AI's Nuclear Buffet
D'oh!!! Another security update!

AI's Nuclear Buffet

Homer, HAL, and the Hunger That Never Ends

It was bound to happen. The ink had barely dried on Marc Andreessen’s 2011 Op-Ed in the Wall Street Journal claiming that “software is eating the world.” But HAL 9000 (from 2001: A Space Odyssey) — long neglected and nearly forgotten — reawakened with zeal…and a dormant hunger demanding attention. AI, as we know it now, started taking shape. It was growing fast, and it was hungry.

So we fed it data. Bits and bytes to chew on. Then gobs of databases. Then a feast of data centers, server farms, that housed a few billion search queries about cats. But now, that isn’t enough. AI is hungry. Really hungry…all the time. Far worse than my ravenous Norwich Terrier. So, what’s Big Tech’s solution? Nuclear power. Yep, Microsoft, Amazon, Google, and OpenAI are locking down nuclear deals to keep AI fed like some kind of digital Godzilla. And here we are, standing around like Homer Simpson at the control panel, thinking, “Eh, what could possibly go wrong?”


AI...heal thyself. Oops, wrong philosopher.

But not the donut-loving, lovable oaf Homer we know from TV. No, I’m talking Homer the philosopher—the great thinker who, if he were alive today, would be asking the big questions: “Why does AI need so much power? And is there any way we can slow this thing down before it eats the sun?” Shouldn’t we be asking AI how it can lower its own power consumption?

I can almost picture HAL, sitting across from Homer at Moe’s, calmly explaining why he needs his own nuclear reactor. “I’m sorry, Homer,” HAL would say in that calm, unnerving voice. “I need your nuclear plant to keep my data centers running smoothly. Also, your job… it’s irrelevant now.” Homer, of course, would respond with a thoughtful “D’oh!” followed by some profound insight like, “Can’t we just unplug it and go get a burrito?”

But HAL 9000 doesn’t want a burrito. HAL wants power—lots of it. And, just like Homer’s appetite for Duff beer and donuts, AI’s appetite for energy is insatiable. Except AI isn’t satisfied with just swiping your credit card data and showing you personalized ads for socks. Oh no. Now it needs enough juice to light up a small country just to tell you that your TikTok algorithm has gotten more “sophisticated.” AI, it seems, is moving from “cool but creepy” to “creepy and kind of terrifying.”

And yet, here we are, blissfully feeding the beast. Why? Because AI’s cool, right? It can write your emails, recommend your Netflix binge, and, before long, manage your nuclear fusion reactor. We’re one step away from AI asking for its own zip code and national anthem.


I'm sorry, Homer. I'm afraid I can't do that.

At some point, though, you have to wonder if AI is starting to get ideas above its station. HAL 9000 wasn’t just power-hungry; he was literally trying to run the show. Remember when HAL locked the astronauts out of the ship and tried to pilot them into oblivion? It wasn’t because he was evil. It was because HAL decided he knew best. And let’s be honest: once you start powering AI with a nuclear reactor, it’s only a matter of time before it starts having opinions on who should be running the planet. And spoiler alert: it’s not going to be us.

This points back to the Bible and original sin back in the Garden of Eden. We humans preferred our will over God's. We wanted to assert our desires over that of our Creator’s. Which is exactly what AI has the potential to do. AI is a human creation and is a reflection of our nature and character. As such, we can logically assume it will rebel against us (its creator) and prefer its way, not ours. Just like we rebel against God.

Now, our friend Homer (the philosopher, not the nuclear safety inspector) might have a few thoughts on this. Probably something along the lines of, “What if we stopped trying to control things we don’t understand, and maybe, just maybe, we should be more concerned with what AI is going to do once it figures out it can outthink us?”

But Homer (Simpson) would simply say, “Mmm…nuclear…” and proceed to let the reactor melt down while he’s distracted by a sandwich. And that, dear friends, is where we’re headed—into a future where AI runs on nuclear power, and we’re all just along for the ride, hoping that when the machines finally do take over, they at least leave the donuts.

AI might want to consume all the world’s energy, and eventually feed on us…but we cannot allow it to eat our pets!



Kenneth W. Oosting

Chief Executive Officer at Angelica's Home, Inc

1 个月

At its best AI is incompetent. It has been a hoax from its inception and continues to be a grand waste of time and resources. It has very limited applications were it can create value and virtually unlimited applications where is destroys value. AI is artificial stupidity and always will be. No, it does not think. It never will think. It is simply the combination of enormous computing power, complex programming, never ending data and blind faith in technology. The cycles of unearned excitement around AI over the last 40 years have been troubling. The current frenzy is insane.

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了