I moved from OpenAI to Meta Superintelligence Labs. Here are my tips for getting hired at a top AI company. (www.businessinsider.com)

Prakhar Agarwal, an applied researcher who moved this year from OpenAI’s API team to Meta Superintelligence Labs, shares practical advice for landing roles at top AI labs. He emphasizes that these teams hire for high-autonomy work: small headcounts, rapid problem selection, and ownership of research-to-product decisions. Experience helps — many roles go to people who’ve already shipped difficult projects — but the hiring focus is on your ability to identify real gaps in large language models (LLMs), propose metric-driven solutions, and operate in ambiguous domains rather than follow rigid directives. His concrete tips are technical and tactical: know LLM nomenclature and fundamentals, keep coding skills current in job-relevant scenarios, and use models frequently so you can spot failure modes and quantify gaps (e.g., what Llama should improve next, and by how much). Demonstrate trend awareness — what capabilities will emerge in 3–6 months — and strong, high-bandwidth communication (whiteboard sessions, concise one-to-one discussions). A Ph.D. signals comfort with abstraction, but equivalent hands-on contributions at startups or in product engineering can substitute. Finally, learn outside formal coursework: read blogs, follow researchers on social media, ask questions, and build intuition by doing.