Nobody Is a True AI Expert Yet, and You Should Be Suspicious of Anyone Who Claims Otherwise

If you’ve spent any time looking for a development partner to help you build an AI-powered product, you’ve probably noticed something: everyone is an AI expert now.

Agencies that were building Shopify stores two years ago are leading with AI practices. Consultancies have rebranded entire service lines. LinkedIn is full of people who have been “AI specialists” since last Tuesday.

We think this is worth naming directly, because it affects how you should evaluate anyone you’re thinking of building with.

The field is genuinely new

Practical, production-grade AI development, the kind where you’re building real features into real products used by real people, has a meaningful history of roughly two to three years. The models that made this possible came out in late 2022 and early 2023. The tooling, patterns, and hard-won lessons around how to build with them reliably have accumulated since then.

That’s not long enough for anyone to have twenty years of experience. It’s not long enough for university programs to have produced graduates who specialised in it. It’s not long enough for the “best practices” to have fully settled.

Anyone presenting themselves as a deep, established expert in production AI development is either stretching the truth or talking about something narrower than you probably need - like a specific framework or a specific model family, not the whole problem of building products that work.

What experience actually looks like in this field

The teams doing the most credible work in AI product development right now are the ones who have been building with it continuously since the tooling became viable - learning what breaks, what doesn’t, what the models are genuinely good at, and what they confidently claim they can do but can’t.

That experience is real, and it matters. But it looks like two or three years of intensive, hands-on building, not a decade of mastery. The honest version of expertise here is: we have been doing this since it became possible, we have shipped things with it, and we have made the mistakes that teach you what not to do.

We can say that about ourselves. We have built AI features into our own products, My Chamber Buddy and My Member Buddy, and applied them to operational problems in industries we know well. We didn’t do it for a demo. We did it because we wanted to see what actually works when real users rely on it.

That’s a different thing from claiming mastery of a field that is still being invented.

What this means for your product

If you want to build a product with AI in it, the right question to ask a potential partner isn’t “how long have you been doing AI?” It’s closer to: what have you actually shipped, what broke, and what did you learn?

The teams worth working with will have honest answers to all three.

They’ll also be clear about where the edges of the technology are today. AI is genuinely powerful, but it has real limitations - around reliability, around cost at scale, around the kinds of tasks where it adds value versus the ones where it creates more problems than it solves. A team that tells you AI can do anything is not a team that has built much with it.

The advantage of building now

Here’s the upside: because this field is still emerging, the gap between a team that has been building with AI for two years and one that picked it up six months ago is significant. You don’t need decades of expertise to find a team that knows substantially more than everyone else - you just need to find one that has been doing it seriously and continuously.

The other upside is that the businesses that build AI-powered products now, while the field is still finding its shape, are the ones that will have a meaningful head start in three years. The patterns, the user behaviours, the integrations that work - these are all still being established. Getting in now means helping to define them, not trying to catch up later.

What we actually offer

We build products. We’ve done it for our own software and for clients across hospitality, associations, and operations-heavy businesses. We’ve been building with AI since the tooling became viable, and we apply it where it genuinely improves the product, not where it looks impressive in a pitch.

We won’t tell you we’re experts in a field that doesn’t have established experts yet. We’ll tell you we’re deep in it, we know what works today, and we’re honest about what doesn’t.

If that sounds like the kind of partner you want to build with, let’s talk.