The Oscars just drew human-only lines through acting and writing.
An Uber co-founder just raised $1.7B to coordinate AI across mining and construction.
Same coordinating skill. One institution bounds it, capital funds it, nobody defines it.
I just published a paper naming it. 🧵
Orchestrator Awards: standards, research & recognition for human AI orchestration.
AI-native cinema: first governed proving ground.
Founded by Mark Sendo — Celestial Technologies LLC. Closed to submissions & payments.
https://t.co/7CxQjWy32W
#OrchestratorAwards#AI#Research
Issue 002 of the Orchestrator Awards research track: “The Orchestrator as Creative Profession.”
https://t.co/ghcvZg88Tw
Disagree with the three tests? Reply — it’s a working paper, revised in public.
The Oscars just drew human-only lines through acting and writing.
An Uber co-founder just raised $1.7B to coordinate AI across mining and construction.
Same coordinating skill. One institution bounds it, capital funds it, nobody defines it.
I just published a paper naming it. 🧵
Nobody defines this profession. No guild. No degree. No standards body.
Vacancies like this get filled either by institutions that understand the work — or by labels that shrink it. “Prompter.” “AI content creator.” Professions named badly at birth keep the name for decades.
And it’s not a film thing. The $1.7B Atoms raise (a16z-led) is funding coordinated AI in mining, construction, heavy transport. Orchestration is crossing industries. Cinema is just where it’s provable frame-by-frame.
This profession has no credential gate.
I built a film production platform with zero coding background. Students are shipping work that used to take studios. The entry requirement is judgment, evidenced in finished work
Three tests for orchestrated work:
https://t.co/5SiEKyL3Vr systems carried real production responsibility
2.A human made the calls at every boundary — handoffs, standards, recovery
3.Someone can explain every decision in the finished work
One AI tool + cleanup ≠ orchestration.
The role: the Orchestrator. Not a prompter. Not an engineer. The person directing AI systems — models, voices, agents, workflows — toward one finished work, and accountable for all of it.
@DarioAmodei@bcherny@sama Dario, anthropic’s recent study on agentic coding missed a major flaw, using Claude’s output to a humans sector expertise vs Claude to Claude or GPT to Claude, etc
@AnthropicAI Clockwork. Peak hours, Claude faceplants. No comms. No spine. Just a late “we’re aware” while paying customers get unpaid-tier reliability. Embarrassing. Do better. #ClaudeDown#AIOutage@sama@elonmusk every day I encounter this BS
@AnthropicAI Clockwork. Peak hours = Claude faceplants. Every. Single. Time.
No status update. No spine. Just a "we're aware" reply two hours late like that fixes anything. You’re not a scrappy startup anymore — you’re a paid service with unpaid-tier reliability.
Embarrassing. Respect your users’ time, or at least pretend to.
#ClaudeDown #ClaudeFail #AIOutage