OpenAI Hits Automated Research Intern Goal — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI said it hit its 'automated research intern' goal, with researchers now averaging 3.1 agent-workdays per human workday and top users spending $7,000+/day on tokens.
- OpenAI is on track for a fully autonomous AI researcher 18 months from now, targeting March 2028, per researcher Andrew Curran summarizing the blog post.
- Agent workloads at OpenAI concentrate on writing research and infrastructure code, technical help and review, and monitoring training, RL, and eval runs, per researcher Christopher Ong.
- Simon Willison flagged a huge uptick in token spend in mid-July, speculating on X it coincided with Astra becoming available to OpenAI employees.
- Charlie O'Neill noted that the median OpenAI researcher was spending $0/day on coding agents as recently as February 2026, underscoring how steeply adoption has scaled.
- Kevin Liu framed the data as transparency into 'recursive self-improvement,' calling it potentially the most important contributor to AI capabilities over the next few years.
Why it matters: OpenAI's top researchers now spend $7,000+ daily on tokens versus $0/day at the median just seven months earlier in February 2026, illustrating explosive agent adoption inside the lab. Researcher Andrew Curran states the company is on track for a fully autonomous AI researcher by March 2028, making recursive self-improvement a concrete roadmap item rather than theory.
Ask SkimNews
