Iām Osmani, Madrid-based, focused on AI Safety education. I founded AI Safety Madrid and Women4AISafety EspaƱa, and coordinate BlueDot Impact Spain and co-facilitate ENAIS/āAIS Collab Spanish cohorts. Iām interested in how we can make AI Safety education work across languages and support professionals who want to contribute but donāt know where to start.
Outside of this, I mentor at 42 Madrid and teach AI Safety/āGovernance at executive programmes in Madrid.
@gergo congrats!
Wow Tristan, Wow!
Thanks for putting this together, I read it carefully and itās incredibly useful for career planning.
A fully structured overview like this is rare.
I especially appreciated the split between career choice (aptitudes > causes) and job search (treating rejections as normal, building capital outside EA), and how clearly you highlight ops/āmanagement as a real gap.
The āno sorting hatā, āfit firstā, and āapply widely but expect slow vettingā framing is spotāon. It also matches what we see locally in terms of timelines and expectations.
Iāll be using this as a reference in intro sessions and advising conversations here in Madrid, itās one of the clearest maps of career landscape Iāve seen. š„³
Wow, great job, I really enjoyed how you integrated the five considerations into a coherent framework, especially the parts on moral seriousness and the careerādonations relationship.
I know several people working fullātime in AIS who had never considered that their donations might need to reflect other parts of their values.
I think this kind of structure would be especially useful in newer or nonāEnglishāspeaking communities (e.g., EA groups in Spanish), where many people arrive with strong intuitions about global poverty but much less exposure to longtermism.
If you ever turn this into a more pedagogical resource (a graphic, a short checklist, concrete splits like 60/ā25/ā15, etc.), it will be very helpful to use it in intro sessions here in EA Madrid and in other Spanishāspeaking groups. šŖš½
Hello @Toby Tremlett Iām planning to write about building AI Safety education infrastructure for Spanish-speaking communities.
The post would cover: why the current English-only ecosystem creates real talent bottlenecks, whatās needed beyond just running study groups (facilitator training, community infrastructure, sustainable funding models), and what Iām learning as I start coordinating EA Madrid and exploring these questions with the community.
Would this be valuable for the Forum?
Content:
More visibility for non-English-speaking EA communities and their challenges
Discussions on alternative career paths in EA beyond traditional org roles
Conversations about building EA capacity in underserved regions/ālanguages
Features:
Better ways to connect people working on similar regional/ālanguage-specific projects
Space for collaborative project development, not just finished work
Meta:
Highlight work from EA communities outside UK/āUS, thereās great work happening globally that deserves more visibility
I really appreciate you sharing this. The work youāre building speaks to me, and Iād like to apply for the role. Wishing you all the best with Impact First and the New Year š„³
SimĆ³n, gracias por abrir esta conversaciĆ³n tan necesaria.
Participo en EA Madrid y estoy estudiando con BlueDot Impact, y una de las primeras barreras que vi fue exactamente esta: la ausencia casi total de recursos en espaƱol para quienes quieren profundizar mƔs allƔ de lo bƔsico.
Coincido completamente en que no se trata solo de traducir, sino de generar contenido original que dialogue con las realidades de nuestros contextos. Cuando intento explicar AI safety o cost-effectiveness a colegas en Madrid o a mi red en Colombia, constantemente me encuentro traduciendo no solo palabras, sino conceptos que necesitan ser reenmarcados para que resuenen en nuestras realidades.
Tu punto sobre la producciĆ³n acadĆ©mica es crucial. He visto de primera mano cĆ³mo la āapariencia forĆ”neaā de EA genera resistencia en cĆrculos profesionales y acadĆ©micos en EspaƱa y LatAm. Sin un corpus robusto en espaƱol que demuestre la aplicabilidad de estas ideas a nuestros contextos especĆficos, seguiremos siendo percibidos como un movimiento importado.
Respecto a la pregunta de AgustĆn sobre sacrificar alcance: creo que hay espacio para ambos. Posts sobre experiencias de community building en contextos hispanohablantes, adaptaciones de programas como BlueDot Impact a LatAm, o anĆ”lisis de cost-effectiveness en polĆticas pĆŗblicas especĆficas de nuestra regiĆ³n son inherentemente mĆ”s valiosos en espaƱol, incluso si su alcance global es menor.
Me sumo a la invitaciĆ³n. Soy cloud architect en transiciĆ³n hacia roles de operations en EA, con experiencia en gobernanza de IA, y estarĆa encantada de contribuir escribiendo sobre:
Experiencias construyendo comunidad tech en Madrid con perspectiva EA
Barreras y oportunidades para profesionales tech hispanohablantes entrando a EA
AdaptaciĆ³n de frameworks de AI governance para contextos regulatorios europeos y latinoamericanos
ĀæSiguen activos los esfuerzos de UPB? Me interesarĆa conocer mĆ”s sobre la convocatoria de artĆculos.
This resonates deeply, especially the line: āOrganizations without clear stories hit friction, even when doing excellent work.ā
Iāve seen this in my own career transition into EA, I had the skills and the commitment, but until I could articulate why my background in international partnerships and data operations connected to AI safety and global health work, I struggled to make others see the fit.
Your framework around Mission ā ToC ā OKRs ā KPIs ā Team is brilliant because it shows that organizational storytelling isnāt just āmarketingā ā itās strategic clarity that enables faster alignment, better partnerships, and ultimately more impact.
The section on authenticity being non-negotiable particularly stands out in the EA context. Being transparent about uncertainty and limitations isnāt weakness; itās what builds trust in a community that values epistemic rigor.
Thank you for writing this. Itās a reminder that even in a movement focused on evidence and outcomes, we still need to remember weāre communicating with humans who understand the world through narrative.
Focused on scale of suffering and cost-effectiveness across my votes. Particularly excited about Latin American organizations (ACTRA, Vida Plena) given the regionās potential and underinvestment. Animal sentience weighs significantly in my cause prioritization.
Thank you for this. Career transitioner here, spent weeks applying to roles Iām marginally qualified for instead of working on what Iām actually positioned to do: AI governance consulting for EU startups navigating the AI Act.
Your point about āimmediate, direct actionā hit hard. I have a mentor meeting in 2 days for this venture. Going to stop the application cycle and focus there.
Would love to hear more about your journey if youāre open to connecting.
Thank you for this.. Karen thank you. Iāve been trying to say what you wrote for a while now.
I run an AI safety group, in Spain. I started getting into AI Safety as most people do. What really made sense to me was BlueDot ImpactĀ“s AGI Safety course. Before that I knew AI safety was a concern. I didnāt know what to do about it.
After that course I did the AI Safety Operations Bootcamp also from BlueDot Impact, because I needed to know how to do this work in a way and I couldnāt find that information anywhere else. It helped me. It also showed me how much I still had to learn.
One thing Iāve noticed is that people donāt think ops work is unimportant. It gets ignored when its done well. Someone has to keep things running so researchers can do their job. I thought about doing research myself, but honestly I think Iām more useful doing ops work. I just wish it was clearer how to do this work.
Directors of regional groups Iāve spoken with all say the same thing: the work is real and there are people doing it, but without a clear path or credentials, good people drift toward roles where that structure exists. The field loses them quietly.
I agree with AgustĆn that we need to understand the context of our work. You canāt just run a group on good intentions. It takes time and knowledge of the field.
Iām really glad this program exists.