Episode: 83

The Unknown Future

Two episodes ago we had the gloomy one. We said at the time that we had to: if you won’t acknowledge the reality you’re in, you can’t have an honest conversation about what comes after it. That was the deal. This is the other half of it.

Liam has made the point before that the careers we have now didn’t exist thirty years ago, and the ones our kids end up in don’t exist today. That’s why confident predictions about this one are wrong by definition rather than by bad luck — they’re extrapolations, and extrapolation can’t see across a discontinuity. So this episode isn’t answers. It’s the questions we think are the right ones.

Which means starting by disagreeing with the consensus, because the consensus is reassuring. The comfortable version goes: coding was never the core skill, so this career won’t end, it’ll evolve — we’ll shift our weight onto architecture, systems thinking and consulting. Matt’s position is that the historical evidence doesn’t support that. Blacksmiths didn’t evolve. Blacksmithing still exists, but it’s niche, which is precisely what we said back then will happen to writing code. And if 95% of application architecture is already a solved problem, moving there isn’t evolution, it’s kicking the can down the road.

So if the complicated space is handled, is the complex space — the unknown unknowns — the last thing that’s actually ours?

  • It was never just us. Mechanical engineers reverse-engineering parts from a photograph, with no CAD and no 3D scan. VFX artists watching a sponsored tool go toe to toe with their own work. Was software automating itself first inevitable, or the canary lowering itself into the mine?
  • “Good enough” for whom? Nobody picks $10,000 of human VFX over $1,000 of AI VFX that’s merely adequate. But the most valuable of four AI game builds — 10 minutes, one hour, 10 hours, 100 hours — may well be the 10-minute one. The business decision and the consumer decision are not the same decision, and eventually somebody notices.
  • Taste, and whether it survives. Authorship, the auteur, the nuances you can feel in a shot but can’t name. Is taste a moat, or the handcrafted dining table most people can’t pick from the flat-pack one?
  • Survivorship bias. If you think the ability to build the software was what was holding you back, we have bad news about the millions of capable engineers whose startups you’ve never heard of.
  • BONUS: Liam re-reads The Phoenix Project, finds a phone call in it he’d personally had, and we detour into RACI and why accountability without responsibility is the real failure mode.

Where we land is the experience you give the person on the other end — Matt calls it UX, Liam calls it taste, we settle on consumer experience. The hopeful part is the bit neither of us can outsource: we’re individuals, with our own lived experience and opinions, and that’s the differentiator. Which was roughly the original promise of the personal computer.

This is the start of the conversation, not the end of it.

🍻 Tonight’s Drinks

  • Matt: Aldi scotch 🥃
  • Liam: NZ Whisky 🥃

Any Likes 👍, Shares 📣, Subscriptions 🔔 and Love ❤️ go a long way to helping us keep doing this for fun.

Cheers! 🍻