Before 2028, will there be enough inference capacity to generate 30T frontier model tokens per day? | Manifold

Before 2028, will there be enough inference capacity to generate 30T frontier model tokens per day?

Plus

7

Ṁ469

2028

40%

chance

1D

1W

1M

ALL

In "Situational Awareness: The Decade Ahead", Leopold Aschenbrenner claims:

Another way of thinking about it is that given inference fleets
in 2027, we should be able to generate an entire internet’s worth of
tokens, every single day.

Resolves YES if by the end of 2027, there is enough deployed inference capacity to generate 30 trillion tokens in a 24-hour period using a combination of frontier models. "Frontier models" in the sense that GPT-4 is a frontier model today in mid-2024.

This is one of a series of markets on claims made in Leopold Aschenbrenner's Situational Awareness report(s):

Other markets about Leopold's predictions:

/ZviMowshowitz/will-we-develop-leopolds-dropin-rem

This question is managed and resolved by Manifold.

#Technical AI Timelines

#Situational Awareness

Get

1,000

and

3.00

Sort by:

Edited title to match language of other [LA:SA] markets. (No change to meaning.)

Related questions

Will an AI achieve >85% performance on the FrontierMath benchmark before 2028?

Will anyone train a TokenFormer model at scale before 2026?

100GW AI training run before 2031?

Will a US Department of Energy high performance computing cluster be used to train a foundation model of more than 500B or more by January 1st 2025?

Before 2028, will any AI model achieve the same or greater benchmarks as o3 high with <= 1 million tokens per question?

Will a new lab create a top-performing AI frontier model before 2028?

Will an AI achieve >85% performance on the FrontierMath benchmark before 2027?

Will OpenAI inference costs fall by 100x over the next 18 months?

$1T AI training cluster before 2031?

10GW AI training run before 2029?

Related questions

Will an AI achieve >85% performance on the FrontierMath benchmark before 2028?

Will a new lab create a top-performing AI frontier model before 2028?

Will anyone train a TokenFormer model at scale before 2026?

Will an AI achieve >85% performance on the FrontierMath benchmark before 2027?

100GW AI training run before 2031?

Will OpenAI inference costs fall by 100x over the next 18 months?

Will a US Department of Energy high performance computing cluster be used to train a foundation model of more than 500B or more by January 1st 2025?

$1T AI training cluster before 2031?

Before 2028, will any AI model achieve the same or greater benchmarks as o3 high with <= 1 million tokens per question?

10GW AI training run before 2029?

© Manifold Markets, Inc.•Terms + Mana-only Terms•Privacy•Rules