Generative AI Fundamentals Map

Map/Agents and organisations/Wave 10

Agents designing agents

Agents build new agents, organisations and evaluations.

stageSpeculative Open on the map →

Facts

Lane: Agents and organisations
Wave: 10 of 12 (build order, not a date)
Needs: 21 boxes all the way back through hard links
Unlocks: 1 boxes downstream through hard links
Leverage: 0.1 points · stands in the way of 1 of 53 stages · rank 54 of 57
Cluster: Oversight and alignment
When: Researcher estimate: 2029 to 2038

Evidence as of October 2026

The later stages of AI 2027, where AI research is mostly automated and progress compounds.

01AI 2027 takeoff forecast

Watch
Agent-designed architectures or training methods adopted by labs.

Timeline note
My estimate, following automated AI research.

Around it

Needs · 4

Superhuman AI researcherstage · Speculative
Agent-native organisationsstage · Emerging · helps, not required
Scalable oversightcapability · Emerging · helps, not required
Alignmentcapability · Emerging · helps, not required

Unlocks · 3

Long-latency civilisationstage · Speculative
Superintelligencestage · Speculative · helps, not required
Human-agent symbiosisstage · Speculative · helps, not required

Loops and off-ramps

Capex pullbackoff-ramp · cut off
Taiwan chip disruptionoff-ramp · cut off
Power shortfalloff-ramp · cut off
Capability plateauoff-ramp · cut off
Training data ruled outoff-ramp · slowed
Incident-driven haltoff-ramp · slowed
Misuse catastropheoff-ramp · slowed
Alignment failure in deploymentoff-ramp · cut off
Public backlashoff-ramp · slowed

Openings researchers' judgement, October 2026 scouting

Startup room 40/100 · rank 61 of 124 Research room 54/100 · rank 51 of 124

Crowding2 Lab overlap4 Small-team fit4 Buyer today2 Research openness4

Depth: light
Players: ADAS (Hu, Lu, Clune); Autoheal; lab scaffolding teams
Wedge: Agent-design search that generalises at affordable evaluation cost, a natural fit for metaheuristic search. research

01ADAS (arXiv 2408.08435)