The 55% Regret Club: How AI-First Companies Are Learning Lessons the Hard Way

17 tickbyte 9 5/31/2025, 11:26:09 PM groktop.us ↗

Comments (9)

Havoc · 9h ago
You don’t really want to replace whole people’s jobs.

You automate 10% of the work across the workforce of say 100 and then you can reduce headcount by a couple

The CEO proudly announcing we just fired all of our category X people are doing it for publicity and become part of a statistic like this

vouaobrasil · 8h ago
> AI-first strategies aren't just failing individual companies—they're creating systematic corporate regret on an unprecedented scale. 55% of companies that replaced humans with AI now admit they made wrong decisions about those layoffs.

The real problems is that companies will just learn not to be so blatant about it. The replacement will happen more strategically, over a longer term, with promises about AI-augmentation rather than replacement. But over a longer period of time, jobs openings will go down nonetheless, and over decades, we will see the same thing until it will be still quite hard to get a new job. And I don't think UBI is a good answer either, because too many people are not self-directed enough to be happy without expending effort towards meaningful goals of obtaining necessities.

jdlshore · 10h ago
Just marketing spam. The message is interesting, but there’s so much shilling of their product and blatant bias , I wasn’t able to finish the article.
Spivak · 10h ago
But 45% of companies don't regret replacing their workers with AI? You can't tell me that those numbers aren't a massive success story for AI vendors.
rsynnott · 19m ago
Honestly, it’s kind of amazing that 55% of companies are admitting to it. “Hey, CEO, did you do something really stupid?” It is surprising that in 55% of cases the answer is “yup”.

And bear in mind that most corporate messing around with LLMs is only really in the last couple of years. Even where companies are being honest, figuring out whether something works takes time.

sputknick · 8h ago
That number will likely go up over time. Probably doesn't reach 100%, but more than 55%
tickbyte · 8h ago
55% that will even admit it.
nothercastle · 8h ago
They are probably lying because that’s bad for stocks
nemomarx · 10h ago
maybe a little petty, but it's funny how the top image in the article has the characteristic AI yellow. You'd really think people could pop a color balance filter on the output now?