Bill Gates published a long, sober essay this week arguing that the move into the AI era will be one of the most turbulent periods in human history, and that right now we are not preparing for it. The line he keeps coming back to is that the world needs a plan, and there isn’t one yet.
He is writing about society, about jobs and inequality and security and the effect all of this has on children and on the way we think. Those are the right things to be worried about, and we are not going to pretend a technology consultancy has the answers to any of them.
It’s already inside your business
But there is one part of his argument that lands squarely where businesses actually operate. Gates points out that this transition is different from the ones that came before it, because unlike the PC we don’t get to adapt to AI at our own pace. It adapts to us instead. It runs on the devices we already have and it speaks our language, which means it is already inside your business whether you planned for it or not. There will be a tool being used in one team, someone quietly coding with it in another, and a pilot running in a corner that nobody has really joined up.
“The question was never whether to start with AI, because it already started, in pockets, without anyone signing it off. The real question is whether you have a plan for what’s already happening, and by that I don’t mean a maturity score or a slide full of ambitions. I mean an honest read on where you genuinely are, what’s actually worth doing, what it’s worth in real terms, and the order things need to happen in.”
Matt Feeney, Chief AI Officer, Enablis
That planning gap is the thing worth acting on, and it is the one part of the picture Gates paints that you don’t have to wait for a government to fix.
Security has to hold whatever the model does
There is a second point in the essay that we would want to underline for anyone working in technology. Gates writes that the cybersecurity experts he knows are worried, because the attackers are gaining capability faster than the defenders can close the gaps, and the same model that finds a flaw so you can fix it will just as happily help someone else exploit it.
“At Enablis, we’ve worked hard to build deterministic controls - the parts of the platform that behave exactly the same every single time, so the guardrails hold whatever the model decides to do on a given day. That’s what makes security a first class citizen rather than a review at the end. It means the safe path is the default path, and teams get to move quickly precisely because the controls underneath them aren’t up for negotiation.”
Ed Marshall, Chief Technology Officer, Enablis
The jobs picture is shifting, not disappearing
The tone we would borrow from Gates, though, is his grounded optimism. He is careful not to collapse into doom, and neither should we. A good deal of the essay is about jobs, and it is worth being precise about what is actually happening in the market rather than assuming AI simply erases roles.
“In the tech jobs market, the demand we’re seeing hasn’t dried up, it’s shifted. What’s changing is what good looks like. Software engineers who can genuinely build with these tools are the ones everybody is trying to hire, and demand for them has risen sharply. Junior roles are the real question - if AI absorbs the work new technologists used to learn on, we lose the pipeline that produces the senior engineers everyone is competing for, and tech needs a deliberate answer to that.”
Martin Jones, Managing Director, Fruition Group (Enablis’ sister company)
The businesses that come through in the best shape
So the businesses that come through the turbulence in the best shape won’t be the ones with the most AI, and they certainly won’t be the ones that tried to hide from it. They will be the ones who knew where they stood, knew what to do next, understood what had to be true before each step could start, and backed the people who can actually build with this technology.
It is the whole reason we work the way we do at Enablis. We start with an honest read on where a business actually is with AI, we turn that into a plan it can act on rather than a score it files away, we build security in from the start rather than bolting it on later, and we map the dependencies so you know the real order of things instead of just a wish list.
Gates is right that the turbulence is here, and he is right that the answer is a plan. That part, at least, is already within reach.
