HCS ▋

I'm a mentor for a few going into cybersecurity (1 about to do Masters after CompSci-Psych degree, 1 currently doing a Masters in GRC, 1 a career changer dev doing a bootcamp going to SOC). The subject of getting work experience has come up. What advice can I pass along? They are based in Europe. We are working on some targeted topics (1-2 hr calls per month) mainly focusing on CV, upskilling, and some discussions of what industry is like.

Tools around agentic AI and LM engineering that I have been using for the past two months: OpenAI, Pisjover, Resend, Langchain, Serper, Polygon... now just got a Google Gemini API key for some tinkering.

HuggingFace was down earlier. So, two things noted for resiliency are usual SDLC - fallback on supplier downtimes, and ensuring to specify pipelines rather than going to a supplier default. PS: if you are on it, my profile is at https://huggingface.co/hcs0

Playing around with Google CoLab and HuggingFace pipelines for translation examples, sentiment analysis etc Learnt yesterday you can spinup instances on CoLab and run Jupyter notebooks on it. You can spin up a free instance of T4 to run some pipelines from HuggingFace. My profile at: https://huggingface.co/hcs0/

One thing I learned tinking with CoLab and HuggingFace - acceptance of using certain models. i.e Meta Llama, Google Gema requires approval on HF platform prior to using it with acknowledging their license. Makes you think about resiliency or backup if there are license issues.

ERA, in partnership with the Oxford Hardware AI Governance lab and the Heron AI Security Initiative, is now accepting applications for our new Frontier AI Security Residency: an 8-week, fully-funded residency for researchers & engineers working to build the security backbone for frontier AI systems. https://airtable.com/appX4HJtse5Ck8YbI/pagTn8kHWFL3Q4sHx/form