AI has flattened the attacker pyramid, and defenders trail
CrowdStrike CEO George Kurtz argues attackers now use unguarded open-weight models at machine speed, while defenders blocked by frontier-model refusals need purpose-built security AI.
Loading articles…
The complete subject coverage could not load. Published articles remain below.
OpenAI says computer-use agents near superhuman speed; control becomes the constraint
OpenAI's Ari Weinstein and James Sun argue that agents operating software like humans are nearing expert speed, shifting the hard problems to enterprise controls, auditability, ad-funded web economics, and adoption.
Anthropic's founders say trust and incentives, not rules, make safer AI
Anthropic's Dario and Daniela Amodei argue that scaling points toward AI surpassing humans, and that trust, principled training and business incentives, not rules alone, determine whether it stays safe.
When AI makes design output cheap, selection becomes the job
Shopify's Marvin Schwaibold, speaking with the show's hosts, argues designers now steer AI agents and prototype in code, while startup website reviews show how abundance without curation produces clutter.
Why overpaying for power-law winners inflates venture bubbles
Matt of Bain Capital Ventures argues that rational price insensitivity toward power-law winners inflates bubbles, even as he judges AI's frontier capability gains, not valuations, as the deciding variable.
Rival AI labs race to build what they fear
Christiane Amanpour interviews Kevin Roose, author of The AGI Chronicles, on why AI labs that fear superintelligence keep racing to build it, and whether unelected insiders should decide.
After OpenAI's proof dump, mathematicians' job shifts from solving to understanding
Computer scientist Scott Aaronson argues AI's solution of hundreds of open problems cannot be undone, pitting mathematics' ideal of open participation against grief over lost careers and unreviewed proofs.