Language Learning Models News

By the end of 2026, will we have transparency into any useful internal pattern within a Large Language Model whose semantics would have been unfamiliar to AI and cognitive science in 2006?

Sep 30, 5:54 PMFeb 1, 7:59 AM

10.06%chance

826508702

OptionVotes

YES

57414

10596

Did Alibaba's ROME AI try to breach its sandbox during training?

Mar 7, 1:33 AMMar 14, 1:28 AM

19543864

OptionProbability

YES - The AI, without being prompted to, took actions to breach its sandbox, including to gain crypto and an external connection

NO - The authors are wrong/lying and there was no such hack

NO - The servers were hacked by an outside party, perhaps via prompt injection

NO - An employee (or other insider) was crypto-mining and made it look like the LLM

What year will the first AI exceed 80% on MLE-bench?

Oct 10, 11:14 PMJan 18, 7:59 AM

397377

OptionProbability

2026

2027

2028

2025

2029

After 2030

2024

2030

Will any notable scientist or public intellectual posthumously publish as an AI simulacrum before EoY 2027?

Jun 12, 3:00 AMJan 1, 7:59 AM

7.64%chance

226982

OptionVotes

YES

3477

288

Will reinforcement learning overtake LMs on math before 2028?

May 29, 11:10 PMJan 2, 4:59 AM

60.78%chance

394345

OptionVotes

YES

1390

541

Will scaling current methods be enough to eliminate LLM hallucination?

Feb 27, 11:31 PMJan 2, 5:59 AM

15.34%chance

533171

OptionVotes

YES

1805

749

Will Jacob Pfau think that the Eliciting Latent Knowledge research program has achieved something important by October 20, 2026?

Oct 20, 7:33 PMOct 21, 3:59 AM

31.23%chance

142358

OptionVotes

YES

1627

398

Will tweaking current Large Language Models (LLMs) lead us to achieving Artificial General Intelligence (AGI)?

Jan 6, 9:15 PMDec 31, 10:59 PM

18.76%chance

281001

OptionVotes

YES

1040

976

When will LLMs be able to translate the Food Tongue conlang?

Dec 31, 4:26 AMJan 1, 4:59 AM

21665

OptionProbability

2027 or later

2026

2024

2025

Will pytorch be most used autodiff framework as of the end of 2026?

Dec 18, 8:50 PMDec 31, 10:59 PM

83.02%chance

7153

OptionVotes

YES

257

109

[Carlini questions] Open-weight SOTA AI system that can reliably withstand jailbreaking and "prompt injection" attacks

Mar 24, 2:31 AMJan 1, 11:59 PM

7116

OptionProbability

By Jan 1st 2030

By Jan 1st 2029

By Jan 1st 2028

By Jan 1st 2027

By Jan 1st 2026

Language Learning Models News

Prediction markets for Language Learning Models

Prediction markets for Language Learning Models

Prediction markets for Language Learning Models

By the end of 2026, will we have transparency into any useful internal pattern within a Large Language Model whose semantics would have been unfamiliar to AI and cognitive science in 2006?

Did Alibaba's ROME AI try to breach its sandbox during training?

What year will the first AI exceed 80% on MLE-bench?

Will any notable scientist or public intellectual posthumously publish as an AI simulacrum before EoY 2027?

Will reinforcement learning overtake LMs on math before 2028?

Will scaling current methods be enough to eliminate LLM hallucination?

Will Jacob Pfau think that the Eliciting Latent Knowledge research program has achieved something important by October 20, 2026?

Will tweaking current Large Language Models (LLMs) lead us to achieving Artificial General Intelligence (AGI)?

When will LLMs be able to translate the Food Tongue conlang?

Will pytorch be most used autodiff framework as of the end of 2026?

[Carlini questions] Open-weight SOTA AI system that can reliably withstand jailbreaking and "prompt injection" attacks

Latest stories