[JohnFen]

> The threat of these models isn't that they'll go all Skynet and kill everyone, it's that they'll cause a lot of economic devastation to people who make a living

Yes. This is pretty much my only concern about these models, and I'm powerfully concerned about this. It's hard to see how this will lead to a good place. It seems more likely that this will lead to increased poverty and multiple socioeconomic crises.

I am even more concerned that very few people are talking about this, and none of the power players in this space are, except for occasional mentions in passing of fantasies like UBI.

[bloppe]

> I am even more concerned that very few people are talking about this.

People have been talking about the threat of automation since the very beginning of the industrial revolution. It just never plays out nearly that badly, and short-term disruptions are always outweighed by long-term efficiency gains within ~5 years or so; even those who experience the worst career disruption tend to end up better off within that time frame.

I certainly would not like for my career to be disrupted for ~5 years, but the alternative would be worse.

[JohnFen]

> even those who experience the worst career disruption tend to end up better off within that time frame.

That hasn't been my observation at all. In the US, there are large swaths of the nation that still haven't recovered from the last similar event.

To add additional worry, the last time, everyone was told that the way out was to "upskill" into knowledge and service industries. Which a lot of people did, and those people were fine. But what are people to do this time? "Upskilling" back to physical jobs can only absorb so many workers, particularly since there aren't as many such jobs as there used to be.

This is all why I'm so concerned. I don't think history gives us any real reason to be optimistic here. In the very long term -- a couple of generations, say -- perhaps. But in the meantime? Even ignoring the ethics of some people deciding that others are expendable, the people being kicked to the curb will still have to find a way to eat, keep a roof over their head, etc.

If even 10-20% of the population can't do that, we're in big trouble.

All that said, nothing would make me happier than for you to be right and me to be wrong.

[natechols]

> People have been talking about the threat of automation since the very beginning of the industrial revolution.

And the prediction that only the ownership class would benefit from technological improvements is at least as old as Marx.

[tstrimple]

And it's proven to be unequivocally true hasn't it? Just take a look at the accumulation of wealth since 1970.

[fl0ki]

I am very curious about what pre-1970 hardware and software you posted this from.

[Animats]

I've been asking for years, if we have all these computers, why do we need so many people in offices? Now we seem to have passed "peak office", with much help from the pandemic.

If everything you do for money goes in and out over a wire, be very afraid.

[waffletower]

From my vantage, short-term (3-5 year) fear seems unfounded. As a software engineer, I can clearly see what ChatGPT and its LLM ilk can and can't do that I can easily do myself. LLMs clearly accelerate my access to API documentation and provide excellent outline code. But hallucinations are omnipresent and can often necessitate additional iteration and rethinking of development approaches. I think the productivity boost due to LLM usage is smaller than many credit them for. The intensity of employment displacement fear comes from an illusion that LLMs have agency. AutoGPT is not much more than an experimental repo and there isn't a viable alternative yet. "WHEN YOU COMMAND AN LLM, YOU ARE THE AGENCY", LLMs are mere extensions. Don't sell yourself short, prompt crafting/engineering is where the agency lies and requires real knowledge and context to empower you effectively use them for successful software engineering.

[JohnFen]

> The intensity of employment displacement fear comes from an illusion that LLMs have agency.

I don't think so. My concerns have nothing to do with agency, anyway. Nor are my concerns limited to (or even primarily about) impact on software engineering specifically.

Even if LLMs perform worse, if using them will save companies money over employing people, then those people are gone.

[Animats]

LLMs have agency when given a goal and connected to an API that can do something. This is likely to be a problem now and then.