That’s the kind of argument you’d get from a B-movie mad scientist back in the last century: sure, there might be some adverse side-effects from their experiments, maybe the patients didn’t exactly consent — but how can science advance if you worry about ethics? Now Jeff Bezos has made that exact argument about AI (this clip comes from a twitter account by one Lisa Kippy):
There is a boatload of bullshit packed into that statement like the assumption LLMs will eventually become “a super-intelligence that could solve all our resource problems.” And that the growth of data centers, despite their demands for electricity and water, are a good, no a great thing, because science! We can’t worry about things like humans having adequate water for their needs — after all, it’s not like they’re Jeff Bezos or someone whose thirst matters.
This adds to the reasons for hating AI that Paul Krugman discusses here. For all their talk about the wonderful future ahead, what they tell us, repeatedly, is that AI is going to take all our jobs. Marc Andreesen literally claims it will take all the brain-power and creative jobs except, of course, venture capitalists like himself — no mere computer intelligence could do what he does (Kevin Leman discusses, more accurately, why LLMs can’t do what historians do)! Krugman remembers other techbros predicting 20 percent unemployment due to LLMs. Plus it will replace all creative work so authors, musicians and others can do … what? Somehow I don’t think they imagine the Jetsons kind of future where we can make a good living working an hour a day pushing buttons.
And as Krugman notes, we’re being pushed to use LLMs in various ways, whether we want to our not. My writer’s group had to put in more effort than we should to stop Zoom’s AI from recording us reading our stories. Nothing nefarious — it was an automatic, like taking minutes — but we still don’t fancy someone recording our work without our consent. Silicon Valley screams about how wonderful it is but it’s spreading more from pressure than the enthusiasm with which we adopted home computers or the Internet. Apparently even businesses are discovering the cost is not worth the output that results, as Krugman and 404 Media note.
Or consider California State University insisting they have to invest in AI to train their students for the “AI-driven future of work.” Which assumes the future will be driven by LLMs and that training to use them is the best use of the college’s money. Meanwhile, in the actual working world, the use of LLMs by both job-seekers and hirers is wreaking havoc on the job market.
Perhaps it’s unsurprising tech companies are presenting themselves as victims of some radical conspiracy. Kevin O’Leary, who’s developing a data center in Utah, had to walk back his accusations opponents of his center were proxies for China. Cops are monitoring criticism of data centers because who knows, maybe the critics wi’ll turn violent. One man who spoke out against data centers at a city council meeting was arrested for speaking too long. It would have been appropriate for cops to escort him out for speaking beyond the public-comment limits but arrest? That’s insane. Perhaps cops like AI because companies are so willing to share surveillance data without any Fourth Amendment concerns.
In other LLM notes:
Companies have figured out how to manipulate LLM searches for data.
LLMs are not helpful at answering small-molecule mass spectrometry questions.
How bad will Google going with AI summaries over actual links hurt online companies?
For a nonfiction writer, the only way to ensure LLMs aren’t giving you false data is not to use them.
If LLMs are sentient, so is the computer game Age of Empires.
“AI systems are beginning to replicate the same anti-LGBTQ bias and misinformation problems that have long plagued social platforms, according to a new GLAAD report previewed at Axios’ AI+NY Summit. The problems GLAAD flags — biased training data, privacy risks, automated discrimination, misinformation and the suppression of legitimate speech — extend beyond LGBTQ users to other minorities and groups in political disfavor.”
“Less than a day after President Donald Trump falsely suggested that Ilhan Omar had staged an attack on herself, the images started to circulate. In AI-generated fake photos that soon flooded both X and Facebook, the Minnesota representative is depicted posing next to the man who invaded a town hall meeting and sprayed apple cider vinegar on her from a syringe. In the AI-generated images, Omar and the man are both smiling; in some, the congresswoman is foisting a wad of cash, presumably to suggest that she bribed her attacker. “
I’ll close with some advice from David Dark: “Be the non-AI generated content you want to see in the world.”
