Moonshot releases Kimi K3.1 Pro and says it cut its hacking skills
THE SO WHAT
Open-weight models that nearly match Claude Opus 5.5 on coding while stripping hacking data show how fast the high-end frontier is leaking into the open ecosystem. If you rely on “closed beats open” as a safety or moat assumption, you need a more granular threat model and eval regime.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIOpenAI Sets Target Valuation of $1.25 Trillion
If OpenAI clears a $30 billion raise at a $1.25 trillion valuation, AI labs are no longer just vendors in your stack—they’re macro assets whose incentives will shape pricing, access, and roadmap stability. Enterprise buyers should assume more aggressive monetization, tighter ecosystem control, and longer contract lock-ins as that capital looks for returns.
Applied AIHindsight puts up to $1M behind its AI content moderation as Google shuts down Perspective API
With Google’s Perspective API going dark on December 31 and every pipeline calling it failing on January 1, moderation just became an urgent replatforming project. If your trust-and-safety stack touches Perspective, you need a replacement and a backfill plan for historical scoring this quarter, not next year.
Applied AIEtsy is cracking down on AI slop. Here’s what’s changing.
Etsy banning AI-generated recipes/patterns and forcing real product photos is a clear move to protect marketplace trust as generative content floods the long tail. Any platform that monetizes uniqueness or craftsmanship will need similarly explicit AI policies—or risk eroding both buyer confidence and seller margins.
Applied AISam Altman: ‘I No Longer Think We’re Gonna Be Able to Just Patch All the Bugs’
When a leading lab CEO says we won’t be able to “just patch all the bugs,” he’s telegraphing that AI risk management will look more like safety engineering than app updates. Operators should be budgeting for continuous red-teaming, layered controls, and incident response around models—not assuming vendor patches will close every gap.