Will it cost $30 to train a GPT-3 level model in 2030? | Manifold

Will it cost $30 to train a GPT-3 level model in 2030?

Plus

19

Ṁ1371

2031

19%

chance

1D

1W

1M

ALL

If Ark's prediction proves correct, this market resolves to YES. I'll leave the definitions open, relying in experts (I hope my GF is a PhD in ML by then), but both press releases and opinion of experts, including the specialized press, will count to this market.

I won't bet.

#Technical AI Timelines

Get

1,000

and

1.00

Sort by:

https://www.semianalysis.com/p/google-we-have-no-moat-and-neither

Related questions

Will it cost less than 100k USD to train and run a language model that outperforms GPT-3 175B on all benchmarks by the end 2024?

Will a GPT-3 quality model be trained for under $1,000 by 2030?

Will $10,000 worth of AI hardware be able to train a GPT-3 equivalent model in under 1 hour, by EOY 2027?

100GW AI training run before 2031?

Will it be possible to disentangle most of the features learned by a model comparable to GPT-3 this decade? (1k subsidy)

Will a GPT-3 quality model be trained for under $10.000 by 2030?

Will a GPT-4 quality model be trained for under $10.000 by 2030?

Will a model costing >$30M be intentionally trained to be more mechanistically interpretable by end of 2027? (see desc)

Will there be an LLM (as good as GPT-4) that was trained with 1/100th the energy consumed to train GPT-4, by 2026?

Before 2028, will anyone train a GPT-4-level model in a minute?

Related questions

Will it cost less than 100k USD to train and run a language model that outperforms GPT-3 175B on all benchmarks by the end 2024?

Will a GPT-3 quality model be trained for under $10.000 by 2030?

Will a GPT-3 quality model be trained for under $1,000 by 2030?

Will a GPT-4 quality model be trained for under $10.000 by 2030?

Will $10,000 worth of AI hardware be able to train a GPT-3 equivalent model in under 1 hour, by EOY 2027?

Will a model costing >$30M be intentionally trained to be more mechanistically interpretable by end of 2027? (see desc)

100GW AI training run before 2031?

Will there be an LLM (as good as GPT-4) that was trained with 1/100th the energy consumed to train GPT-4, by 2026?

Will it be possible to disentangle most of the features learned by a model comparable to GPT-3 this decade? (1k subsidy)

Before 2028, will anyone train a GPT-4-level model in a minute?

Terms & Conditions•Privacy Policy•Sweepstakes Rules