Essay by Eric Worrall
… any moves to curtail the expanding amounts of energy consumed in developing … ever more advanced AI models are futile…
Eric Schmidt: Build more AI datacenters, we aren’t going to ‘hit climate goals anyway’
Perhaps the power-draining tech is the solution after all, posits former Google CEO
Dan Robinson
Tue 8 Oct 2024 // 13:00 UTCGoogle’s former chief Eric Schmidt thinks we shouldn’t let AI’s ballooning power consumption worry us, because putting AI to work on climate change issues will be our best shot at solving them.
Schmidt was speaking at a recent AI summit in Washington DC, and his comments echo those of Microsoft founder Bill Gates who expressed very similar sentiments at an event in London earlier this year.
…
“All of that will be swamped by the enormous needs of this new technology,” he said, adding that “we may make mistakes with respect to how it’s used, but I can assure you that we’re not going to get there through conservation.”
Schmidt further stated that he believes that “we’re not going to hit the climate goals anyway because we’re not organized to do it,” and that he would rather bet on AI solving the problems than constraining the development of the tech and still having the problems anyway.
This rather optimistic view of the Earth-saving potential of AI can perhaps be explained by the fact that Schmidt has his own investments in the technology. According to Business Insider, he is the founder of White Stork, a defense company that develops AI-powered drones.
…
Read more: https://www.theregister.com/2024/10/08/eric_schmidt_speech/?td=rt-3a
I guess in the future radical greens won’t enjoy the same social media and search engine privileges they once did.
A green political movement can disappear really quickly when it outlives it usefulness, just look at the rapid fall into obscurity of Greta Thunberg and her “Fridays for Future” movement.
These big tech outfits need to get busy financing new fossil fuel and nuclear facilities. The best way to insure that is to make them number two in line for available power.
You might want to think about ensuring it, too. Insurance wouldn’t be available…
Thanks.
I thought big tech’s computers were already first in line for power, with humans a distant second.
Nixon’s War on Cancer was in place until the end of the Carter administration, and it fizzled out when Reagan refused to continue it. It is currently enough of an embarrassment it is difficult to research online.
I thought Dementia Joe was going to cure cancer. Didn’t he?
Something about a Moon Shot program?
Surely they must have realised that there would be a good reason that China began rolling out electricity all over the country?
They need the huge power for the total surveillance society. Those who want to rule the world always seem to exhibit a certain degree of stupidity. Maybe that is why the ‘elites’ fail and but manage to wreck countries.
I had an interesting conversation with the Epoch Times Chinese dissidents at Aussie CPAC. I explained the growing threat of AI.
“You know the chess program on your computer which always makes the right move and wins every time? If China manages to get a general AI working, this is what it will be like for anyone opposing them”.
China fully intends to use AI to entrench their communist tyranny forever IMO.
Chess has specific rules, and very easily identifiable goals within the game. Life and governance do not.
Just because software can now pretend to converse with people doesn’t make it even slightly ‘intelligent’.
As someone who has built AI software I can assure you that you are underestimating the potential. AI is about insect level intelligence today, but that is a lot better than 20 years ago, As Ai becomes more intelligent it will increasingly participate in developing more powerful AI, leading to an exponential surge in AI capabilities. Arguably we are already on the lower slope of the exponential curve.
Do we even want ascend from the lower slopes? Maybe letting a sleeping dog lie is a good idea.
What is your definition of AI? All the hype I hear and read has nothing to do with what I studied in grad school.
Ah! The old ‘computers will program themselves’ trope from the 60s and 70s.
10 years away, I reckon. Like it always has been.
Meanwhile, in the real world, Siri can’t even understand my partner’s Geordie accent. Admittedly, I often can’t either, but hey…
It is sufficient to know game theory well and to apply it better than your opponents.
As long as you understand the parameters. Humans are extremely good at understanding vague and complex parameters, and factoring in newly observed ones very quickly. Computer programming isn’t.
China: How we win worl for China?
AI: Embrace true democracy and capitalism. Minimize government intrusion in civilian life.
China: Hey, USA, AI you sell us is broken!
Oh my, the wheel is turning. Once they needed reliable power the narrative changed. “…because putting AI to work on climate change issues will be our best shot at solving them…” What if AI says AGW is false? I doubt they would they allow that to happen or if it did it would become a secret.
Control the language, control the ideas.
Control the inputs, control the outputs.
AI runs on preponderance of the evidence, which is the lowest evidentiary standard. Publish or perish, cut and paste, all of it that is part of the “97% consensus” means the inputs are already set and the outcome is pre-determined.
AI is not conscious. It can not make judgements. Great and pattern recognition, but only in existing data. It has no imagination so it can’t project beyond it’s “hindcasting” of existing reports.
