Conversations about where AI is headed and who decides its value.
Over the last couple of weeks I've had a number of conversations about where AI is actually headed. Two of them stuck with me. The first was with a technically-solid ex-colleague who told me, matter-of-factly, that he could probably rebuild a system that has been developed over two decades - production-ready and deployed - in around six months, working largely alone with AI coding tools. The second was someone who had concluded that money and jobs were effectively finished.
The first of those feels credible to me. The challenge isn't the technology - it's being able to clearly define the outcome you want and validate what the AI produces. But if you can do that, the implications are significant. A lot of people's working value is built on tribal knowledge - an understanding of how a particular organisation works, its processes, its quirks. Once AI can navigate that as well as the people inside it, that value diminishes. The work doesn't disappear, but the number of people needed to do it does. The second conversation took that further - perhaps tongue-in-cheek, but not entirely. If the work most of us do becomes redundant at scale, what happens to the economic model that underpins how value is exchanged? He wasn't sure we'd need money anymore. I'm not sure I agree, but I'm not sure I can entirely dismiss it either.
There's something more fundamental worth considering. The purpose of the societies we set up, businesses, governance, interactions is ultimately human. These things exist to fulfil a human need or deliver value to a human. That means there is always a human in the loop - the demand side, at least, is inherently human. That feels obvious when you say it out loud, but it's easy to lose sight of when the conversation moves quickly to automation and replacement.
Which brings us to the supply side - who or what delivers that value. This is where the disruption is real and the uncertainty is genuine. As I described earlier, tribal knowledge and process familiarity are genuinely at risk. But the things that are harder to replicate - trust, judgment, empathy, accountability - retain their value precisely because the recipient is human. People aren't yet comfortable handing over responsibility for outcomes that affect them to something they don't fully understand or trust. That may change, but I don't think we are collectively ready for the human element to disappear.
I've had these conversations and done the thinking, and I still don't fully know where this lands. I suspect most people are in the same position - not through lack of interest or intelligence, but because the picture genuinely isn't clear yet. Even Sam Altman, the person arguably most responsible for where AI is headed, acknowledged this week that the fear and anxiety are justified and that the economic transition will be difficult. If he doesn't have all the answers, the rest of us can probably forgive ourselves for the same. I'm trying to position thoughtfully for whatever comes next - the fractional career path I'm pursuing is part of that - but I'm not certain I'm doing it as well as I could. What I keep coming back to is staying close to what people actually need. That feels like the most reliable compass when everything else is uncertain.