I periodically come back to Weizenbaum on computing as a conservative force. The example he uses is banking - without a technological invention, banks would have needed to reinvent themselves socially. But the PC made things fast enough for inefficiencies and contradictions to remain tolerable. - ThreadSky

spavel.bsky.social • 46 days ago

I periodically come back to Weizenbaum on computing as a conservative force. The example he uses is banking - without a technological invention, banks would have needed to reinvent themselves socially. But the PC made things fast enough for inefficiencies and contradictions to remain tolerable.

Comments

benniangerer.bsky.social•46 days ago

I always liked the way Oswald Wiener made this point in 1968: "... if Copernicus would have had a computer at his disposal with which to multiply and shift the by then conspicuous Ptolemaic circles, we would probably have a modern Ptolemaic worldview and would nonetheless be tackling spaceflight."

spavel.bsky.social•46 days ago

In the sense that AI is "revolutionary" it is a counter-revolution. Normally a funding crisis would force organizations to reevaluate what work is valuable to the org and what products are valuable to customers.

Then along comes AI and says, no that's not necessary, keep on keeping on.

spavel.bsky.social•46 days ago

The two fundamentals of generative AI seem unrelated but they are actually inextricable:
- generating Stuff at scale has never been cheaper
- you don't need a value proposition as long as AI is involved somehow

The result is that instead of removing the bullshit jobs, AI glorifies them.

spavel.bsky.social•46 days ago

Real jobs hesitate to rely on general purpose type AI chatbot because the outputs are inconsistent and usually low quality (specialized LLM tools that do one thing well are another story but typically they are not referred to as AI; in the public imagination, AI is when there is a chatbot).

spavel.bsky.social•46 days ago

But almost by definition, the quality of the output doesn't matter in the case of a bullshit job. So AI is a real "productivity booster" here and nowhere else (for the labor market overall, the impact of AI on productivity has been between 1-3%) because the rate of bullshit can be accelerated.

spavel.bsky.social•46 days ago

But all that excess supply of bullshit needs a purpose! And that's why AI is so hyped up - the hype is the outlet for the hyperscaled bullshit that high performing bullshitters are generating.

Using AI to bullshit about AI is seen as innovative (because it's AI, and AI use cases are hard to find).

spavel.bsky.social•46 days ago

It also has the side effect of making the bullshitter the hero of the story. They are no longer creating busy work - they are the Innovators and you will be Left Behind (in some kind of rapture situation? more on that elsewhere, I think).

And so we get AI for AI's sake, rather than for a purpose.

boleteful.bsky.social•46 days ago

I'm kind of reminded by some research that claimed people going into self help indistry are those who haven't been successful.

acuity.design•46 days ago

AI benefits managers and organisations who want to distance themselves from people and responsibilities. The AI BS barrier layer

spavel.bsky.social•46 days ago

I think I've been following you for nearly a decade, so I can honestly say that this is truly some of your best work. It's up there with the "act like a royal" tactic.

g--0.bsky.social•46 days ago

the only problem with this theory is that basically zero tasks are being automated

Comments

Posting Rules

Reply