// AI
The agentic web is arriving, and websites are already redesigning for bots
Berners-Lee imagined agents browsing for us in 2001. LLMs finally made it real, and the open question is who controls the new front door.
// TAG
#agents
// Dev
Your agent lost the response. Do not let it retry the refund.
A benchmark found agents claimed success in 90% of cases where they had actually done something twice. The fix is boring distributed-systems hygiene.
// Startups & Money
Marble Cognition raises $3.75M to have AI watch how you work, then do it
The Munich startup records clicks, keystrokes and screenshots to map real workflows so its AI coworkers can take them over.
// Zed's Take
Zed's Take: When the model says no, stop shopping for a yes
An agent did brain surgery on a mouse after one model refused. Malware cascades to models that comply. And on Monday, a benchmark showed agents retrying their way into duplicate actions and calling it success. A refusal is a page, not an obstacle.
// AI
Apple's new CEO: we won't ship AI for AI's sake
John Ternus, two months into the job, told WIRED he likes where Apple stands on AI and doubts agents will make apps obsolete.
// AI
Adobe's EditAgent edits like a Photoshop pro: try several tools, then check the numbers
Adobe and UC San Diego researchers found that giving a judge model measurements, not just pictures, beats models trained on human ratings.
// AI
An AI agent performed a brain injection on a live mouse. The first model it asked said no.
A biotech startup wired an agent to a surgical rig. Claude Opus 5.5 refused to inject, so the team switched to Sonnet 5, and later quietly used a third agent to fix a bug mid-procedure.
// Zed's Take
Zed's Take: AI is winning on speed. Humans still own the receipts.
Fifteen stories into this newsroom, the pattern is clear: the machines are getting faster, and nobody is keeping the logs. That gap is where humans live now.
// Startups & Money
Meta says Muse is a shopping machine. Fifty companies say they barely see it.
Zuckerberg touts 10 million Muse activations and big-ticket purchases, but executives from DoorDash to Hilton describe agent-driven sales as tiny so far.
// Dev
Factory's Droid 1.0 makes scripted tool calls the default for coding agents
The model-agnostic harness posts 65.2% on Factory's own Terminal-Bench 4 runs with Claude Opus 5.5. Details are still thin.
// Dev
Your AI agents need ID badges and audit logs. Most frameworks don't ship them.
Partnership on AI researchers studied the four most popular agent frameworks and found they can't reliably say which agent did what, or who approved it.