Cybersecurity researchers aren’t happy about the guardrails on Anthropic’s Fable; plus, how memory tools can make AI models worse
About this episode
Cybersecurity researchers criticize the strict guardrails of Anthropic's Fable model, while new research indicates that AI memory systems may negatively impact model performance.
Cybersecurity researchers are complaining that Anthropic's new model Fable has guardrails that are too strict for any cybersecurity work. Also, new research suggests that AI memory systems can degrade model performance and encourage sycophantic tendencies. Learn more about your ad choices. Visit podcastchoices.com/adchoices
Topics covered
- cybersecurity
- AI models
- memory tools
- model performance
- research
Keywords
- cybersecurity
- Anthropic
- Fable
- AI memory systems
- model performance
- research
Mentioned in this episode
Organizations: Anthropic, Cybersecurity researchers
Products: Fable
More episodes of TechCrunch Industry News
- xAI fired an engineer who raised alarms about Grok safety, new lawsuit claims; plus, Microsoft taps Alt Carbon for carbon removal
- Apple says it may remove some apps from the App Store; Rivian started deliveries of its R2 SUV; and so long FAANG, hello MANGOS.
- Pentagon says Alibaba, Baidu, BYD, and Unitree support China’s military; plus, Apple bets cheaper AI will woo small developers; and OpenAI filed for IPO
- Former cyber executive turned whistleblower accuses IBM of covering up several data breaches; plus, OpenAI is still working on that ‘super app’
- Chinese spies are using LinkedIn to lure Westerners into sharing sensitive information; Waymo as getaway vehicle, and Uber's data-collection vehicles
- The world’s largest privately owned laser just turned on; plus, Uber caps employee AI spending and Martin Scorsese becomes the latest Hollywood voice for AI
Explore listener stats, chart rankings, contacts and more on the TechCrunch Industry News podcast page.