The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch
00:28:35 · Jason Lemkin, Rory O'Driscoll, Harry Stebbings
[00:28:35] S0: I just think it's a hundredth the cost, but I agree with all [00:28:37] S1: of You're right. A 100... Which which... In which case... Yeah. [00:28:40] S2: We're gonna see, like, the unbundling of ChatGPT, and we're have users go, oh, well, this one [00:28:44] S0: would be No. Good [00:28:45] S1: The the the problem is not not ChatGPT, the app, but the OpenAI API and the Entropic API. You're right. This is a developer product. You as an end user, I as an end user, don't use it. You can go on and try and use it, but it's full this morning, I tried. But it's really, I mean, if you listen to the launch announcement, it's very focused on developers. And the idea is, and this is an interesting thing, almost comes back to Muse, that I think you might see, I could be wrong on this, that the needs of humans, especially consumers for AI, are gonna continue to diverge from the needs from software developers for AI. And this is a core software developer thing. You'll never need it, Harry. But someone who's building a software app might realize that a significant portion of the core intelligence they want is system one intelligence. Just tell me if this is an a or a b, you know, animal, mineral, or vegetable. Is eggs better than y? Give me a quick answer. I don't want to blather and have you tell me that's a great question, Rory, like a sycophantic LLM. Just give me the damn answer. So it's not for you, the user, but it's unbundling at the developer level where it's more likely to happen. [00:29:50] S2: You're right. I require far more intellectual answers.
00:29:50 · Harry Stebbings, Rory O'Driscoll
[00:29:50] S2: You're right. I require far more intellectual answers. [00:29:53] S1: No. You just require a bundled product. Wait. Wait. Actually, Harry, I think what you really like is the sycophantic part of the LLM when they tell you, you're so smart, Harry.
00:30:01 · Jason Lemkin
[00:30:01] S0: But to answer your question, what I did learn from using JEV all weekend long, failed on Saturday, figured out the prompts and the use case, got it to work on Sunday. So cheap it doesn't, you don't even to measure it 10 times faster is the use case, but only for these, for these, for these limited use cases. What I did learn, may almost sound tangential is man, using a harness to pick a model and getting it right is effing exhausting. This is going to get even harder and harder and harder. When should I use JEV? When should I not use it? What about... You have to run so many evals, so many tests because there's so many things that I do with JEV. You know what doesn't work in JEV? Between Harry, between Rory and Jason, who's the better person to join twenty VC as a partner? It can't answer. You'll find it's gonna get that wrong. Okay. So between that and should Rory and Jason meet for coffee, it's gonna get that one right. It's not gonna guess. It's actually going to know, hey, we're both in the Bay Area. We've known each other. We should meet for coffee. You have to QA and test every single use case to get those benefits. It's exhausting. And then I got switched over on Replit to Autorouter where switches between Astra, Fable, the open source ones and everything. Now I don't even know which one it's using. And then I saw some performance degradation. So then I had to switch it all back to Astra. Do you trust your harness? Can you QA 10,000 uses of JEV? This isn't like a bad thing, but this is like the renaissance of DevOps or something. Everyone's gonna need this massive team to optimize it. And the needs are going up with JEV and friends. Like your team's gonna have to get more. It's not just benchmarks and evals. Like we're gonna be running these 20 fourseven across countless permutations and it's good, but it's for me it's it's too much. Like I can't do it anymore. I can't pick these models anymore. I'm tapping out. These harnesses are only so like Databricks is it's easy to save money with the harness. Hey, we route 80% to open weights models and now we're doing 10% like this. It sounds great to the CFO, but I found it doesn't work for me. And a lot of folks were saying they're using JEV as an instant router. It's like, JEV, JEV, you make the decision. Here's what we're doing. You pick the model in milliseconds. Great.
00:31:56 · Jason Lemkin, Harry Stebbings, Rory O'Driscoll
But if JEV is like wrong, like 20% of the time, if you pick the wrong model for me coding a mission critical feature 20% of the time, I'm going to be trying to build this thing bug ridden all day long. Listen, smart people will figure this out. It's just getting more complicated to pick your model, not easier. And, it's it's good it's good for investors, but, it's, makes building more complicated. [00:32:15] S2: If you're Anthropic or OpenAI, what do you think the conversation is internally when they look at Jeff? [00:32:20] S1: OpenAI have become ruthlessly commercial. Their take will be like, screw it. If someone's gonna do that, we should do that. We can get something like this out in a few weeks. Let's compete and have a low cost offering too, because they're trying to make a buck. And Anthropic, they're trying to build God, and this isn't God, so why would they even bother? In fact, the launch video for JEV was prod not God, right? So they're deliberately saying not God. And if you talk to Anthropic, the mission is AGI slash god. So they're like, you are just a minor nothing thing. Why would I even deign to sully my hands on vulgar commerce?
00:32:56 · Jason Lemkin
[00:32:56] S0: I think that Anthropic and OpenAI have made a very strategic decision. They will not play in this market for now. They have... OpenAI and Anthropic each have the two worst models that exist. They're called OpenAI Mini and Haiku. They're terrible. They're terrible. Now, if you ask Claude, what should I use for simple workflows? Haiku. And if you ask ChatGPT, it will tell you Mini is great. It's so cheap. And at least my little evals, they fail a 100% of the time. Should Rory and Jason have coffee? Yes. Send them to London. Worst answer... I mean, I'm exaggerating. These are the worst models I've ever used, Mini and Haiku. And so they've decided to launch crippled models that they can lightly promote, but they don't want to play in these low margin businesses for now. They could build this, right? They could build, I think they could build a version of Java in an hour or over the weekend, but they've intentionally decided to cripple the low end of the market. And I think they're puckered about the open weights models that just got a little bit closer earlier than they planned. Right. Because that is more of a threat than abandoning the bottom of the market. This is just my view, but I don't think they've ever taken the bottom of the market seriously. They have checked the box offerings that no serious developer uses, I don't think. Go on Claude and switch to Haiku if you can and ask it a question. You... It it it doesn't remember anything since 2023, and it doesn't know know where anything is. And that
00:34:10 · Rory O'Driscoll, Harry Stebbings, Jason Lemkin
[00:34:10] S1: in a nutshell, it is the JEV opportunity. It's not quite one for one, but, yeah, you're right. [00:34:15] S2: I'm I'm going slightly off on a tangent on this, but people do like it when we talk about Venture. It was a $40,000,000 seed round for JEV. And, you know, I just had a conversation with my team right before this, and they're like, dude, we never see anything less than a $20,000,000 first raise for anything anymore. The seed rounds, minimum are, like, eight to 10 with someone spinning out of a good company. Are we seeing the evaporating of traditional seed? [00:34:39] S0: Why do you think Andreessen just did a university? They need to go pre inception. You think I'm kidding? No. Mean, finally someone figured out that Peter Thiel got this right eighteen years ago at the Thiel Fellowship and the dorm room fund, which sort of did cursor. You got to go pre inception if you want to do seed now. Inception Inception's too expensive. You got to go pre inception. But Andreessen Horowitz did launch Horowitz Andreessen University today with $40,000,000 to do pre inception investing, right? To do the Thiel Fellowship on steroids. I think it's a very rational response to your point. But I do think Peter Thiel saw this space before anybody did and executed. The only thing he didn't want to do was scale it up. He didn't want to build a Thiel Rabois University like Horowitz and Dreeson. He just didn't want to scale it It could have been. I mean, they have the best people in the Thiel fellowship. They get the best... Today too, they get the best people.