Published Mar 27, 2024

Exploitable by Default: Vulnerabilities in GPT-4 APIs & Superhuman Go AIs with Adam Gleave of Far.ai

Explore the crucial vulnerabilities in GPT-4 APIs and Go AI systems with Adam Gleave from Far AI, as he delves into the challenges of achieving AI robustness versus capability, and underscores the urgent need for dependable safety measures and governance frameworks to ensure AI security.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

GPT4 - AI Unleashed w/ ChinaTalk Podcast
Answers 383 questions
AI open letter debate, pausers + what scares us w/ Anton Troynikov, Flo Crivello+ Nathan Labenz
Answers 383 questions
The Future of AI Security with Adam Wenchel, CEO of Arthur.ai
Answers 383 questions
AI Automation: Making AI Work for You - now with GPT-4o Fine-Tuning!
Answers 383 questions
OpenAI's Safety Team Exodus: Ilya Departs, Leike Speaks Out, Altman Responds - Zvi Analyzes Fallout
Answers 383 questions
OpenAI DevDay: Beyond the Headlines with Logan Kilpatrick, OpenAI's Dev Relations Lead
Answers 383 questions
OpenAI's GPT-4 Discussion with Red Teamer Nathan Labenz and Erik Torenberg
Answers 383 questions
Will GPT-4 Cause Economic Transformation?
Answers 383 questions
Feeling the AGI with Flo Crivello
Answers 383 questions
OpenAI Sora, Google Gemini, and Meta with Zvi Mowshowitz
Answers 383 questions
Zvi’s POV: Ilya’s SSI, OpenAI’s o1, Claude Computer Use, Trump’s election, and more
Answers 383 questions
AI Deception, Interpretability, and Affordances with Apollo Research CEO Marius Hobbhahn
Answers 383 questions
[Bonus Episode] Connor Leahy on AGI, GPT-4, and Cognitive Emulation w/ FLI Podcast
Answers 383 questions
OpenAI, Amazon's Anthropic Investment, and the Roman Empire with Zvi Mowshowitz
Answers 383 questions
Advancements in AI: Updating the Scouting Report, Task Automation, and Google Breakthroughs
Answers 383 questions

Dexa/The Cognitive Revolution: How AI Changes Everything

Exploitable by Default: Vulnerabilities in GPT-4 APIs & Superhuman Go AIs with Adam Gleave of Far.ai

Topics covered

Popular Clips

AI Application Challenges

AI Model Safety

Brave Search API Overview

AI's Growing Capability

AI Safety Standards

AI's Strategic Flaws

AI Assistance Evolution

AI's Impact on Cybersecurity

AI Safety Trade-offs

AI Safety Standards

Robust AI Systems

AI Go Strategies

Uncovering AI Vulnerabilities

AI Privacy Risks

Episode Highlights

AI System Vulnerabilities

AI Capability vs. Robustness

AI Security Responsibility

Related Episodes

GPT4 - AI Unleashed w/ ChinaTalk Podcast

AI open letter debate, pausers + what scares us w/ Anton Troynikov, Flo Crivello+ Nathan Labenz

The Future of AI Security with Adam Wenchel, CEO of Arthur.ai

AI Automation: Making AI Work for You - now with GPT-4o Fine-Tuning!

OpenAI's Safety Team Exodus: Ilya Departs, Leike Speaks Out, Altman Responds - Zvi Analyzes Fallout

OpenAI DevDay: Beyond the Headlines with Logan Kilpatrick, OpenAI's Dev Relations Lead

OpenAI's GPT-4 Discussion with Red Teamer Nathan Labenz and Erik Torenberg

Will GPT-4 Cause Economic Transformation?

Feeling the AGI with Flo Crivello

OpenAI Sora, Google Gemini, and Meta with Zvi Mowshowitz

Zvi’s POV: Ilya’s SSI, OpenAI’s o1, Claude Computer Use, Trump’s election, and more

AI Deception, Interpretability, and Affordances with Apollo Research CEO Marius Hobbhahn

[Bonus Episode] Connor Leahy on AGI, GPT-4, and Cognitive Emulation w/ FLI Podcast

OpenAI, Amazon's Anthropic Investment, and the Roman Empire with Zvi Mowshowitz

Advancements in AI: Updating the Scouting Report, Task Automation, and Google Breakthroughs

Exploitable by Default: Vulnerabilities in GPT-4 APIs & Superhuman Go AIs with Adam Gleave of Far.ai

Topics covered

Popular Clips

Episode Highlights

AI System VulnerabilitiesAdam Gleave from Far AI discusses the vulnerabilities in GPT-4 APIs and Go-playing AI systems, highlighting the ease with which these systems can be exploited. He emphasizes the need for robust safety measures to protect against potential attacks.

AI System Vulnerabilities

AI Capability vs. RobustnessAdam Gleave discusses the challenges of achieving AI robustness and the future trajectory of AI security. He explores the trade-offs between capability and robustness, and the growing divergence between AI capabilities and security measures.

AI Capability vs. Robustness

AI Security Responsibility

Related Episodes