How a socialist senator, an Effective Altruist organizer, a former White House strategist, and the Chinese Communist Party ended up demanding the same thing.
Not so strange. I have never trusted Bannon and the rest are peas in a pod. We do need to address local concerns about the data centers, especially power. AI is coming whether we do it or not. Really a separate issue from AI. More akin to big industrial projects. Can't leave it to local officials since they can be bought and the price is low.
The “danger” of “AI” is every bit as hyperbolic and exaggerated as the benefits. Yes, it’s going to be a useful tool, and will change things. But the danger comes from foolish misapplication rather than anything intrinsic to “AI” itself.
“AI” is “dot com” revisited. Remember all the hyperventilating books and articles from the late 1990s (if you’re old enough — I am) wildly claiming how “the old rules of economics no longer apply”? How profit didn’t matter, only “eyeballs” did? How did that work out?
The internet changed things, for sure. But pretty much every mainstream prediction about *how* it was going to change things was wildly off base: “Your refrigerator will automatically contact the grocery store when you’re low on milk” was one that was taken seriously despite it’s obvious ludicrousness. Nobody foresaw “social media”, or if they did it never was mainstream thinking.
And “AI” will change things: I use it extensively for coding (and other site reliability tasks such as analyzing failures) at work. It generates journeyman-level code and useful insights if used correctly, but still requires experienced oversight to maintain the quality of results. I wouldn’t want to be an entry-level (or subcontinental) coder because “AI” can churn out good-enough code much faster and cheaper than unaided humans can.
Where I see the risk is that poor AI generated code will start to be training fodder for future models, and, intuitively, this sounds like it has the potential for a death spiral that might accelerate quickly because of the volume of AI code being generated.
The other risk is from some of the financial shenanigans around AI and data centers. There will be a sorting when the economics reach equilibrium, and with that will come abandoned data centers and many billions in destroyed value. Let’s hope for a big nuclear power build-out before that happens so that, after the collapse, power prices to consumers and goods-producing industries fall as a consequence of the loss of data center demand.
Get these geezers out of government. Enough!
Not so strange. I have never trusted Bannon and the rest are peas in a pod. We do need to address local concerns about the data centers, especially power. AI is coming whether we do it or not. Really a separate issue from AI. More akin to big industrial projects. Can't leave it to local officials since they can be bought and the price is low.
They can all be bought. Track AIPAC proves that.
Most of the ordinary citizens I know here in Florida are opposed to the data centers being built and proposed.
Commies doin' Commie Stuff....
Took the words out of my mouth...
NOT STRANGE AT ALL. THIS IS 5G WARFARE.
The “danger” of “AI” is every bit as hyperbolic and exaggerated as the benefits. Yes, it’s going to be a useful tool, and will change things. But the danger comes from foolish misapplication rather than anything intrinsic to “AI” itself.
“AI” is “dot com” revisited. Remember all the hyperventilating books and articles from the late 1990s (if you’re old enough — I am) wildly claiming how “the old rules of economics no longer apply”? How profit didn’t matter, only “eyeballs” did? How did that work out?
The internet changed things, for sure. But pretty much every mainstream prediction about *how* it was going to change things was wildly off base: “Your refrigerator will automatically contact the grocery store when you’re low on milk” was one that was taken seriously despite it’s obvious ludicrousness. Nobody foresaw “social media”, or if they did it never was mainstream thinking.
And “AI” will change things: I use it extensively for coding (and other site reliability tasks such as analyzing failures) at work. It generates journeyman-level code and useful insights if used correctly, but still requires experienced oversight to maintain the quality of results. I wouldn’t want to be an entry-level (or subcontinental) coder because “AI” can churn out good-enough code much faster and cheaper than unaided humans can.
Where I see the risk is that poor AI generated code will start to be training fodder for future models, and, intuitively, this sounds like it has the potential for a death spiral that might accelerate quickly because of the volume of AI code being generated.
The other risk is from some of the financial shenanigans around AI and data centers. There will be a sorting when the economics reach equilibrium, and with that will come abandoned data centers and many billions in destroyed value. Let’s hope for a big nuclear power build-out before that happens so that, after the collapse, power prices to consumers and goods-producing industries fall as a consequence of the loss of data center demand.