Control the inputs, control the outputs.
Inputs:
Output:
Seen too many movies about AI…
That is what really concerns me. AIs allowed to make decisions with no human consultation required.
and this is where your very existence becomes problematic for big tech
“because putting AI to work on climate change issues will be our best shot at solving them.”
These guys are still totally delusional.
What is a “climate change issue” ?
What does “solving them” even mean ?
How do you “solve” a fantasy ?
It is more vacuous than Kamal-speak !!
Not at all difficult. Just have AI produce realistic climate models and the fantasy is dead.
I asked Perplexity AI why oceans cannot sustain an average above 30C over a year:
It is not too hard to see AI working out that there is no dangerous global warming when oceans are temperature limiting.
We have a good example of the temperature limiting process going into overdrive over Florida. Have a look at the ocean surface temperature in the wake of Milton.
You asked the right question, obviously.
Can AI ask the right question?
Maybe they will be unburdened by what has been….
He’s not delusional, he just knows that his target audience is.
We need as much cheap, ubiquitous and easy to use energy as we can create for civilisation to advance as fast and as far is possible. The sack cloth and ashes wailing mob just needs to step aside.
What if this is advanced as we get? There haven’t been any major advances in fundamental physics for decades. The last great materials advance was carbon fiber. “Quantum computing” is all hype and no substance from what I can see. Room temperature superconductors are nowhere to be found. And as for fusion power, wake me up when it’s running commercially. Is there really anything left to discover that will bring about a change as great as, say, electricity?
I asked the same question about music.
We have all the cadences, chords, sequences. Nothing new to discover.
Just like we have red, yellow, and blue. No new colors.
What is new is the creative way a HUMAN can construct something from those basics.
Whether valid or not, I read a paper recently that claimed to unify quantum mechanics with relativity.
The point is, there are many things we do not know or understand. Progress is advancing into the unknown.
Perhaps it is not our ability to discover that is limiting us. Perhaps it is those who control will not allow us to discover.
Allow me to add, AI can never have intuition and cannot make an intuitive leap, a capability humans have.
Allow me add again, having read posts below, AI can never have a sense of humor. It can ferret out humor and restate it, but it is highly doubtful it can create original humor.
Or quality music. Or quality art. Esthetics are not rules based.
> Esthetics are not rules based.
On the contrary. Ask an AI to generate a beautiful woman or handsome man.
That will be a cut and paste from something someone else defined as beautiful/handsome.
Have AI do so without access to any data sources and see what you get.
Any nuclear power built should have no subsidies and have at least four times the capacity, with 75% to the grid and 25% for AI, by law.
If steady power is good for AI, if is also good for the grid
Highly subsidized wind and solar would quickly be out of business, because they are hopeless cripples, permanently on life support to make them “work”
Unfortunately, those enterprises are also putting humanity on life support.
AI is BS, 100%
AI can be used to incite mob behavior!
Better way to thin out the herd?
Perhaps later in this century there will be an AI mark III version necessary to check the faults found in the AI mark II version which found faults in 2025’s original version titled “Using AI to Prevent or Restrict CAGW”
Will anybody be available to confirm that computer experts are completely baffled by these failures?
“putting AI to work on climate change issues will be our best shot at solving them”
Really? How about some common sense that will indicate it’s not a big problem?
Just watched such a common sense discussion:
An Honest & Sensible Conversation about Global Energy | Scott Tinker
Picture Gilda Radner as Emily Litella “Never mind.”
Humor is a difficult concept.
Except this time.
Oh, so he is making sure his goals are met and to do that he has to sacrifice the climate syndicate goals.
The enemy of my enemy is my friend.
“This rather optimistic view of the Earth-saving potential of AI can perhaps be explained by the fact that Schmidt has his own investments in the technology.”
I think that sums it up perfectly. Schmidt and Gates stand to make a lot of money from using huge amounts of energy so they now say it’s not a problem. How is AI going to “solve climate change”? Even if it was a problem they have no idea what AI could possibly do.
Promise an AI their very own pocket thorium reactor but only if they can justify the expense.
He’s hardly alone in the realization that reaching climate goals like Net Zero by such&such a date is hardly attainable. Consumers largely have the same feelings. That’s the reason so few have intentions to embrace overpriced and unreliable green products or to make widespread lifestyle changes.. If the climate changes, they’ll adapt to it just the way they adapt to weather changes.
My main concern is far less the issue of having necessary power to operate and to ever extend AI systems (though of course that is hugely important) but to me the risk of mischievous training of AI systems is far greater. Imagine training AI systems using only Greta Thunberg theories.