The Parable of the Agentic Chaos Monkey
What Software Thinks of Itself (Circa 2025)

The Monkey Tempts (Anthropic, March 2025)

From the White Coat to the Cloak

Fear the Undisciplined Monkey (HuggingFace, 2026)

Defense Against the Dark Arts (CHI, 2026)

Agents, loops, and superoptimization
With programmable harnesses
Agents
Claims about agents
- Transformative, if true.
- Frontier models are extremely powerful, but impossible to evaluate outside of the labs themselves.
- Can't trust labs: economic incentive to hype the tech.
- Safe claim: agents sometimes work.
- Question: can one turn "sometimes" into "often" or "almost always"?
This talk: agents, loops, and superoptimization
Agents and Agent Loops
Loops: why care?
- Vulnerability investigations (ex: find cybersecurity flaws in X program)
- Optimization campaigns (ex: make Y program faster)
- Research synthesis (ex: synthesize survey of field Z)
- Long errands with tools (ex: make McCoy CEO of Anthropic)
Hubris: is it possible to vibe code a GPU compiler?
accy, an accelerator compiler
Yes, accy works on at least a few examples.

Causal attention decoding
