"How AI could ensure the smart get smarter while everyone else gets left behind"

[Innula Zenovka]

The Great Cognitive Divide (Evernote Link because paywall)

An interesting and thoughtful analysis of how AI both helps develop some skills and abilities while causing others to atrophy.

Artificial intelligence can provide a function that’s a bit like humanity’s cognitive gym. Rather than having equal, across-the-board effects, AI can act like an amplifier, further driving a decisive wedge between two groups of people: those who have an abundant supply of critical thinking skills and a willingness to learn—and those who do not.

Artificial intelligence has already created a higher floor to what humans produce: mediocre outputs are now absurdly easy to create. Anyone can, with a few keystrokes, “write” a moderately compelling, cringe-laced LinkedIn post or a decent technical report. PowerPoints and spreadsheets are child’s play.

But artificial intelligence has also raised the ceiling of possibility for those who already have critical thinking skills, a willingness to learn, and are knowledge workers who can amplify their existing skills with powerful digital companion tools.

Certainly when I'm using ChatGPT to help me code, I find not only that it's a great time-saver when it's writing boilerplate stuff for me (though often not in the way I'd naturally do it) but also that it frequently comes up with techniques and solutions I'd not previously considered. It would be very easy just to use ChatGPT's solutions are they are, but I'm learning a lot by stopping to ask it explain to me how these particular methods work and why they're better than the way I'd naturally do things, which I can do because I'm saving so much time on rest of the project.

But I guess you need the basic knowledge to start with, as well as the inclination to learn. Though this, of course, is true of using any tool. He suggests

Imagine that a person comes up to you at the gym with some helpful advice while you’re lifting weights or running on the treadmill.

“A forklift could lift the same amount with way less effort,” they tell you. “And you’re not optimizing when you’re on that treadmill; you could cover five kilometers a lot faster on a bike, or better yet, in a car.”

which is all very well, but if you need regularly to shift heavy loads around, as opposed to enhance your physical fitness, you're probably going to want to learn to use a forklift rather than spend time at a gym.

 

[Beebo Brink]

The biggest problem for me with AI is that no matter how real and how far-reaching its utility may be, none of it is worth -- in monetary terms -- the use of resources it takes to run the damn thing. Data centers are coming at the most critical period in our human history, a time when we should be lowering energy and water resources, and it is jacking usage up instead.

All the debate and worries about AI's long-term effect on human productivity and whether or not it's eroding human skills? We'll be damn lucky in the short-term if we still have a functioning civilization by mid-century.

As water resources dry up and the cost of energy goes up (amidst rolling blackouts) and weather events upend distribution of goods, data centers will become a lightning rod for the rage of an impoverished citizenry. We'll see how long they last. (Well, someone will see. The mob chaos scenes of the future will probably happen after my time. Unless FTE (faster than expected) strikes again.)

 

[GoblinCampFollower]

The Great Cognitive Divide (Evernote Link because paywall)

An interesting and thoughtful analysis of how AI both helps develop some skills and abilities while causing others to atrophy.





Certainly when I'm using ChatGPT to help me code, I find not only that it's a great time-saver when it's writing boilerplate stuff for me (though often not in the way I'd naturally do it) but also that it frequently comes up with techniques and solutions I'd not previously considered. It would be very easy just to use ChatGPT's solutions are they are, but I'm learning a lot by stopping to ask it explain to me how these particular methods work and why they're better than the way I'd naturally do things, which I can do because I'm saving so much time on rest of the project.

But I guess you need the basic knowledge to start with, as well as the inclination to learn. Though this, of course, is true of using any tool. He suggests



which is all very well, but if you need regularly to shift heavy loads around, as opposed to enhance your physical fitness, you're probably going to want to learn to use a forklift rather than spend time at a gym.

Great article. Thanks for sharing! And yeah... I've very begrudgingly come to accept that AI can be used to enhance productivity and can be very useful when used right. I also think it's important to try to do things the old fashioned way first then use AI to improve output or to help when stuck in a particular point. Almost everyone I know uses it at least a little bit.

...but I feel very lucky to have finished my education without it. I think it's just necessary to make sure young people often have to do certain tasks without it and some things without electronic help at all! AI just doing all the work is not comparable to a lot of previous crutches that people worried about.

Beebo is also of course also very much correct that I really don't think ANY of this is worth what we are doing to energy prices and the climate. A lot of tech bros either don't believe in climate change or are just hoping the "super intelligence" emerges and magically solves the problem. I just don't think that's realistic at all. Even when you acknowledge the potential of AI, it's likely not quite worth it and should be shut down or scaled WAY back for the foreseeable future.

 

[Noodles]

The biggest problem for me with AI is that no matter how real and how far-reaching its utility may be, none of it is worth -- in monetary terms -- the use of resources it takes to run the damn thing. Data centers are coming at the most critical period in our human history, a time when we should be lowering energy and water resources, and it is jacking usage up instead.

All the debate and worries about AI's long-term effect on human productivity and whether or not it's eroding human skills? We'll be damn lucky in the short-term if we still have a functioning civilization by mid-century.

It also feels like a compounding problem. As we dry up.and cook farmland that has been around for centuries probably in many cases, is it even ever recoverable? Or is it just gone? Food shortages and mass immigration out of the hottest regions are just going to become a worse and worse issue.

