Open Source Models

“Today we can feel a little bit safe because OpenAI and Anthropic have this and no one else does, but there's open source models. They're six months behind.”

“I think calling open source models Chinese models is a psyop by the frontier labs to basically trick people into thinking that they're scary and otherize them.”

“And if you're looking at a Frontier model with double dollar digit cost per token, and you're looking at an open source model that's 10, 11 cents per token, while all tokens aren't created equal, it's still enough of a difference that there's going to be a mas…”

“If you just compare open source models Versus the closed frontier models. There is no way for us to understand what the thinking tokens are, meaning you give it a prompt, you get back something, but all the stuff in the middle is obfuscated. Clot obfuscates it…”

“When you take a closed source model, it does really well on a benchmark. When you wrap it in its own harness, it decays in capability. When you take an open source model, And you use any other open source model, it improves in capability.”

“You would say, you know, six months, 12 months, maybe at the most between the closed wait frontier and available open source model.”

“The open source models will very soon, if not already become capable of lending meaningful assistance to destructive uses that some people might pursue already cyber offensive capabilities has been a concern, right? With mythos, for example, that was withheld …”

“And actually, if you look at the amount of value captured, it's actually the closed source frontier models. You know, the whole pie is growing, but it's growing so fast that even as the open source models like share expands, the frontier models are still able …”

“We don't have any mythos level open source model for sure, but we definitely have models that are not super far from the opus category or depending also it's more spiky.”

“In OpenSoup Mobile, you can download it. Nobody can, like no country could take it out from you once you download it. You can find it yourself. If you operate it on, on your data center, like local ground data center, I feel like you start to have the beginnin…”

“At any given point in time, including now, I think models from open AI and Anthropic are kind of more widely used and more critical kind of in general than open source models.”

“I think in two, three months from now, open source have distilled all these capabilities, and we'll find open source models out there, which you can find to you if you're an attacker, to actually do this on a task basis.”

“You can, in a lot of cases, get slightly better outcomes at half the cost. But again, that half the cost, I think a lot of people hear that, they're like, that's bad for AI demand. It's actually not at all because the cost the user pays has, you know, is just …”

“What if the other side that's interacting with the chat bot can, you know, build in a certain code word or a certain like character sequence that then jail breaks that model and gets it to reveal all the data to me. So it can sort of like vomit out all of the …”

“We've seen some customers maybe six months ago or a year ago, they'll take an open source model and fine tune it or something. And then, you know, they're now like, okay, this is like, shit, this is like really behind and it's like nightmare to maintain and I …”

“I think they'll get there, but it'll probably take longer than people think because, you know, even with our team, fine tuning these models is non-trivial.”

“But I think the point is that at a certain point, it's strictly better to use open source models because when your use case is solidified and you're in production at scale and you're pretty sure this is the sort of shape of the agent, then there's no reason no…”

“We don't just build our set of open source models and then, you know, it's done, we can move on to our next thing, and maybe we'll revisit this in two years. You often need to train new models all the time because As the frontier changes, as the capability of …”

“When we have open source open source models that we want to fine tune, we find that if we can clearly tailor our evals to customer outcomes, it's way better than just looking at, like, loss curves over time, right?”

“If we ran our entire business exclusively on frontier models, I would care a lot about costs, and I would think a lot about that, but once you're already, once you've already made the jump to saying, okay, now, now we know how to think about open source models…”

“When I speak to the mate who CFO, which is a leading like creative AI company he was saying, look, we don't really need the most like the biggest model. We don't need the best model in that sense, but we need the best model fine tune for our use case. So we ca…”

“And there will be clearly an important place for open source models in the world and people that will want their own weights for all sorts of reason, the ability to modify those.”

“I think the foundation models, let's say specifically Anthropic, have such special models. This can be hard for somebody to just kind of say, I've used open source with my data. It's going to be functional and positive for some amount of what you're doing, but…”

“So I think, you know, you're going to have the combination of something like that, a family of models from Anthropic, and then a set of open source models and things that you train with your own data, and you're going to look across that whole tapestry and say…”

“I wouldn't say that open source model improvements cannibalize our core business because data is most valuable on the frontier of model performance, so each of our customers has their own unique goals and is purchasing eval and training data sets To fill gaps …”

“I think where it's all leading to is an effort to ban open source models.”

“You see this rhetoric already in Anthropix blog posts. Any threat that they describe, they kind of go out of their way to take that shot at open source models.”

“Well, look, I mean, I have it on good authority that there is no decision by the white house to ban open source models.”

