AI Safety Level News

Will we have better-than-human-aggregate forecasting AIs by the end of 2024?

Feb 22, 8:28 PMDec 1, 10:59 PM

4.03%chance

326944

OptionVotes

YES

4176

679

Will AI xrisk seem to be handled seriously by the end of 2026?

May 25, 5:47 PMDec 31, 10:59 PM

23.21%chance

425408

OptionVotes

YES

1101

940

When will there first be an AI system that qualifies as ASL-3?

Oct 3, 2:03 AMJan 1, 4:59 AM

115267

OptionProbability

Other

2025 Q1

2023 Q4

2024 Q1

2024 Q2

2024 Q3

2024 Q4

The first Anthropic-trained AI system to qualify as ASL-3 qualifies because of misuse risks rather than autonomy risks

Oct 3, 2:12 AMJan 1, 4:59 AM

89.7%chance

212044

OptionVotes

YES

1993

798

Aging cured by 2040 if AGI by 2030?

May 7, 9:30 PMDec 30, 9:09 PM

18.71%chance

231420

OptionVotes

YES

208

Will California pass an AI Safety bill in 2025?

Jan 11, 12:14 AMJan 1, 7:59 AM

51.88%chance

131307

OptionVotes

YES

1038

963

When will Anthropic first train an AI system that they claim qualifies as ASL-3?

Oct 3, 1:58 AMJan 1, 4:59 AM

131126

OptionProbability

2025 Q1

2024 Q4

Other

2024 Q3

2024 Q2

2024 Q1

2023 Q4

Never

Sometime before 2031, AI is going to cross a generality threshold that allows for recursive self-improvement, and then FOOM to superhuman capability. Before that point humanity must solve AI safety or be helpless.

Dec 24, 3:53 AMFeb 1, 7:59 AM

27.15%chance

17792

OptionVotes

YES

990

970

If most of the economy becomes automated without us all dying, will Eliezer Yudkowsky admit that he was wrong?

May 15, 12:35 AMJan 2, 7:59 AM

40.15%chance

21386

OptionVotes

YES

122

❓ If AGI turns out to be a disaster, what will be the main cause?

Aug 3, 5:44 AMAug 11, 3:59 AM

10332

OptionProbability

🕵️‍♂️💣 Used by bad actors — hackers, authoritarian regimes, or terrorists weaponize AGI.

🎮❌ Loss of control — humans can't predict or stop AGI’s actions anymore.

🧠⚠️ Misaligned goals — AGI misunderstands human values and acts against our interests.

💼📉 Job loss & chaos — AGI wipes out jobs, leading to mass instability.

👺🐒 Humans are evil — AGI can only align the universe by destroying us.

Other

🚀💡 Runaway self-improvement — AGI upgrades itself faster than humans can follow.

🌐🤝 No global coordination — nations or companies rush ahead without cooperation.

🌳💥 Environmental harm — AGI decisions damage ecosystems or speed up climate change.

🤯🔀 Unintended consequences — AGI does something harmful without meaning to.

🛑🔧 Weak safety protocols — not enough safeguards or oversight before release.

🧠📲 Misinformation at scale — AGI spreads lies, manipulates societies.

Will Anthropic announce one of their AI systems is ASL-4 or higher before the end of 2025?

Aug 23, 6:42 PMJan 1, 7:59 AM

7.49%chance

13321

OptionVotes

YES

351

Will a major AI company publish a “responsible scaling policy” for AI consciousness by 2030?

Feb 4, 11:21 PMJan 1, 7:59 AM

54.85%chance

11205

OptionVotes

YES

180

150

AI Safety Level News

Prediction markets for AI Safety Level

Prediction markets for AI Safety Level

Prediction markets for AI Safety Level

Will we have better-than-human-aggregate forecasting AIs by the end of 2024?

Will AI xrisk seem to be handled seriously by the end of 2026?

When will there first be an AI system that qualifies as ASL-3?

The first Anthropic-trained AI system to qualify as ASL-3 qualifies because of misuse risks rather than autonomy risks

Aging cured by 2040 if AGI by 2030?

Will California pass an AI Safety bill in 2025?

When will Anthropic first train an AI system that they claim qualifies as ASL-3?

Sometime before 2031, AI is going to cross a generality threshold that allows for recursive self-improvement, and then FOOM to superhuman capability. Before that point humanity must solve AI safety or be helpless.

If most of the economy becomes automated without us all dying, will Eliezer Yudkowsky admit that he was wrong?

❓ If AGI turns out to be a disaster, what will be the main cause?

Will Anthropic announce one of their AI systems is ASL-4 or higher before the end of 2025?

Will a major AI company publish a “responsible scaling policy” for AI consciousness by 2030?

Latest stories