- cross-posted to:
- workreform
- cross-posted to:
- workreform
cross-posted from: https://lemm.ee/post/39685922
Earlier this year I got fired and replaced by a robot. And the managers who made the decision didn’t tell me – or anyone else affected by the change – that it was happening.
The gig I lost started as a happy and profitable relationship with Cosmos Magazine – Australia’s rough analog of New Scientist. I wrote occasional features and a column that appeared every three weeks in the online edition.
It didn’t. In February – just days after I’d submitted a column – I and all other freelancers for Cosmos received an email informing us that no more submissions would be accepted.
It’s a rare business that can profitably serve both science and the public, and Cosmos was no exception: I understand it was kept afloat with financial assistance. When that funding ended, Cosmos ran into trouble.
Accepting the economic realities of our time, I mourned the loss of a great outlet for my more scientific investigations, and moved on.
It turns out that wasn’t quite the entire story, though. Six months later, on August 8, a friend texted with news from the Australian Broadcasting Corporation. In summary (courtesy of the ABC):
Cosmos Magazine used a grant to build a ‘custom AI service’ to generate articles for its website.
The AI service relied on content from contributors who were not consulted about the project and, as freelancers, retained copyright over their work.
Contributors, former editors and a former CEO, including two co-founders, have criticized the publishing decision.
Cosmos had been caught out using generative AI to compose articles for its website – and using a grant from a nonprofit that runs Australia’s most prestigious journalism awards to do it. That’s why my work – writing articles for that website – had so suddenly vanished.
If your only experience is with text or photo generation in consumer models, that is a fun toy, but it’s very small compared to the total possibilities. And it gets better exponentially. It is already replacing software engineers and paralegals. And it’s better than radiologists at spotting cancers and such.
It’s the only thing people know. I haven’t seen it replacing engineers, though I’ve seen a lot fired because they thought they could.
What system replaced an engineer?
For that matter, a paralegal?
Name? Company? Is there anything that presently exists in any state of relative useful completion? No. And there isn’t going to be.
What we’re going to see as these gargantuan investments start to dry up is lots of horrific kludging of some-AI-some-non-American-behind-the-curtain magic that very much wishes us to think of it as advanced AI. And it’s not going to work very well either.
It doesn’t have to replace every engineer to have replaced engineers. It massively increases productivity, and therefore companies can “trim the fat” and get the same level of productivity from fewer employees. I haven’t written a function from scratch in months, and we lost 10 team members six months ago.
As long as you don’t break any new ground or require complex problem solving, great. Trimming the “fat” is a valid use case.
Ideally, the “fat” would be re-focused on things only humans can do, but of course business doesn’t work like that.
I honestly hope you are right. But as an expert, I am confident you are not.
There are existing successful products in these and other areas. I am not going to do your research for you, because I have burned a bunch of time and energy on that in these types of discussions in the past and rarely has anyone changed their mind. If you want, you can start your search with the term “AI Agents”
Realistically, the near future is going to look like 1 person managing 10+ AI Agents that do desk jockey jobs in a supervisor role. Long term future, as confidence rises, the managers will not be needed either.
Fair enough, I’ll look into it. But I’m expecting a lot of BS.
Which is what you’ll find if that’s what you’re looking for. There’s plenty of it out there. But those of us working in impacted industries have been witnessing the early impacts for a year or more already, and the systems get a dozen times better every few months.