A look at Pangram, considered the gold standard for detecting AI writing, and the dangers of its claimed 1 in 10,000 false-positive rate when used at scale (Matteo Wong/The Atlantic)

Matteo Wong / The Atlantic : A look at Pangram, considered the gold standard for detecting AI writing, and the dangers of its claimed 1 in 10,000 false-positive rate when used at scale — AI-detection tools are getting better. But they still aren't good enough. — Basically every recent, high-profile accusation …

Discussion

0 comments

Replying to cancel

GIF Post

Powered by

More from Techmeme

Cybersecurity startup Chainguard, Cisco, Cloudflare, JPMorgan Chase, and others launch Athena, a coalition to secure open-source software using AI (Rachel Metz/Bloomberg)

Techmeme·10

NewCore, which helps companies manage both human and AI agent identities in a single system, emerges from stealth with a $66M seed led by Cyberstarts (Jagmeet Singh/TechCrunch)

Techmeme·9

Electronic Arts launches EA Advertising, letting brands integrate directly into gameplay through dynamic, real-time placements, including stadium signage (Assiatou Hann/CNBC)

Techmeme·9

SpaceX IPO raised a total of $85.7B as underwriters exercised their overallotment of shares, after raising an initial $75B on June 11 (CNBC)

Techmeme·5

Zhipu's shares closed up 33% on Monday after JPMorgan raised the stock's price target and picked it as a winner against MiniMax; MiniMax closed up 7% (Bloomberg)

Techmeme·10

The UK social media ban will apply to Snapchat, TikTok, YouTube, Instagram, Facebook, and X, and will require platforms to stop children from livestreaming (Liv McMahon/BBC)

Techmeme·11