George Sivulka

Hebbia has actually pioneered something different where a year, almost 18 months ago, we said, hey, we can't wait for these models to catch up. What we'll do is infer simple, single question. Let's actually run hundreds or even thousands of submodels of the best models in the world to compute over every single document to answer the same question.

Hebbia has actually pioneered something different where a year, almost 18 months ago, we said, hey, we can't wait for these models to catch up. What we'll do is infer simple, single question. Let's actually run hundreds or even thousands of submodels of the best models in the world to compute over every single document to answer the same question.

Hebbia has actually pioneered something different where a year, almost 18 months ago, we said, hey, we can't wait for these models to catch up. What we'll do is infer simple, single question. Let's actually run hundreds or even thousands of submodels of the best models in the world to compute over every single document to answer the same question.

And so ultimately, if you can't train larger and larger models fast enough, you could take whatever state of the art or cutting edge and run it more times to get more compute, i.e. more computational power, better decision making for the same user right now. And so this is an idea that we pioneered. It doesn't matter if you're using Claude 3.5 or if you're using O1 itself, i.e.

And so ultimately, if you can't train larger and larger models fast enough, you could take whatever state of the art or cutting edge and run it more times to get more compute, i.e. more computational power, better decision making for the same user right now. And so this is an idea that we pioneered. It doesn't matter if you're using Claude 3.5 or if you're using O1 itself, i.e.

And so ultimately, if you can't train larger and larger models fast enough, you could take whatever state of the art or cutting edge and run it more times to get more compute, i.e. more computational power, better decision making for the same user right now. And so this is an idea that we pioneered. It doesn't matter if you're using Claude 3.5 or if you're using O1 itself, i.e.

scaling at inference at the orchestration layer with something that was scaled at inference with the training layer, but you get way better results. And it's a way to

scaling at inference at the orchestration layer with something that was scaled at inference with the training layer, but you get way better results. And it's a way to

scaling at inference at the orchestration layer with something that was scaled at inference with the training layer, but you get way better results. And it's a way to

We've seen that for certain types of documents, like the dense legalese or more colloquial documents, Anthropic works better. But for other types of documents, like OWAN or OpenAI 4.0 works better. And it's always trade-offs between accuracy and speed and all kinds of things. Actually, a lot of the time when we're decomposing a task, we'll use mixes of OpenAI, Anthropic, even Gemini.

We've seen that for certain types of documents, like the dense legalese or more colloquial documents, Anthropic works better. But for other types of documents, like OWAN or OpenAI 4.0 works better. And it's always trade-offs between accuracy and speed and all kinds of things. Actually, a lot of the time when we're decomposing a task, we'll use mixes of OpenAI, Anthropic, even Gemini.

We've seen that for certain types of documents, like the dense legalese or more colloquial documents, Anthropic works better. But for other types of documents, like OWAN or OpenAI 4.0 works better. And it's always trade-offs between accuracy and speed and all kinds of things. Actually, a lot of the time when we're decomposing a task, we'll use mixes of OpenAI, Anthropic, even Gemini.

This makes me think of the story of Bloomberg, which has the best financial services training set of all time. And they trained a GPT 3.5 class model. It was called a Bloomberg GPT. And they released an archive paper and everyone on LinkedIn was like, wow, Bloomberg is, you know, cutting edge and they're going to steal finance. Why did they not?

This makes me think of the story of Bloomberg, which has the best financial services training set of all time. And they trained a GPT 3.5 class model. It was called a Bloomberg GPT. And they released an archive paper and everyone on LinkedIn was like, wow, Bloomberg is, you know, cutting edge and they're going to steal finance. Why did they not?

This makes me think of the story of Bloomberg, which has the best financial services training set of all time. And they trained a GPT 3.5 class model. It was called a Bloomberg GPT. And they released an archive paper and everyone on LinkedIn was like, wow, Bloomberg is, you know, cutting edge and they're going to steal finance. Why did they not?

They did have all they've got the best data in finance. So then GPT-4 was released, I think, like a few weeks later. I don't know exactly the right timeline, but it just destroyed Bloomberg GPT at every single finance task. And so you saw the idea of post-training or kind of like this refined, verticalized model creation just always would lose to scaling loss.

They did have all they've got the best data in finance. So then GPT-4 was released, I think, like a few weeks later. I don't know exactly the right timeline, but it just destroyed Bloomberg GPT at every single finance task. And so you saw the idea of post-training or kind of like this refined, verticalized model creation just always would lose to scaling loss.

They did have all they've got the best data in finance. So then GPT-4 was released, I think, like a few weeks later. I don't know exactly the right timeline, but it just destroyed Bloomberg GPT at every single finance task. And so you saw the idea of post-training or kind of like this refined, verticalized model creation just always would lose to scaling loss.

And maybe we're at the end of scaling law as a training. But I actually think, you know, Hebbia and now OpenAI and a variety of other companies are starting to pioneer the idea of scaling laws and inference. And I actually think that it will make nothing that that other players can do to fine tune models will ever catch up.

And maybe we're at the end of scaling law as a training. But I actually think, you know, Hebbia and now OpenAI and a variety of other companies are starting to pioneer the idea of scaling laws and inference. And I actually think that it will make nothing that that other players can do to fine tune models will ever catch up.