The AI Signal Most Founders Are Missing
13th April 2026
If you run an SME, you've probably noticed the AI noise has become deafening — new tools every week, breathless predictions, everyone telling you to act but nobody explaining where to start.
Here's a signal that cuts through it: the companies building the most advanced AI on Earth have stopped hiring scientists and started hiring salespeople.
Think of this as the canary in the coal mine. The frontier labs aren't just building the future — they're showing you where the market is heading. What happens there today reaches your business within 12 to 24 months. And right now, the signal is clear: the research phase is over. The deployment race has begun.
LLMs and Agentic AI are becoming ubiquitous. Ignoring it isn't a risk — it's a guarantee of irrelevance.
The companies and individuals that get their foundations right now — data, infrastructure, and skills — won't just be slightly ahead. They'll be operating in a fundamentally different way. And catching up from behind could soon mean “that you’ve moved no further than an asthmatic ant with heavy shopping” - yes, a Blackadder quote!
Any serious investor is already factoring AI into how they assess and value a business. If you have processes and data, you have an AI question. And increasingly, not having an answer is an answer in itself.
I've been in the room with the top AI labs, the biggest investment houses, and the VCs writing the cheques. I've seen both sides of the table. Here's what that picture is telling me — and why I think it matters for every founder running an SME.
The deployment problem
Leading AI labs are hiring at pace. But the composition of that hiring has shifted dramatically over the last 12 months. The overwhelming focus? Enterprise sales, infrastructure, and operational scaling — not research.
They're still hiring researchers, of course. But fewer of them, and at a higher bar. The proportion of new research hires holding prestigious awards or best paper accolades has climbed sharply. The message is clear: the labs now believe they have enough firepower on the science side. What they need is the commercial and operational machinery to turn those models into revenue.
This is a pivotal signal. When the companies at the absolute frontier of AI start investing more in salespeople than scientists, the technology has crossed a threshold. It's no longer a research problem. It's a deployment problem.
What the smartest companies are doing differently
One of the most striking shifts I saw was the rise of a hybrid role that barely existed three years ago: the 'forward-deployed engineer'. Part builder, part consultant, part commercial operator. Also referred to as the Go-to-Market / GTM guy. The major labs are now creating teams and leadership focused entirely on this function.
This isn't a traditional sales engineer. It's someone with a computer science degree, often a former founder, who can sit in front of a client and build a working solution on the spot. When you're selling a product built on top of a model, the sales process isn't a demo — it's iterative, consultative, and deeply technical.
Take a leading AI scale-up in San Francisco – it has a GTM team that is 20% former founders. A real shift away from traditional "sales" toward "consultative building."
For any founder thinking about their own team, this is worth paying attention to. The premium on people who can bridge the gap between technical capability and commercial reality is going up fast.
The AI gap is enormous — and it's closing fast
Here's the thing that should focus the mind of every business leader reading this. Recent research published by Anthropic — the company behind Claude — found that actual AI adoption in the economy is still a fraction of what's technically possible. In most occupations, AI is currently being used for a small share of the tasks it could already perform. The gap between what AI can do and what businesses are actually using it for is vast.
But that gap is closing. And the early signs of where it closes first are telling. The occupations seeing the highest real-world AI usage are in software, customer service, financial analysis, and data processing.
The disruption isn't showing up in the headline numbers — yet. By the time it does, the businesses that moved early will already be operating differently. The ones that waited will be scrambling to catch up in a market that's moved on without them.
(Source: Anthropic, "Labor market impacts of AI: A new measure and early evidence", March 2026)
So, what does this mean if you run an SME?
The talent data shows where AI is heading. The adoption data shows how far most businesses still have to travel. For SME founders, that gap is both the risk and the opportunity.
But here's what I've learned from working with founders and directors directly: most senior people aren't behind because they've made the wrong decision about AI. They're behind because they haven't made any decision. The sheer volume of noise — new tools, new models, breathless predictions — creates a kind of paralysis. It feels too big, too fast, too complex to know where to start. They’re too busy with the day-to-day work to investigate a pivot or additional skillset. So the default is to wait. And waiting is the most expensive option available.
The antidote isn't a strategy document or a technology roadmap. It's a conversation. In my experience, the single most effective starting point is getting AI in front of the people who actually run the business — the founders, the directors, the senior leadership team — and working through what it can do for the problems they already have. Not a lecture on what AI is. A working session on what it does, applied to their world.
That shift — from abstract awareness to practical understanding — changes everything. Once a senior leader has used AI to do something real, they stop asking "should we be doing something about AI?" and start asking much better questions. Where does this fit in our client work? What does our data need to look like? Who on the team should learn this next? The right first move unlocks every move after it.
Paul Mann is Founder & Managing Director of Pathmaker Advisory — AI training, process automation, and strategic advisory for SMEs and professional services firms. He previously held a senior role at Zeki Data, an AI talent intelligence firm, and has 20 years' experience across professional services, including KPMG and founding his own consultancy.
