You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
$290,000 left to reach the funding goal.
I run Simple Programmer, a career brand for software developers (blog, YouTube, courses; the books Soft Skills and The Complete Software Developer's Career Guide; 55+ Pluralsight courses). I want to point that audience at a real bottleneck in AI safety: experienced engineers who can build evals, infrastructure and tooling for safety teams.
"Developer to AI Safety Engineer" is a 12-month program (Dec 2026 to Nov 2027):
A free online course, "From Software Engineer to AI Safety Engineer", plus a public YouTube/blog series for the Simple Programmer audience.
Two 10-week mentored cohorts of about 25 senior engineers each. Every participant ships a capstone contribution to an open-source safety project or eval suite.
Hands-on placement help into safety orgs, fellowships (MATS etc.) and lab safety teams.
Plain statement of my main weakness: I'm not an AI safety researcher. That's why a big chunk of the budget pays for a part-time technical advisor from the safety field who owns the curriculum's accuracy. What I bring is 15 years of reaching working developers and moving them into new careers.
Goal: more experienced engineers doing useful AI safety work. There are decent on-ramps for students and researchers, but very little aimed at mid-career software engineers, who are scarce on safety teams and can contribute quickly.
Months 1-3: recruit the technical advisor, build the curriculum (ML fundamentals for SWEs, threat models and why safety matters, eval design and harnesses, interpretability tooling, safety-relevant infrastructure, how to get hired), publish the free course and the video series.
Months 4-6 and 8-10: two cohorts, weekly mentoring, each person ships a capstone to an open-source safety project.
Ongoing: application coaching and warm intros to hiring orgs and fellowships.
2,000+ free course completions
50 cohort graduates
40+ merged open-source safety contributions
15+ graduates in safety roles, paid fellowships or funded independent projects within 6 months of finishing
Mainline $290,000 for 12 months:
$90k: my stipend as program lead at 0.5 FTE, including self-employment tax
$60k: part-time AI safety technical advisor / curriculum lead
$40k: cohort mentoring and placement coordinator
$35k: course and video production
$25k: participant scholarships and compute credits for capstones (50 x $500)
$10k: outreach to senior engineers
$5k: course platform and software
~$25k: buffer
Minimum $120,000: the free course plus one cohort, with the advisor at reduced hours. Budget sheet: https://docs.google.com/spreadsheets/d/16m0xdw77tkYjq4DD-d9QHQ0XDZymeGJNygFC-5vsl8Y
Me, John Sonmez (LinkedIn, simpleprogrammer.com). Founded Simple Programmer in 2013; its blog, YouTube channel and courses have reached millions of developers. Author of Soft Skills (Manning) and The Complete Software Developer's Career Guide. 55+ courses for Pluralsight. I run Rockstar Developer University, a developer career-coaching program built on the same course + cohort + placement model this project uses. I also run the Bulldog Mindset YouTube channel, and I run my businesses with a self-built system of about 7 autonomous AI agents (email, marketing, video, finance), so I deal with the oversight problems of agentic AI every day.
To hire with the grant: a part-time technical advisor from the safety field and a cohort/placement coordinator. Neither is hired yet; recruiting the advisor is the first milestone.
Can't recruit a strong enough technical advisor, so the curriculum isn't credible to safety orgs. I won't run cohorts until that's solved.
Graduates can't find safety roles because the hiring pipeline is narrower than I think. Fallback outcome is still a free course and a set of real open-source contributions.
Alumni drift into capabilities work. Mitigated by curriculum framing, selecting for motivation and only facilitating placement into safety-focused roles.
Some of my earlier products and channels didn't reach scale. I learned to validate demand with free content first, which is why this starts with a free course.
$0 raised for this project; I've had no prior grants. Pending applications for this same project (if one funds it I'll reduce or withdraw the others): EA Funds Transformative AI Fund ($290k), Coefficient Giving capacity-building RFP ($290k), Lightcone Commons ($290k), Middle Powers & Transformative AI RFP ($390k, international version), BlueDot Impact Rapid Grant ($20k, first 3 months), Halcyon Futures intake, and FLI Digital Media Accelerator ($25k, video series pilot only). Separately, a personal Basis Fellowship application ($50k, not tied to this project).