← sources

Senior devs are in their Kasparov era

source: reddit (reddit:ClaudeAI)published 2026-10-26 09:34ingested 2026-10-27 02:36

retained text

I keep coming back to how chess dealt with engines, because it feels like we’re replaying it almost step by step.

Quick recap. Deep Blue beats Kasparov in ‘97 and he does *not* take it well. For years the top players say the same thing: ok engines calculate, but humans *understand*. Then Kasparov himself comes up with “Advanced Chess”, where a human and an engine play as a team. His point was never really “humans beat engines”. It was “human + engine beats engine alone”.

And he was right! For about a decade. Then engines fixed their blind spots, and humans stepping in started making the games *worse*. Today if you ask Carlsen or Gukesh whether they could beat Stockfish they’ll just laugh at you. Engines are their prep, their sparring partner, their coach. Nobody in that generation wastes energy pretending otherwise, and they’re stronger for it.

Now listen to what we say about agentic coding: “sure the agent writes the code, but it needs an experienced engineer steering it.” That’s literally the centaur position. It’s probably true today. Chess says the real question isn’t *whether* it’s true, it’s for how long.

One detail i love: in a 2005 freestyle tournament (engines allowed), the winners weren’t grandmasters. They were two amateurs with three regular PCs. They just had a better process for steering the engines. So the best “centaur architect” might not be the person with 30 years of architecture experience. It might be whoever’s best at driving agents. Not the same skill.

The big difference is that chess has a scoreboard. You can play a thousand games and *see* when human overrides start hurting. In software nobody can really prove whether the senior’s override made the system better or worse. Which means we could stay convinced we’re adding value long after we stop.

But here’s what i think survives. Engine vs engine chess exists, and it’s useful, engines get better and humans steal ideas from it. But nobody really watches it. Chess matters because *people* play it. Software matters because people *need* it. A flawless system nobody needs is just engine vs engine chess: impressive, and pointless. Deciding what’s worth building, and for who, feels like the part that stays ours. For now at least lol.

So i’m curious:

* Do you think we’re in the centaur phase right now? How long does it last?
* Seniors: what do you bring when steering an agent that you’re confident *won’t* get automated?
* People who learned to code with agents: does “it’s better than me at writing code” feel like a threat, or just a fact?
* With no scoreboard, how would we even know when we stop adding value?

PS: Yes I used Claude to help me put this parallel into perspective, irony intended. Perhaps at some point agents will be better than humans at posting on Reddit. Oh wait..!

PPS: Thank you to those of you who engaged with the core idea, and those who pointed out the ways in which my analogy was flawed (or uninformed as it comes to chess). The argument is probably a bad one and I want to acknowledge that. 

document fetched 2026-10-27 00:23 · canonical https://reddit.com/r/ClaudeAI/comments/1xfs6h7/senior_devs_are_in_their_kasparov_era