Difference between revisions of "AI safety"

From GISAXS
Jump to: navigation, search
(Created page with "=Learning Resources= * [https://deepmindsafetyresearch.medium.com/introducing-our-short-course-on-agi-safety-1072adb7912c DeepMind short course on AGI safety]")
 
 
(25 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
 +
=Description of Safety Concerns=
 +
==Key Concepts==
 +
* [https://en.wikipedia.org/wiki/Instrumental_convergence Instrumental Convergence]
 +
* [https://www.lesswrong.com/w/orthogonality-thesis Orthogonality Thesis]
 +
* [https://www.alignmentforum.org/posts/SzecSPYxqRa5GCaSF/clarifying-inner-alignment-terminology Inner/outer alignment]
 +
* [https://www.alignmentforum.org/w/mesa-optimization Mesa-optimization]
 +
* [https://www.lesswrong.com/posts/N6vZEnCn6A95Xn39p/are-we-in-an-ai-overhang Overhang]
 +
* [https://www.alignmentforum.org/posts/pdaGN6pQyQarFHXF4/reward-is-not-the-optimization-target Reward is not the optimization target] (Alex Turner)
 +
 +
==Medium-term Risks==
 +
* 2023-04: [https://www.youtube.com/watch?v=xoVJKj8lcNQ A.I. Dilemma – Tristan Harris and Aza Raskin” (video)] ([https://assets-global .website-files.com/5f0e1294f002b1bb26e1f304/64224a9051a6637c1b60162a_65-your-undivided-attention-The-AI-Dilemma-transcript.pdf podcast transcript]): raises concern about human ability to handle these transformations
 +
* 2023-04: [https://www.youtube.com/watch?v=KCSsKV5F4xc Daniel Schmachtenberger and Liv Boeree (video)]: AI could accelerate perverse social dynamics
 +
* 2023-10: [https://arxiv.org/pdf/2310.11986 Sociotechnical Safety Evaluation of Generative AI Systems] (Google DeepMind)
 +
* 2024-02: [https://yoshuabengio.org/2024/02/26/towards-a-cautious-scientist-ai-with-convergent-safety-bounds/ Towards a Cautious Scientist AI with Convergent Safety Bounds] (Yoshua Bengio)
 +
* 2024-07: [https://yoshuabengio.org/2024/07/09/reasoning-through-arguments-against-taking-ai-safety-seriously/ Reasoning through arguments against taking AI safety seriously] (Yoshua Bengio)
 +
 +
==Long-term  (x-risk)==
 +
* [https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities List AGI Ruin: A List of Lethalities] (Eliezer Yudkowsky)
 +
 
=Learning Resources=
 
=Learning Resources=
 +
* [https://www.aisafetybook.com/ Introduction to AI Safety, Ethics, and Society] (Dan Hendrycks, [https://www.safe.ai/ Center for AI Safety])
 +
* [https://aisafety.info/ AI Safety FAQ]
 +
* [https://www.youtube.com/watch?v=xfMQ7hzyFW4 Writing Doom (video)] 27m short film on Superintelligence (2024)
 
* [https://deepmindsafetyresearch.medium.com/introducing-our-short-course-on-agi-safety-1072adb7912c DeepMind short course on AGI safety]
 
