Technical AI Safety Specialist

# Technical AI Safety Specialist Get exceptional technical talent working on the most important problems in AI safety.

Apply Now
1. Home2. Work with us3. Technical AI Safety Specialist##
** Who is Blue Dot Impact?
** We're a nonprofit startup
** building the workforce that protects humanity**.Since 2022, we've trained over 7,000 people in AI safety and biosecurity. Our alumni work at Anthropic, Deep Mind, the UK AI Safety Institute, and dozens of other organisations shaping how transformative technologies affect the world.

We run courses, offer 1-1 advising, train over a 1,000 people a month, and deploy over $1M a month in grants. We're a small team headquartered in San Francisco. We raised $25M from Coefficient Giving in 2025.##
** Why this role matters
** There are a huge number of open technical problems we must solve to make AI go well. Model evaluations are still brittle, jailbreaks are still common, we still can’t reliably align models with human values and have even less idea how to do this for super-intelligent systems.

The number of people contributing to these problems is in the low hundreds, and many people who want to do this lack the strategic clarity to actually push the needle on them. There is also a capacity problem. There aren’t many safety organisations and open roles, and there aren’t enough places on fellowships to absorb the talent.

Your job is to tackle both problems — and Blue Dot puts you in an unmatched position to do it. Drawing on one of the largest talent databases and alumni networks in AI safety, you'll identify and accelerate top technical talent through advising and strategic clarity, while building programs and deploying significant capital to support them full time. We're taking an especially ambitious approach in the months ahead, with over $5M in grants in the next year as we scale our programs and grantmaking to new heights.##

** What you'll do
**** Shape how the field thinks
*** Write publicly about technical AI safety strategy for large audiences: canonical guides for newcomers, career guides, and strategy docs for researchers.
* Build a platform so your reach compounds over time.
** Get the best people up to speed
*** Run our technical AI safety course and project sprints. Make them sharper, faster, and more useful.
* Test new formats — research bootcamps, mentorship intensives, lab-specific bespoke training. Whatever works.
* Work alongside leading researchers to keep the content world-class.
** Place them where they'll have impact
*** Do one-to-one advising calls with the strongest people in our pipeline. Give them opinionated guidance on what to work on and where to apply.
* Cultivate relationships with hiring managers at labs, evals orgs, and independent research outfits.
* Help mobilize researchers towards areas you think are neglected.
* Track where graduates end up and whether they're actually shipping safety-relevant work.
** Deploy capital to unblock people
*** Review and decide on rapid grants and career transition grants for technical talent.
* Identify people who could be doing important work but are stuck on funding — and unstick them in weeks, not months.
** Decide what TAIS actually needs — and do it
*** Map what's being done in technical AI safety and where the bottlenecks are. Which problems are starved of talent? Which orgs can absorb more researchers? Where is the leverage?
* Talk to leading researchers and lab safety teams every week. Form strong, specific opinions.
* Think from first principles about the low-hanging fruit that the field needs, whether it be fellowships, incubation programs, technical research collaborations, hacker houses, or something else the field hasn’t ever seen before. Build and run these programs end-to-end.

A strong first six months looks like owning the next iteration of our technical AI safety course, running ~50 one-to-one advising calls, and shipping at least one new program format we've never tried before.##
** About you
** We're looking for someone with deep technical AI safety chops who's energised to move from doing research to building the field around it. You might be a great fit if you:
* ** Have a strong technical AI…