Before July 1, 2023, will an agentized-LLM system autonomously improve itself?

12

250Ṁ4778

resolved Jul 5

Resolved

NO

1H

6H

1D

1W

1M

ALL

From https://metaculus.com/questions/16618/agentized-llm-improves-itself/

The emergence of "agentized" large language models (LLMs) such as AutoGPT, HuggingGPT, Reflexion, and BabyAGI may be a significant focus in AI development's near future.

"Agentized" LLMs use multiple LLM systems to construct a cognitive architecture, with each LLM instance assigned a specific role, such as determining goals or offering executive function.

Get

1,000

to start trading!

🏅 Top traders

#	Name	Total profit
1		Ṁ66
2		Ṁ46
3		Ṁ17
4		Ṁ14
5		Ṁ13

Sort by:

why can't something dumb happen like: "How would you improve yourself, as a large language model"? "As a large langauge model, to improve myself, I would train myself for a thousand more gpu-hours".

People are also trading

By the end of June 2025, will closed-source LLMs increase access to pandemic agents?

Will one of the major LLMs be capable of continual lifelong learning (learning from inference runs) by EOY 2025?

Will LLMs (or similar AI systems) be meaningfully integrated into US public school education by 2025?

Will an LLM improve its own ability along some important metric well beyond the best trained LLMs before 2026?

Will an LLM agent complete >50% of the lab tasks on the Factorio Learning Environment benchmark in 2025?

Will the most interesting AI in 2027 be a LLM?

Will an AI Tutor (LLM personalized for each student) replace conventional teaching by 2035?

Will there be major breakthrough in LLM Continual Learning before 2026?

By 2025 end, will it be generally agreed upon that LLM produced text/code > human text/code for training LLMs?

Will a LLM-based AI be used for a law enforcement decision before 2025?

Related questions

By the end of June 2025, will closed-source LLMs increase access to pandemic agents?

Will one of the major LLMs be capable of continual lifelong learning (learning from inference runs) by EOY 2025?

Will LLMs (or similar AI systems) be meaningfully integrated into US public school education by 2025?

Will an LLM improve its own ability along some important metric well beyond the best trained LLMs before 2026?

Will an LLM agent complete >50% of the lab tasks on the Factorio Learning Environment benchmark in 2025?

Will the most interesting AI in 2027 be a LLM?

Will an AI Tutor (LLM personalized for each student) replace conventional teaching by 2035?

Will there be major breakthrough in LLM Continual Learning before 2026?

By 2025 end, will it be generally agreed upon that LLM produced text/code > human text/code for training LLMs?

Will a LLM-based AI be used for a law enforcement decision before 2025?

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