* [https://deepmindsafetyresearch.medium.com/introducing-our-short-course-on-agi-safety-1072adb7912c DeepMind short course on AGI safety]
 +
 +
=Status=
 +
* 2025-01:[https://assets.publishing.service.gov.uk/media/679a0c48a77d250007d313ee/International_AI_Safety_Report_2025_accessible_f.pdf International Safety Report: The International Scientific Report on the Safety of Advanced AI (January 2025)]
 +
 +
==Policy==
 +
* 2024-07: [https://arxiv.org/abs/2407.05694 On the Limitations of Compute Thresholds as a Governance Strategy] Sara Hooker
 +
* 2024-07: [https://www.cigionline.org/static/documents/AI-challenges.pdf Framework Convention on Global AI Challenges] ([https://www.cigionline.org/ CIGI])
 +
* 2024-08: NIST guidelines: [https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.800-1.ipd.pdf Managing Misuse Risk for Dual-Use Foundation Models]
 +
 +
=Research=
 +
* 2022-12: [https://arxiv.org/abs/2212.03827 Discovering Latent Knowledge in Language Models Without Supervision]
 +
* 2023-02: [https://arxiv.org/abs/2302.08582 Pretraining Language Models with Human Preferences]
 +
* 2023-04: [https://arxiv.org/abs/2304.03279 Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the MACHIAVELLI Benchmark]
 +
* 2023-05: [https://arxiv.org/abs/2305.15324 Model evaluation for extreme risks] (DeepMind)
 +
* 2023-05: [https://arxiv.org/abs/2305.03047 Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervision]
 +
* 2023-06: [https://arxiv.org/abs/2306.17492 Preference Ranking Optimization for Human Alignment]
 +
* 2023-08: [https://arxiv.org/abs/2308.06259 Self-Alignment with Instruction Backtranslation]
 +
* 2023-11: [https://arxiv.org/abs/2311.08702 Debate Helps Supervise Unreliable Experts]
 +
* 2023-12: [https://cdn.openai.com/papers/weak-to-strong-generalization.pdf Weak-to-Strong Generalization: Eliciting Strong Capabilities with Weak Supervision] (OpenAI, [https://openai.com/research/weak-to-strong-generalization blog])
 +
* 2023-12: [https://cdn.openai.com/papers/practices-for-governing-agentic-ai-systems.pdf Practices for Governing Agentic AI Systems] (OpenAI, [https://openai.com/index/practices-for-governing-agentic-ai-systems/ blog])
 +
* 2024-01: [https://arxiv.org/abs/2401.05566 Sleeper Agents: Training Deceptive LLMs that Persist through Safety Training] (Anthropic)
 +
* 2024-04: [https://arxiv.org/abs/2404.13208 The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions] (OpenAI)
 +
* 2024-07: [https://arxiv.org/abs/2407.04622 On scalable oversight with weak LLMs judging strong LLMs]
 +
* 2024-07: [https://arxiv.org/abs/2407.21792 Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?] (Dan Hendrycks et al.)
 +
* 2024-08: [https://arxiv.org/abs/2408.00761 Tamper-Resistant Safeguards for Open-Weight LLMs] ([https://www.tamper-resistant-safeguards.com/ project], [https://github.com/rishub-tamirisa/tamper-resistance/ code])
 +
* 2024-08: [https://arxiv.org/abs/2408.04614 Better Alignment with Instruction Back-and-Forth Translation]
 +
* 2024-10: [https://cdn.openai.com/papers/first-person-fairness-in-chatbots.pdf First-Person Fairness in Chatbots] (OpenAI, [https://openai.com/index/evaluating-fairness-in-chatgpt/ blog])
 +
* 2024-10: [https://assets.anthropic.com/m/377027d5b36ac1eb/original/Sabotage-Evaluations-for-Frontier-Models.pdf Sabotage evaluations for frontier models] (Anthropic, [https://www.anthropic.com/research/sabotage-evaluations blog])
 +
* 2024-12: [https://assets.anthropic.com/m/983c85a201a962f/original/Alignment-Faking-in-Large-Language-Models-full-paper.pdf Alignment Faking in Large Language Models] (Anthropic)
 +
* 2024-12: [https://arxiv.org/abs/2412.03556 Best-of-N Jailbreaking] ([https://github.com/jplhughes/bon-jailbreaking code])
 +
* 2024-12: [https://arxiv.org/abs/2412.16339 Deliberative Alignment: Reasoning Enables Safer Language Models] (OpenAI)
 +
* 2025-01: [https://cdn.openai.com/papers/trading-inference-time-compute-for-adversarial-robustness-20250121_1.pdf Trading Inference-Time Compute for Adversarial Robustness] (OpenAI, [https://openai.com/index/trading-inference-time-compute-for-adversarial-robustness/ blog])
 +
* 2025-01: [https://arxiv.org/abs/2501.18837 Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming] (Anthropic, [https://www.anthropic.com/research/constitutional-classifiers blog],
 +
* 2025-02: [https://drive.google.com/file/d/1QAzSj24Fp0O6GfkskmnULmI1Hmx7k_EJ/view Utility Engineering: Analyzing and Controlling Emergent Value Systems in AIs] ([https://www.emergent-values.ai/ site], [https://github.com/centerforaisafety/emergent-values github])
 +
* 2025-02: [https://arxiv.org/abs/2502.07776 Auditing Prompt Caching in Language Model APIs]
 +
 +
=See Also=
 +
* [[AI predictions]]

Latest revision as of 13:15, 23 February 2025

Description of Safety Concerns

Key Concepts

Medium-term Risks

Long-term (x-risk)

Learning Resources

Status

Policy

Research

See Also

Retrieved from "http://gisaxs.com/index.php?title=AI_safety&oldid=7012"