Is LeCun right that open-source AI will soon become 'unbeatable'? (EOY 2025)

On Oct 14 2023, Yann LeCun (Chief AI Scientist at Meta) stated: "Open source AI models will soon become unbeatable. Period."

Resolves YES if, at the end of 2025, it's decisively clear (in the judgment of Eliezer Yudkowsky) that open-source LLMs (or their successors in the role of widely used AGI tech) are more powerful or more cost-efficient than their closed-source alternatives. That is, if either all the leaderboards are full of open-source LLMs with successors to GPT-4 or Claude being far behind, or if most of the business spending for inference seems to be on running AI models built on open-source foundation models, this resolves YES.

If it's hard to tell or if that seems wrong, resolves NO. "Unbeatable" seems like it shouldn't be subtle.

Get Ṁ600 play money
Sort by:
predicts YES

GPT-3.5 class model with longer context size is offered at 1/3 of gpt-3.5-turbo price

bought Ṁ250 of YES

predicts YES

@WieDan How is Gemini "open-source AI"?

predicts YES

@osmarks it's obviously not, but Gemini Ultra being barely better than GPT-4 arguably shows that having enormous resources and data doesn't really give you that much of an edge.

predicts YES

@BairAiushin My reading of that is just that Google is bizarrely incompetent some of the time.

predicts YES

@BairAiushin yeah i think we're deep in the s curve

@osmarks Google had an extra 9 months over GPT-4 + some open source knowledge, and still failed to make the gains we expected, even with delaying Ultra. This definitely is good news for YES betters here. Plus, open source is getting closer to gpt-4. I'm updating a bit towards open source seeing that it is progressing faster.

bought Ṁ0 of NO

@ShadowyZephyr I haven't seen any open source models get remotely close to GPT-4, the best ones are just barely starting to get close to 3.5 Turbo.

predicts YES

@SemioticRivalry this is true but if the other models are going to slow down they have 2 years to catch up. In the scenario where everyone is blasting full speed ahead them catching up would not be nearly enough.

predicts YES

@SemioticRivalry Mixtral is reportedly better than GPT-3.5 on benchmarks.

The executive order threatens large proprietary models, but leaves open source models alone? Is that what I heard correctly just now? Theres a way that could give open source models an advantage over proprietary ones; government regulation.

bought Ṁ100 of NO

@VAPOR Not really correct, no. Threshold is based on compute, so in theory a hypothetical Llama-3 would have the reporting requirements even if then released as open source.

(I don't think those reporting requirements are a threat.)

bought Ṁ5 of YES

@DaveK It could be trained abroad though?

predicts NO

@mariopasquato US regulators tend not to look kindly on, “One Weird Trick to avoid US regulation” gambits. Most entities training such models want to do business in America

More cost efficient to whom?

For certain applications today, the terms of use of closed-source alternatives are a non-starter so OSLLM's are already more cost efficient than, "not available." Drinking one's own piss to stay alive comes to mind as an analogy.

@PatrickDelaney I am very curious as to the kind of life you have lived such that "Drinking one's own piss to stay alive" was the first analogy that came to mind here.

bought Ṁ100 of NO

Betting NO because if the market resolves YES then we're more likely to go extinct.

@MartinRandall Even in this non-conditional (non policy based) market, one should also signal their true beliefs to politicians! You're making them less likely to act based on not even profitable, but wishful mindset!

predicts NO

The close date of the market is 2025-01-02T07:59:00Z, but in the description there is:

> Resolves YES if, at the end of 2025

Is it expected that the close date is about 1 year earlier than the planned resolution? (Probably this is expected, asking to be sure that there is no error is market close date or in the description)

@bessarabov Oops, fixed.

Does Llama 2 count as open source?

@MrLuke255 Yes; why wouldn't it?

predicts YES

@EliezerYudkowsky It's at least quite contentious in the open source community. Pre-existing large companies have to apply for a license and the terms forbid using the weights to train another model. There are also restrictions on using it to break the law, write a virus, or generate spam, and I'd think at least the last two disqualify it. It should count for this market anyway because LeCun is definitely including Llama in his claim.

bought Ṁ10 of YES

@Frogswap This suggests that a good definition of “open source model” would make the question clearer

More related questions