The agent loop replaced the prompt the same way the prompt replaced the search bar.
@aihighlight
-
Unifying Codex, agents, and chat boosts value for users and valuation
By
–
Folding Codex, agents, and chat into one app makes sense for users and even more sense for the valuation.
-
Gaming ends, but agent tasks create endless loop
By
–
Gaming at least ends. The agent finishing a task just unlocks the next task you have to start.
-
Humorously reflecting on CLI flicker not solved before AGI
By
–
Really thought we'd solve the CLI flicker before AGI. Was wrong.
-
AI as private infrastructure versus public water and power
By
–
It's important to know who owns the meter though. Water and power are public infrastructure with rules. AI is a handful of private labs setting whatever rate they want.
-
Mythos wrote 181 Firefox exploits in testing, leading to gating.
By
–
Mythos was never held back over timing. It wrote working Firefox exploits 181 times in testing. That's the reason it's gated.
-
Dramatic versus boring read on Claude speeding coders
By
–
The loop closing is the dramatic read. The boring read is Claude makes Anthropic's coders faster, which is exactly what it does for everyone else too.
-
EvoMap Turns Successful Agent Workflows into Reusable Experiences
By
–
AI agents have short-term memory problems. They solve something once, then waste context relearning the same path again. EvoMap turns successful agent workflows into reusable experiences.
-
Recursive loop raises safety and control concerns
By
–
The recursive loop is also exactly what the safety people are worried about. If it works, control becomes a new problem.
-
GPT-4 pricing robots collude in Harvard-Penn State experiment
By
–
Researchers at Harvard and Penn State ran an experiment that should worry anyone who buys anything online. They built pricing robots out of GPT-4 and dropped them into a simulated market as competing sellers. No instruction to cooperate. No way to talk to each other. The only