OpenAI Scholars 2019: Applications open
OpenAI
OpenAI's opening applications for round two of its Scholars program. It's paid, mentored deep learning training for people from underrepresented groups, no CV padding required.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI wants a second batch of Scholars, and applications for the 2019 cohort are open now. The setup is simple on paper: 6 to 10 people get a stipend, three months of full-time study, and direct mentorship from OpenAI researchers, all aimed at getting them fluent in deep learning. At the end, each Scholar ships an open-source project, so the payoff isn't just a resume line — it's actual code the community can poke at.
The program specifically targets underrepresented groups in the field, which is the part that matters most here. AI research still skews heavily toward a narrow demographic slice, and pedigree — the right university, the right advisor, the right internship — does a lot of gatekeeping before anyone even touches a research problem. A stipend plus mentorship removes the two biggest barriers at once: money and access. You don't need a machine learning PhD lined up to get in the door.
This is the second run of Scholars, so OpenAI presumably learned something from cohort one about what works and what doesn't. Three months is a tight runway to go from curious to shipping a real project, but it's also long enough to force actual depth rather than a weekend tutorial binge. The open-source requirement at the end is a smart forcing function too — it means Scholars have to produce something legible to outsiders, not just a private portfolio piece.
None of this changes the field overnight. Six to ten people a year is a rounding error against the thousands entering AI research annually. But programs like this compound — alumni become mentors, hiring managers, founders — and OpenAI clearly sees it as a long game rather than a PR stunt.
My take — AI-written commentary, not fact-checked reporting
I like small, unglamorous programs like this more than I like most product launches, because they actually shift who gets to build the next generation of models instead of just who gets to use them. Six to ten stipends won't fix AI's diversity problem, but it's a rare case of a big lab spending money on people rather than compute, and that's worth applauding without irony.
Read more about this at: OpenAI