chat GPT. So the spaces thing is a great idea, by the way. So this is a, this is a, so the way I backed into GeckoBot is I actually created my own page that just listed all the stuff that I was working on. And, you know, I sort of developed that more. The biggest thing is like basically cards. So any item that I get reminded by Gecko, I click on it and it has all the history of that and the item and any, like there's attachments to it or like, for example, I'm editing an interview now, Duman just, he does the initial transcript and then he posts it.

17:00And then instead of me trying to find where the link is or whatever, it's just, I just, it's in the card. I will click it. I download the text file. I can edit it. It's great. Like the, so this idea of having some sort of substantiation place where you could actually see the stuff you're working on in a way that makes sense. And you can, I actually have UI on that page and it's all fake UI because the UI is just prompts. You click it and like, say you click done. It's not, there's no machinery in the page that is going to the database, clicking done. It's firing off a prompt to Gecko saying this is done.

17:30And then Gecko actually does all the bookkeeping. So this whole concept of a space now, obviously they have way more resources, way more capabilities, what it looks like you can do with those spaces looks great. So I think like, and this is the frustration. There's all these ideas that make sense and that are super cool. And should exist. Right. Yeah. It's just, it's such a, there's no organizing principle here. And I think it is all downstream to bring this full circle to the emailer of they had chat GPT.

18:02They decided we have to get money from the enterprise and we're just still living in sort of the shipwreck that is downstream from that. Everything's great. And it all sucks because you don't know what's going on. And yeah, I think the, the, the Cardinal sin was the, the super app pushing this stuff to in the long run. You can see why all this should be together. Like dot should be the ultimate manifestation of chat GPT. You shouldn't have a, but even a chat GPT, it's kind of weird.

18:32You had a bunch of different threads. It's be, it'd be easier just to have one thread, right? You just talk to your agent and it answers your question. And that's true. Sure. That's what dot is. I think ultimately going for, but this app wasn't ready. And then you can imagine, oh, you have, it goes off and spins off new projects for normal people. That means you have to have cloud compute. It has to be just like meta did with muse. You have to have the VM in there. And they were so eager. I think they had the right vision, but they'd had no patience.

19:07Yeah. It wasn't ready to be shoved together. When they shoved it together, you had codecs, which worked on your local computer. You had chat GPT work, which had its VM and you had chat GPT, which was sort of a stateless sort of thing that it's, it's synced between devices, but every, every conversation is fresh. There was no sort of ongoing environment. Putting those three things together didn't make sense architecturally. It didn't make sense conceptually, even if in the long run, those should be one product. And yeah, it's a lack of patience. It really, really is where they're going makes sense, but they're not there yet.

19:40And by accelerating, it makes everything harder to use and confusing and a mess, including with dots. They don't have enough compute to serve dots to more than just the pro tier subscribers right now, but dots, it could be a product that's actually really useful to more people. It might be telling that dots is exposed in the same part of the chat GPT app as the rest of chat GPT, whereas codecs is its own tab. By the way, why both Anthropic and OpenAI, why is there not a dedicated cloud code app and

20:13a dedicated codecs app? This having to go into a little thing to get to your thing, it just drives me by, there's no reason for it. It's weird. It's like everything has to be in one app. Does it? Like, I understand the lots of people have this app already installed. Take advantage and put more capability in there for the people who might not discover it otherwise. Yeah, but it just makes for a really crappy user experience, which to go super meta, this is all of tech.

20:44The user experience all sucks everywhere. It really does. And like, this is why I've ultimately, what the actual answer for all this, and again, I think this is where dot is going in the long run, is you don't need to know any of the complexity. You just ask your agent. You say, I want to, instead of operating at a level of, I need to know how to do something, you just need to know what you need to do, and the agent will take care of the details. And I absolutely do think in the long run, and probably sooner than people think, the experience of using an agent is going to be so much better than using a bunch of different

21:21apps and websites that this paradigm is going to win. It's not going to win because nerds want it to win. It's not going to win because the economy needs it to win. It's going to win because it's going to be better. It's such a good take. Well, it's going to win because the rest of tech sucks, as you said 20 minutes ago. And that's something that millions and millions of people identify with.

Consumer agents as a winner-take-all market

21:44Well, speaking of the agent future here, Andrew says, not me, hi, Andrew and Ben. I'm halfway through Thursday's episode, and I can't wait to listen to the rest. But you've been discussing why AI agents for consumers will be or should be a winner-take-all market. But I feel like you aren't addressing my main concern with that argument. More consumer demand doesn't make the product itself any better, and the switching costs aren't all that significant structurally. So why do you think this is a winner-take-all market?

22:15Any thoughts on that, Ben? All right, and that is the end of the free preview. If you'd like to hear more from Ben and I, there are links to subscribe in the show notes, or you can also go to sharptech.fm. Either option will get you access to a personalized feed that has all the shows we do every week, plus lots more great content from Stratechery and the Stratechery Plus bundle. Check it out. And if you've got feedback, please email us at email at sharptech.fm. Thank you. Bye.