Anthropic Fellows: a paid AI safety research on-ramp, no degree

My notes

Summary

A TikTok creator promotes the Anthropic Fellows program, a paid pathway into AI safety research that requires no degree and no prior AI experience. The substantive detail: it pays roughly 3,850 USD/week for four months (~61,000 USD), and selects for technical curiosity and troubleshooting ability rather than credentials. The video itself is engagement-bait (“comment AI”) wrapped around a real program.

Key Insight

• The real program (Anthropic Fellows) targets people who like to tinker and find vulnerabilities in software, not credentialed ML researchers.
• Stated comp: ~3,850 USD/week for 4 months, framed as ~61,000 USD annualised. It is a fixed-term fellowship, not a salaried role.
• Selection bar described as “technical curiosity and troubleshooting” - i.e. debugging/problem-solving instinct over a CS degree or published AI work.
• The angle worth noting: an AI lab actively creating an on-ramp into the field while the dominant narrative is “AI is taking jobs.” That contrast is the only non-obvious takeaway here.
• Treat the creator’s claims as unverified marketing. The “comment AI for the link” mechanic is lead-capture, not a real shortcut - the program is publicly searchable.