The AI Great Leap Forward

In 1958, Mao ordered every village to produce steel. The steel was useless. The crops rotted. Today's top-down AI mandates are producing the same pattern: ba...

Han, Not Solo
@ai6yr Highly recommend anyone even tangentially adjacent to technology read this and understand what is being willingly cast aside
@robert Yeah, thanks for calling attention to it!!

@ai6yr Quoting from that:

So they adapt to build anti-distillation agent skills, just as the intellectuals adapted after the Anti-Rightist trap.

We are already seeing agent skills built specifically for job security. The performative skill looks comprehensive and demos well but omits the 20% of edge-case knowledge that makes it work in production — you are now more indispensable, not less. The poison pill encodes expertise faithfully but with subtle dependencies on context only you hold — internal wikis you maintain, terminology you coined, data pipelines you own — so removing you causes outputs to drift quietly until someone says “we need to bring them back on this.” The complexity moat makes the skill so architecturally entangled with your other work that extracting your knowledge is harder than keeping you around. You are now a load-bearing wall disguised as a decoration.

I wonder if this is something similar to the last AI boom, the one based on expert systems. I know that the technology had its limitations (but at least it was deterministic and actually included checkable logic) but one major reason that's often given for its failure was the sheer cost of encoding domain-specific knowledge. Maybe an unwritten part of the problem was that experts weren't particularly keen to make themselves redundant?

Anyway, mighty fine points raised there.

@ai6yr the "backyard furnace" simile wrt ai is the one i've too had in mind for a long time.

perfectly sensible people suggest 'using ai' for basic automation tasks that've possible since the 70's or supposedly technically capable ones pasting reams of babble amounting to 'maybe you have an uncaught error somewhere' when others are trying to find where that error is..

@geospaz @ai6yr A peer recently seriously suggested using an LLM to sort a short text list.

In other news, Tony Hoare hasn't been dead a month and this is where we are now.

@arclight
One of my coworkers used Claude to find all occurrences of a team name in a directory of JSON files...
@geospaz @ai6yr
@cambria @geospaz @ai6yr I bought a used copy of UNIX Power Tools for a coworker so he had a clear and simple (and large) reference to learn from. He's a safety engineering team lead, not a dev or ops person but he had hacked one of my test automation scripts to run production cases in his own. I was happy and proud of that. We need to own our tools.
@ai6yr Love it. This is what I always hoped my history students would take away from my class, the ability to apply it to the present.

@ai6yr Exactly. This.

I am living in the middle of this right now, and I've already killed three or four locusts. It's coming.

@ai6yr I'm so glad folks are noticing the connection between the Great Leap Forward and the AI thing.

Fudging the grain numbers looks like workers manually editing the AI output so it actually works

@ai6yr the great leap analogy nails it. Best piece i have read in a long time.

@ai6yr

"Four legs good, two legs bad." - Animal Farm

@ludicity if you haven't seen this yet... Oof.

@llorenzin I didn't know this:

"Oh, and Klarna? The company that loudly announced it would replace Salesforce with internal AI solutions? They quietly replaced Salesforce with another SaaS vendor instead. The backyard furnace couldn’t produce real steel. They bought it from a different mill."

But that is very, very funny