Ben - [C/x] on Warpcast

Content pfp

0 reply

0 recast

0 reaction

July pfp

a crazy idea: the fact that most (if not close to all) models, papers, are all going to be obsolete in a few years even though they feel so new today

7 replies

2 recasts

24 reactions

Kyle Mathews pfp

well or they won't. Progress isn't guaranteed without new conceptual breakthroughs. I don't really see any improvements coming until models get embodied (i.e. exist as robots so can learn from reality)

1 reply

0 recast

1 reaction

Ben - [C/x] pfp

So you think we’re near the end of this cycle?

1 reply

0 recast

1 reaction

Kyle Mathews pfp

quite possibly. All the new models released in the last year have been somewhere between gpt 3 & 4. Lots to do w/ evals, fine-tuning, training on propritary data, efficient inference, etc. but "intelligence" of models seems like it's topped out — see https://warpcast.com/kam/0x1dd60ed3 & linked @vgr post

1 reply

0 recast

0 reaction

July pfp

Agree with you on this, we’ll probably need a step change akin to CNNs did for deep learning that transformers have done for LLMs

2 replies

0 recast

1 reaction

Kyle Mathews pfp

the thing is though that GPT4 is very very good — if LLMs are in case just an extreme type of data mining — then GPT4 is perhaps near the limits of what you can mine from human language.

1 reply

0 recast

0 reaction

Ben - [C/x] pfp

How long until we get a model that can interpret structures data and equations? (i.e., finance)

2 replies

0 recast

1 reaction

//trip pfp

Plenty of specialized models out there that do this. Unsure exactly what natural language (LLMs) add besides a query layer? Crunch can already happen

1 reply

0 recast

1 reaction

Ben - [C/x] pfp

Can you share an example? To clarify: models that can interpret financial statements

1 reply

0 recast

0 reaction

//trip pfp

I can't as it's not my field but friends at big shops all talk about the algos and models they run. It's why earnings are so unrelated to the actual numbers lol. OCR hits statements as they are published and it's pushed thru analysis pipeline that has different models tuned to different intents etc etc. I hate it 😅

0 reply

0 recast

1 reaction