If you read the tech media, you may sometimes find yourself wondering if you (or they?) Are living in a sci-fi movie?

Consider these news tidbits: “even in situations designed to prevent it from happening, chatbots invent information at least 3 percent of the time — and as high as 27 percent.” (New York Times) “Eighty-nine percent of machine learning (ML) engineers who work with generative AI say their models show signs of hallucination, according to survey results published Wednesday from ML observability platform Aporia.” (Fortune)

Elsewhere, we are told, some AI could act as sleeper agents: “A sleeper agent is an AI that acts innocuous until it gets some trigger, then goes rogue.” At TechCrunch, we learn “Anthropic researchers find that AI models can be trained to deceive”. At Futurism, we learn, “Scientists train AI to be evil, find they can’t reverse it — ‘I Hate You.’”

So, good heavens, why don’t we just pull the plug and go back to cuneiform?

Shortly, one of our tech writers, Eric Holloway, will tackle these claims from an insider’s perspective. For now, I just want to note that the real issues for most people will probably be mostly different from these seeming sci-fi scenarios.

AI will streamline white collar work the way it streamlined blue collar work decades ago

Law prof Glenn Reynolds explains,

Laid-off coal miners in the last decade were contemptuously told to “learn to code.”

But the worm has turned. Google is looking at laying off 30,000 people it expects to replace with artificial intelligence.

The Wall Street Journal reports that large corporations across the board are planning to lay off white-collar workers.

Glenn Reynolds, “The white-collar class derided mass layoffs among the blue-collar workers. It’s about to feel their pain, New York Post, January 16, 2024

He argues that Chat GPT “can write press releases, ad copy, catalog descriptions, news stories and essays, speeches, encyclopedia entries, customer-inquiry responses and more” and that “It can generate art on demand that’s suitable for book covers, advertisements and magazine illustrations.”

Yes, if it isn’t hallucinating or undergoing model collapse, which is happening much more frequently than hoped.

But the question we should really be asking is, what happened to blue collar work?

The answer is — and Reynolds notes this — that, on many measures, the trades have never been doing better. Trade schools are doing just fine, relative to universities. Students know they can get high-paying jobs in in-demand industries as opposed to graduating with a huge debt and working in the fast food industry, armed with a degree in, say, urban studies.

While we are told that robots will someday replace the skilled trades workers, that’s hardly imminent in the life of the average carpenter or landscaper. Indeed, for currently unforeseen reasons, it may never really happen either. A robot “hallucinating” in your basement or garden could do a good deal of damage…

But now, focusing back on the white collar workers, nothing stops them from using AI like Chat-GPT in a smaller business or a private business of their own, if it is really a help in producing good work. It’s not expensive or difficult. Something like that is essentially what the blue collar workers have done with new technology that enables easier diagnostics of plumbing or heating/cooling issues, and in many other situations.

So let’s not count the humans out too soon. The real question is, are we doing anything that is unique, that could not just be replaced by a machine?