Turns out that generative AI tools like ChatGPT will impact knowledge workers far more than cooks, mechanics, and others who might earn less. It’ll happen sooner, too.
This is according to a research paper from the University of Pennsylvania and OpenAI (OpenAI makes ChatGPT).
The scope of their predictions is stunning: Four-fifths of the US workforce could have a tenth of their work tasks impacted, and almost a fifth could see half or more of their responsibilities taken over by AI.
It’s brilliant propaganda for OpenAI, which obfuscates its profit-making intentions with a lot of gobbledygook language about purposes other than profit-making. Microsoft reportedly dumped $10 billion into it earlier this year, and I doubt it was a donation for the public good.
I’m intrigued that the research identifies knowledge workers for getting whacked harder and first, though.
The convenient bias about technology innovation has always been that it will replace work that people don’t want to do or, perhaps more sinisterly, work that none of the people involved in developing or promoting said tech would do for a living.
Automation, as an industrial concept, precedes the arrival of robots by at least a century as companies treated workers as cogs in a huge machine. So, the distance from such dehumanizing work to dehumanizing the workers isn’t all that great.
Smart machines simply replace the work done by dumb humans.
Robot is as robot does.
Only now, generative AI means that those knowledge workers are at risk. Financial advisers and brokers of all stripes. Journalists. Many of the tasks now done by lawyers.
What this means is that AI will put people who aggressively promote it out of work. You’d think that they’d think twice about evangelizing its merits now that those, er, benefits will cut much closer to home.
Or maybe there’s just too much money to be made before they get replaced by robots.
Those lower-paying jobs aren’t going to be safe for long, either.
ChatGPT doesn’t possess opposable thumbs, and building robots that are mobile, dexterous, and computationally flexible enough to do that work is mostly only a challenge of materials science.
The parents and grandparents of those machines have been bolted to the floors of factories for years. They’ll step up and out soon enough.
So, what’ll be left for we humans to do once AI does almost everything?
The AI Greek Chorus promises that there’ll always be new jobs serving machines, which is good news because the machines will always catch up and take them over. Think workplace improvement described by Pollyanna.
There’ll also be jobs that are too difficult or dangerous for expensive and complex robots to do. Yucky, horrible stuff.
The insurance coverage on human workers will be cheaper.