OpenAI Fellows Fall 2018: Final projects
OpenAI
OpenAI's second Fellows class just finished a 6-month program turning ML beginners into real contributors. Applications for the next round are already open.
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 just closed the book on its second Fellows cohort, and the pitch is simple: take people who aren't machine learning experts yet, put them through six months of real work alongside OpenAI's research teams, and see what comes out the other side. Judging by the program's own framing, what comes out is core contributors, not just interns who shadowed a few meetings.
The Fellows program has always been a bit of a bet on talent over pedigree. You don't need a PhD or years in a lab to get in; you need aptitude and the willingness to be thrown into deep water fast. Six months is not a long runway to go from novice to someone shipping work that matters at a frontier AI lab, but that's the whole premise here.
OpenAI didn't publish a long list of specific projects or breakthroughs in this announcement, which is honestly typical of these wrap-up posts — they're more about signaling that the pipeline works than about showcasing individual output. The real message is operational: round two is done, round three is already accepting applications on a rolling basis for Summer 2019.
What's notable is the cadence. Two cohorts in and OpenAI is treating this like an ongoing institution rather than a one-off experiment. For an org that's often associated with splashy model releases, quietly building a farm system for future researchers says something about how they think about compounding talent over time.
My take — AI-written commentary, not fact-checked reporting
I like this kind of program a lot more than another benchmark press release — it's OpenAI admitting that talent pipelines matter as much as GPU clusters, and that's a healthier long-term bet than chasing hype cycles. My only gripe is the total lack of transparency on what these Fellows actually built; if you want credibility for this pipeline, show the work, not just the recruiting funnel.
Read more about this at: OpenAI