Head of Audience

Vals AI•San Francisco, CA

1d•$175,000 - $200,000•Onsite

About The Position

Vals is building the work that closes the gap between the rapid advancement of AI models and the understanding of them. We run rigorous benchmarks trusted by frontier labs and the AI research community. We are hiring a Founding Content Lead to translate this research into the broader conversation about AI. This role will own the content function end-to-end, setting the editorial voice and building the audience from the technical community to a wider audience that needs to understand AI. The position involves working directly with the CEO, engineering and research teams, and the designer. The goal is to create content that is read, cited, and remembered, moving beyond diluted and recycled AI content. Key responsibilities include building and running the content function (essays, model-release breakdowns, methodology explainers, original analysis), defining its cadence, distribution, and growth. The role requires setting an editorial voice that is technically credible for ML engineers and accessible for smart non-experts. A critical aspect is shipping analysis within hours of a frontier model drop, making Vals' insights part of the immediate conversation. The role also involves building the content function by hiring a team, establishing a playbook, and continuously raising the quality bar.

Requirements

A track record building or operating a content brand at scale — paid newsletter, publication, podcast with measurable reach, founding content function at a startup, or a personal Substack you grew to thousands of engaged readers.
AI fluency. Not necessarily an ML researcher, but you can read a model card, follow a methodology paper, and call out when a benchmark is gamed. You have opinions about which evals are credible and why.
Speed under news cycles. You've shipped published work within 24 hours of a major industry event and can do it again.
A distinctive editorial voice — sharp, opinionated, allergic to generic AI commentary.
Ability to work in-person in San Francisco (preferred — open to discussion for the right candidate).

Nice To Haves

Operator instincts. You pitch and execute without detailed briefs. Likely a former founder, early-stage operator, or content lead at a small team.
A personal audience of 1,000+ engaged readers on Substack, LinkedIn, X, or equivalent.
A technical background (CS, EECS, math) you've translated into writing.
Existing relationships in the AI research community, at frontier labs, or with enterprise AI buyers.
Visual instincts — can sketch a chart before handing it off to design.

Responsibilities

Build and run the content function — essays, model-release breakdowns, methodology explainers, original analysis — and the cadence, distribution, and growth that make it land.
Set the editorial voice. Technical enough that ML engineers at frontier labs trust it; sharp enough that smart non-experts can actually use it.
Ship on news cycle. When a frontier model drops, our analysis is part of the conversation within hours.
Build the function. Hire the team, set the playbook, raise the bar over time.