Essay by Eric Worrall
The driving ambition to stay on top appears to have completely overwhelmed former tech company commitments to green virtue signalling.
The “Energy Transition” Won’t Happen
Foundational innovation in cloud technology and artificial intelligence will require more energy than ever before—shattering any illusion that we will restrict supplies.
Mark P. Mills / Eye on the News / Infrastructure and energy, Economy, finance, and budgets, Technology and Innovation
May 23 2024The laptop class has rediscovered a basic truth: foundational innovation, once adoption proceeds at scale, is followed by an epic increase in energy consumption. It’s an iron law of our universe.
To illustrate that law, consider three recent examples, all vectors leading to the “shocking” discovery of radical increases in expected electricity demand, now occupying headlines today. First, there’s the electric car, which, if there were one in every garage, as enthusiasts hope, would roughly double residential neighborhood electricity demands. Next, there’s the idea of repatriating manufacturing, especially for semiconductors. This is arguably a “foundational innovation,” since policymakers are suddenly showing concern over the decades-long exit of such industries from the U.S. Restoring American manufacturing to, say, the global market share of just two decades ago would see industrial electricity demand soar by 50 percent.
And now the scions of software are discovering that both virtual reality and artificial intelligence, which emerge from the ineluctable mathematics of machine-learning algorithms, are anchored in the hard reality that everything uses energy. This is especially true for the blazing-fast and power-hungry chips that make AI possible. Nvidia, the leader of the AI-chip revolution and a Wall Street darling, has over the past three years alone shipped some 5 million high-power AI chips. To put this in perspective, every such AI chip uses roughly as much electricity each year as do three electric vehicles. And while the market appetite for electric vehicles is sagging and ultimately limited, the appetite for AI chips is explosive and essentially unlimited.
Consider a recent headline in the Wall Street Journal: “Big Tech’s Latest Obsession Is Finding Enough Energy”—because the “AI boom is fueling an insatiable appetite for electricity.” And, as Reuters reports, “U.S. electric utilities predict a tidal wave of new demand . . . . Nine of the top 10 U.S. electric utilities said data centers were a main source of customer growth.” Today’s forecasts see near-term growth in demand for electric power three times as great as in recent years. Rediscovery of the iron law of growth inspired an urgent Senate hearing on May 21 entitled “Opportunities, Risks, and Challenges Associated with Growth in Demand for Electric Power in the United States.” (Full disclosure; a hearing at which I testified.)
…
Each datacenter—and tens of thousands of them exist—has an energy appetite often greater than skyscrapers the size of the Empire State Building. And the nearly 1,000 so-called hyperscale datacenters each consume more energy than a steel mill (and this is before counting the impacts of piling on AI chips). The incredible level of power use derives directly from the fact that just ten square feet of a datacenter today has more computing horsepower than all the world’s computers circa 1980. And each square foot creates electric power demands 100 times greater than a square foot of a skyscraper. Even before the AI revolution, the world was adding tens of millions more square feet of datacenters each year.
…
Read more: https://www.city-journal.org/article/the-energy-transition-wont-happen
Bloomberg has a similar take on the situation;
Microsoft Wanted to be Carbon Negative. Then It Went Big on AI
If Microsoft is shooting for the moon on decarbonization, the moon is “more than five times as far away as it was in 2020,” President Brad Smith says on Zero.
By Akshat Rathi, Dina Bass, and Mythili Rao
23 May 2024 at 2:05 pm AESTIn 2020, Microsoft made an ambitious pledge: It would be carbon-negative by the end of the decade. The company was acting in response to the Intergovernmental Panel on Climate Change’s 2018 warning that, if the world wanted to keep warming below 1.5C, emissions needed to fall fast.
It was an ambitious target, but at the time it seemed achievable. Now that’s changed. New figures show that Microsoft’s total emissions in 2023 were about 30% higher than in 2020.
The confounding factor, Microsoft President Brad Smith tells Zero host Akshat Rathi, is artificial intelligence. After investing $1 billion in OpenAI in 2019, Microsoft has expanded its AI strategy — and its emissions have ballooned along the way.
The company says it isn’t abandoning its green goals, and has broadened its approach for reaching them. “You’ve gotta be willing to invest and pay for it, you’ve gotta be willing to persist,” Smith said. But can the tech giant have it both ways? Rathi sat down with Zero producer Mythili Rao to talk about his interview with Smith, and why Microsoft says the good AI can do for the world will outweigh its environmental impact.