 

[Beebo Brink]

As we dry up.and cook farmland that has been around for centuries probably in many cases, is it even ever recoverable?

The best American farmland is built on top of THOUSANDS of years worth of bison and antelope manure. That rich soil is a natural resource that we've plundered at a furious rate and without any way to replenish it. The massive herds of bison are gone and not coming back in our lifetimes. Just as bad, we're drawing down the deep water in aquifers, too. They are technically replenishable, but not at the rate we're drawing water from them.

 

[Noodles]

The best American farmland is built on top of THOUSANDS of years worth of bison and antelope manure. That rich soil is a natural resource that we've plundered at a furious rate and without any way to replenish it. The massive herds of bison are gone and not coming back in our lifetimes. Just as bad, we're drawing down the deep water in aquifers, too. They are technically replenishable, but not at the rate we're drawing water from them.

Yeah but not just in the US. There are a lot of things grown in Equitorial regions that can't really be grown elsewhere.

What do we do when we can't grow Coffee anymore.

 

[Innula Zenovka]

The best American farmland is built on top of THOUSANDS of years worth of bison and antelope manure. That rich soil is a natural resource that we've plundered at a furious rate and without any way to replenish it. The massive herds of bison are gone and not coming back in our lifetimes. Just as bad, we're drawing down the deep water in aquifers, too. They are technically replenishable, but not at the rate we're drawing water from them.

This might offer a way forward:

World’s first wind-powered underwater datacentre starts operating in China

 

[Khamon]

The biggest problem for me with AI is that no matter how real and how far-reaching its utility may be, none of it is worth -- in monetary terms -- the use of resources it takes to run the damn thing. Data centers are coming at the most critical period in our human history, a time when we should be lowering energy and water resources, and it is jacking usage up instead.

In context of the OP there are, broadly, two types of use. One is general stupid mediocre AI that requires large data centers compiling and tokenizing the sum of human "facts" so that it can all be related in stupid mediocre memes. The other is expert systems, that are focused on a specific industry or skill set, and utilize vetted (often already existing) data sources to produce productive and useable results. We can advance AI without requiring that the engines have access to massive data centers. Instead, we're building data centers so that Billionaires can own the data, and become Trillionaires.

 

[Caete]

This might offer a way forward:

World’s first wind-powered underwater datacentre starts operating in China

Prof Rick Stafford, a marine biologist at Bournemouth University, said: “An underwater datacentre is likely a good idea. While the cooling using seawater will result in some localised elevated temperatures, these will not be far reaching.”

Maybe just one center wouldn't be a major problem but multiply it by 100, or 1000 and it could be catastrophic. Not to mention sea water is highly corrosive and will steadily weaken the structure over time. How long? Well, given the tofu construction of China where new buildings are falling apart, made with substandard or entirely wrong materials, I'd wager less than 5 years before a leak develops, shorts out the whole center and seriously impacts the "dependacy" on AI. Let us not forget that anything underwater quickly develops into an eco system. How expensive is it going to be to keep the submerged structure clean? Will the marine growth impact the cooling? What happens when the structure gets snagged with fishing nets? A hurricane/typhoon hits it?

Back in 2018 Microsoft launched a pilot in the waters around Orkney in Scotland. Two years later, the company stated they had promising results but progress has since stalled. Why? Why Scotland and not the USA? Why did it stall? Probably tax incentives and that it had a low profit margin or even a negative one.

https://www.tomshardware.com/tech-industry/artificial-intelligence/more-than-75-data-center-build-outs-worth-usd130-billion-have-been-successfully-blocked-in-the-first-four-months-of-2026-bipartisan-opposition-mounts-nationwide-over-fears-of-soaring-power-and-water-costs

 

[Innula Zenovka]

Maybe just one center wouldn't be a major problem but multiply it by 100, or 1000 and it could be catastrophic. Not to mention sea water is highly corrosive and will steadily weaken the structure over time. How long? Well, given the tofu construction of China where new buildings are falling apart, made with substandard or entirely wrong materials, I'd wager less than 5 years before a leak develops, shorts out the whole center and seriously impacts the "dependacy" on AI. Let us not forget that anything underwater quickly develops into an eco system. How expensive is it going to be to keep the submerged structure clean? Will the marine growth impact the cooling? What happens when the structure gets snagged with fishing nets? A hurricane/typhoon hits it?

Back in 2018 Microsoft launched a pilot in the waters around Orkney in Scotland. Two years later, the company stated they had promising results but progress has since stalled. Why? Why Scotland and not the USA? Why did it stall? Probably tax incentives and that it had a low profit margin or even a negative one.

https://www.tomshardware.com/tech-industry/artificial-intelligence/more-than-75-data-center-build-outs-worth-usd130-billion-have-been-successfully-blocked-in-the-first-four-months-of-2026-bipartisan-opposition-mounts-nationwide-over-fears-of-soaring-power-and-water-costs

This explains the history of the pilot and why Microsoft abandoned it

Microsoft confirms Project Natick underwater data center is no more

But says it will use learnings from the project

www.datacenterdynamics.com

However, China clearly takes a different view.