About
AI built around expertise, with human judgement at the core.
AI products and writing, built on one idea: AI in service of human expertise, however good it gets.
What you can expect
Straight answers
We do not force AI into problems where it does not belong.
Human-centred design
The goal is better judgement and performance, not technology theatre.
Measured progress
Every engagement is scoped to move a decision forward and de-risk what comes next.
Why this exists
Most AI problems are organisational before they are technical.
The recurring pattern is familiar: a capable system gets built, but people do not trust it, use it, or maintain it. We work upstream of that failure mode.
Most AI projects don’t fail because the technology doesn’t work. They fail because the people building them didn’t understand the organisation they were building for.
The pattern repeats everywhere: technically sound systems that nobody trusts, uses, or maintains, because they were built around data models and process diagrams rather than the actual expertise of the people doing the work. The result is AI that looks good in a demo and stalls in production.
AI For Human Expertise exists to work out the alternative in public. The starting point is always the expertise people already have: what they actually know, including the things that have never been written down, and how AI can make that expertise more effective, more scalable, and harder to lose.
That idea plays out in three ways. The writing is where the ideas get developed and tested against evidence. The products are where they get built into tools people can use. And a small amount of advisory work keeps the thinking honest by testing it against real organisations and real decisions.
Who's behind this
Sammy Newman, Founder
Sammy Newman is the founder of AI For Human Expertise, where he writes on designing AI that strengthens human judgement instead of eroding it, and is building products on that idea.
Connect on LinkedInPrinciples
How we approach the work.
These principles are practical constraints, not marketing copy. They shape how engagements are scoped, delivered, and judged.
Human-first
Technology serves people, not the other way round. Everything we write and build starts from the judgement of the people doing the work.
Pragmatic
We focus on what works, not what's fashionable. If AI isn't the right answer for a problem, that is the answer we give.
Measurable
Claims should be checkable. That standard applies to our writing, our products, and anyone we advise.
Honest
We won't oversell what AI can do. The value of this work comes from getting it right, not from closing a sale.
Stay in the loop
Follow where this goes
The waitlist is the best way to hear when products ship, plus occasional writing on AI and human expertise.
The waitlist opens soon. In the meantime, new work is announced on LinkedIn , or you can leave your details and we will let you know directly.