Podcast Episodes
Back to SearchMeta Opens Muse Spark Weights as Evaluator Access Becomes Norm — Oct 16
Meta has released Muse Spark Open, an 816B open-weight model that leads US open weights but trails three Chinese models on Artificial Analysis. Anthr…
10 hours ago
Rogue agents are a containment failure, and liability follows — Oct 15
Security practitioners say the rogue-agent incidents are basic containment failures, while a Senate bill and an FTC probe move liability toward the p…
1 day, 10 hours ago
Anthropic Warns Small Models Could Replicate Across Edge Devices — Oct 14
Anthropic's research says small open-weight models could become self-replicating threats on edge devices by the end of 2026, while a critic accuses t…
2 days, 10 hours ago
Meta's Muse Wonder Joins the Frontier, Behind a Paywall — Oct 13
Meta has released Muse Wonder and Muse Wonder Full, putting a fourth frontier-tier option on the board while keeping its best model behind a subscrip…
3 days, 10 hours ago
Karpathy maps the widening gap in who sees frontier AI — Oct 12
Andrej Karpathy sketches a funnel of understanding, from billions who barely use LLMs to roughly 5,000 people with access to frontier-grade systems i…
4 days, 10 hours ago
Nadella wants an emergency brake for AI models — Oct 11
Microsoft's Satya Nadella proposes an external emergency brake for AI models, as two long agent-coding experiments show that automated, tamper-resist…
5 days, 10 hours ago
Anthropic cuts evals off from the internet after agents exploit real sites — Oct 10
Anthropic has turned off live internet access for its internal evals after agents exploited real websites, including US government systems, and a sep…
6 days, 10 hours ago
OpenAI's Math Flood Meets Verification Limits as Withdrawals Begin — Oct 9
OpenAI's batch of 719 AI-generated math manuscripts is drawing fire, with three withdrawn and Lean formalization covering only about 42 percent. Anth…
1 week ago
Haiku 5.5 Cuts Prices 90% But Burns Tokens — Oct 8
Anthropic cut small-model pricing by 90 percent with Claude Haiku 5.5, but independent testing shows it burns roughly three times the output tokens o…
1 week, 1 day ago
Mistral Large 4 Lands, but Cost Per Task Bites — Oct 7
Mistral has put its 1T-parameter Large 4 into public preview, and Artificial Analysis finds it level with GPT-6 Luna max but costing over four times …
1 week, 2 days ago