“If Android represents all open source models, open weights, et cetera, iOS represents frontier labs. I feel like we're going to head down a path with a similar breakdown in the U S and globally. We're like three quarters of the world are going to choose open s…”

“Well, I think open source models are going to dominate AI inferencing. You know, you'll see in the next two years, we'll shift to where's almost no open source to where's going to be almost all open source.”

“Enterprise LLM spend was 19% open source last year, and it's 11% this year. So it's dropped.”

“And the open source models are less than two months behind.”

“You know, the open source models are really like six months behind and they're like 99% cheaper or more sometimes for inference.”

“From a point of view of open source and using the model, everybody's going to be fine. The question is going to be, are they okay with the Chinese model or not? That's the only question here. It's not open source versus closed source.”

“Well, we've been telling customers, I believe that majority, majority of enterprise workloads will actually be on open source models in three years for sure.”

“I think at first people opened up and said, everybody, as much tokens as you want. And in enterprises, there's no open loop. We don't give sort of any resource unconstrained to people. And now we're jumping on and saying, whoa, all right, these guys should hav…”

“In that case, like, you have to use open-source models, because none of the major model companies, or at least OpenAnthropic, don't really do fine-tuning anymore.”

“I think for example, LM inference, thousands more companies are going to post train their own models and deploy open source models for inference.”

“In that world, enterprises and individuals are probably optimizing for a couple of things. I think they're probably optimizing for cost, right? They're starting to think about the trade-offs between intelligence and spend and how they can optimize their token …”

“They're awesome for the AI infrastructure providers. They just shift economic value from the margins of the frontier labs to the infrastructure.”

“The open source models are about three to three to six months behind the frontier models, right? But they're 99% cheaper for inference.”

“We're using OpenAI. We're using Anthropic. We're going to increasingly use open source models to save money, but also because we don't today have a budget constraint on our Anthropic contract.”

“Outside of China, there are basically no good open source models.”

“Increasingly, it's not just which model is the best. It's which model can solve the task at the lowest price point, because as these models get better, there's kind of intelligent saturation where it's like, okay, for some simple tasks, I actually want to know…”

“There's customers like us, where we do need to move some of our traffic to open source models because it's just becoming super expensive to serve like the largest closed frontier model.”

“We are going to start publishing some of our research that we're doing with the labs or the other providers and open sourcing more models, more of the work we're doing.”

“We're training our own models post training it on top of amazing open source models, and that will bring down the cost that we currently spend on frontier model tokens. We expect to continue to use frontier models for designing new experiences and new capabili…”

“Probably, like, 80 to 90%. It's typically the planning that really needs the frontier models.”

“The majority of economic value may continue to accrue to the frontier, and man has it ever accrued to the frontier thus far, and for sure the first six months this year, but the majority of tokens consumed in the world may be open source.”

“In AI, if Anthropix sort of hits a wall and stops improving, or open AI, then the open source models will catch up. And then it might be a race to the bottom, and it might be, you know, it won't be good for the stocks, probably. It could be good for the chip c…”

“In terms of the red capture agenda that you're seeing in Washington, I think where it's all leading to is an effort to ban open source models or open weight models.”

“We're also seeing founders that are running massive agentic stacks, and they're swapping out for open source models, like left, right, and sideways, because it's just like unsustainable for their margins. Like they just cannot pay these luxury prices.”

“You can have basically an inexpensive computer running, you know, what was a state of the art model year and a half ago, running it locally on a device with an open source model, which also means that the Chinese can have it, the Russians can have it, the Nort…”

“Like all the open source models, I think what happened was they got like very hyped and people were very interested in using them. But I think like over time, like there was a spike in usage for these models. And then it goes back to open AI, Anthropic and Goo…”

“Every entrepreneur in my portfolio is like swapping out open source models for everything. And of course you would. Right. Cause that's your margin.”

“The open source models have crossed some sort of chasm in terms of their baseline. Capability, and then I think RL techniques and post-training is for specialized models has become mainstream enough, and, you know, there's enough examples of its work, of it wo…”

“A lot of people were talking last year about, like, oh, open source models are gonna take over, like, even on the AI voices space, like, you're gonna get commoditized and so on. And for a period of time, I kind of believed it. I was like, yeah, I mean, There i…”

“We're like, yeah, we could go with open source models. Excellent. Do it. You try to operationalize it and they've lost three months and then they come back.”

“I've talked to a lot of big companies, 65 to 70% of their token consumption is open source model, right?”

“I think increasingly you know, what you're gonna see is the open source models are probably one step back from the absolute frontier. You know, it usually takes about six months for the open source community to sort of re-implement and figure out what those id…”