Head of Public Understanding
People have real relationships with AI now — for company, for advice, sometimes on the hardest night of their lives. It's happening at scale, and most platforms have no way to tell whether a given conversation is helping someone or quietly harming them. At the sharp end that's someone in crisis nobody notices. In ordinary use it's dependency building over weeks, or a person slowly reshaped in how they think and who they lean on. Regulation is moving slowly, and people fall through the gap.
We're a small team building safety infrastructure for exactly this: open-source classifiers any team can deploy today, proprietary ones for those who need more, and advisory work through NOPE Labs. We have a point of view, an incident database, a live regulation tracker across the EU, UK and US, taxonomies, and original research. What we don't have yet is a public voice carrying any of it. The conversation about what AI is doing to people in everyday use is still happening in the wrong places, in the wrong language.
Head of Public Understanding is the person that voice belongs to. You'd build and lead NOPE's presence in the public conversation about AI safety, and the community of people who care about it: builders, researchers, journalists, policymakers, and the ordinary users who sense something is off. In practice you run our channels and write what goes on them, turn our research and trackers into things people actually read and argue about, and show up at the webinars and events where this is happening. Over time you become a recognizable voice on the subject and bring an audience with you — building a position in public, not babysitting a forum.
We're not an anti-AI project, and we're not cheerleaders either. These relationships are real and won't reverse; the honest response is to make them safer with evidence, not to tell anyone to log off. We want someone healthily cynical and genuinely opinionated, who knows the technology well enough to see its limits. The shape of the role is open, like the rest of what we do — we're not handing you a script. The work is mostly async: you'd share what you're trying and what's landing, and we'd work it out together.
You'd be a good fit if
- You care, deeply, about relational safety and the future of how humans and AI interact
- You're already writing and posting about the limits of these systems and how they show up in everyday conversational AI
- You have a following and a network — X, Bluesky, Substack, LinkedIn — around AI, and especially AI safety
- You write like a person, and you don't lean on AI to think for you
- You understand in broad terms how LLMs are built and where they fall short, and how the companies making them work on the inside
- You follow how policy and policymakers move on new technology, and you're comfortable representing a point of view at an event, not only on a page
Bonus: crossover with trust & safety, policy, healthcare or law — or the rarer background of a builder or engineer who moved into writing and knows from the inside what can really be done about the risk.
This starts part-time on a six-week initial contract, so we can both see whether it's a fit; there'd be more work, and potentially a full-time role, as we grow. Send us a piece of writing you're proud of and a sentence on what you think the AI safety conversation is getting wrong.
Apply
Tell us about yourself. No cover letter needed—just the basics and anything you'd like us to know.
We read every application. If there's a fit, we'll reach out within a week or two. If not, we'll let you know—no ghosting.