By January 2026, will we have a language model with similar performance to GPT-3.5 (i.e. ChatGPT as of Feb-23) that is small enough to run locally on the highest end iPhone available at the time? | Manifold

By January 2026, will we have a language model with similar performance to GPT-3.5 (i.e. ChatGPT as of Feb-23) that is small enough to run locally on the highest end iPhone available at the time?

Plus

24

Ṁ4450

2026

92%

chance

1D

1W

1M

ALL

Any proof that this is possible counts.

This question is managed and resolved by Manifold.

#️ Technology

Get

1,000

and

3.00

Sort by:

Mistral Small seems similar in performance to GPT 3.5: https://mistral.ai/news/la-plateforme/

Should be a matter of days until someone runs it on their iPhone

It runs on a MacBook, fairly trivial to get ~3x gain in tflops and 3-10x gain in model performance over this window

predictedYES

https://twitter.com/simonw/status/1634635007712165888?s=46&t=7MQWoR0BbBi0qMI9dbdAwQ

Federated learning

Might happen much sooner than I expected

https://twitter.com/harmlessai/status/1626769581858758661?s=46&t=-aOs5vi8y_5tlqgbjNoRKA

predictedYES

As evidence in favour:

this type of advance in making existing transformers smaller at same performance: https://twitter.com/arankomatsuzaki/status/1624947959644278786?s=20&t=sWoi47Zz-RRprcyDC-jZog
optimal training data vs parameter count (a la Chinchilla)
the hefty "neural engines" that Apple is continuously making better in their apple silicon
that we may find new architectures that work better for language

Related questions

By January 2026, will a language model with similar performance to GPT-4 be able to run locally on the latest iPhone?

Will a language model that runs locally on a consumer cellphone beat GPT4 by EOY 2026?

Will inflection AI have a model that is 10X the size of original GPT-4 at the end of Q1, 2025?

Will we have an open-source model that is equivalent GPT-4 by end of 2025?

Will a language model comparable to GPT-4 be trained, with ~1/10th the amount of energy it took train GPT-4, by 2028?

Will there be an AI language model that strongly surpasses ChatGPT and other OpenAI models before the end of 2025?

Will there be a language model called GPT-5, released by OpenAI, this decade?

By 2028, will there be a language model of less than 10B parameters that is superior to GPT-4?

Will language models, such as ChatGPT, be integrated into dialogue trees for NPCs in triple-A games by the end of 2025

Will a model as great as GPT-5 be available to the public in 2025?

Related questions

By January 2026, will a language model with similar performance to GPT-4 be able to run locally on the latest iPhone?

Will there be an AI language model that strongly surpasses ChatGPT and other OpenAI models before the end of 2025?

Will a language model that runs locally on a consumer cellphone beat GPT4 by EOY 2026?

Will there be a language model called GPT-5, released by OpenAI, this decade?

Will inflection AI have a model that is 10X the size of original GPT-4 at the end of Q1, 2025?

By 2028, will there be a language model of less than 10B parameters that is superior to GPT-4?

Will we have an open-source model that is equivalent GPT-4 by end of 2025?

Will language models, such as ChatGPT, be integrated into dialogue trees for NPCs in triple-A games by the end of 2025

Will a language model comparable to GPT-4 be trained, with ~1/10th the amount of energy it took train GPT-4, by 2028?

Will a model as great as GPT-5 be available to the public in 2025?

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