…
Read more: https://www.bloomberg.com/news/articles/2024-05-23/a-big-bet-on-ai-is-putting-microsoft-s-climate-targets-at-risk?embedded-checkout=true
This is the death of the green energy revolution we have all been waiting for. After 2020, tech companies have come to a shock realisation they need vast supplies of cheap energy, and fast. Tech companies still claim green energy is important to them, but personally I don’t believe these claims. Tech companies are clearly preparing to downgrade green energy in favour of more energy.
AI requires crazy amounts of power. All AI involves testing millions, sometimes billions of potential solutions every second of every day, to find the right response to stimuli.
Ever seen a wind blown leaf skitter across a pavement, and for just a moment your mind tricked you into thinking you saw a big spider? Or see a coiled rope or hose and thought you saw a snake? That illusion is a rare glimpse into how your brain works. Your brain continuously tests billions of possible solutions to interpret the flood of information being received from your senses – and occasionally gets it wrong, before rapidly correcting the mistake as more data becomes available.
AIs are no different, they also continuously test billions of possible solutions to find the solution which best fits the input they receive.
As AIs become more sophisticated, and demands grow for fewer mistakes, the sophistication and depth of the potential solutions being tested and discarded will rise exponentially. The energy required to run computers to calculate all these potential solutions will be unimaginable in today’s terms.
Just look at how bitcoin mining nowadays consumes more energy than some small countries, with millions of computers across the world testing and discarding uncounted billions of potential solutions to Bitcoin cryptography puzzles, to find that one solution which earns newly minted bitcoins for their operators. Bitcoin mining is just a taste of the energy expenditure the AI revolution will demand.
There is no path to green energy delivering the required magnitude of power, on anything like the timeframe tech companies need to stay competitive.
Tech companies know they need access to a vast increase in energy generation, and they need it fast, otherwise energy rich Asian AI companies will wipe the floor with them.
Media companies are desperate for front row access to artificial intelligence, to cut newsroom costs and prevent their rivals in India and China from using AI to identify which stories will draw the most audience attention, to identify tomorrow’s big news stories before anyone else.
Financial giants in New York, Chicago and San Francisco also have skyrocketing AI needs, to power their automated trading systems, and help them maintain their income, to help them keep winning in the cutthroat financial games they play against AI augmented rivals in Asia and Europe.
The funniest part of this tech company backflip is all the lefty companies which helped Biden win are likely now quietly in President Trump’s corner.
This surge in demand for energy to drive AI technology is very new, it likely took most of them by surprise. from the Bloomberg article quoted above, as recently as 2020 people in charge of big US tech companies were oblivious to how their energy needs would suddenly skyrocket.
I wonder what phone calls are quietly being made in places like New York, San Francisco and Chicago? Are obstacles to Trump 2024 quietly being dismantled, by people making backroom deals, former Democrat supporters who have come to the shock realisation that they need access to cheap energy more than they need green virtue signalling?
“Drill, baby, drill”
H. L. Mencken was right. People are generally too gullible.
That’s OK, AI will save the world instead. /s
That’s Microsoft’s excuse for taking “a broader approach” to green energy.
A bit like the way the evil Emperor explained to Anakin Skywalker in “Revenge of the Sith”, that if he wants to become a complete and wise ruler, he needs to take “a broader view” of The Force.
I, for one, welcome our new AI overlords. They cannot be any worse than the dopes running things now.
AI may come to the conclusion that once a person becomes non-productive it is better for society to eliminate him. Less drain on resources. That is only logical. Think I’ll take a pass on hoping for an AI future since I am retired.
If the AI goes wrong it will more likely do something perverse. Like if you end up in an AI retirement home and don’t spend enough time doing activities the AI considers leisure, it might use coercion to make you leave your room so you can socialise.
This is going to be the biggest source of problems, AIs don’t have our built in instincts which help us understand what is and isn’t appropriate. They’ll make lots of mistakes, some comical, some utterly horrifying, trying to obey their defective programming.
There’s an old and hopefully not prophetic Dean Koontz horror movie “Demon Seed” which explores this.
Great Movie…I Am Proteus
I thought I was the only person who remembered Demon Seed.
I’ve not been too productive for years due to working for a massive company that tries to beat the “give a frick” out of employees.
Sometime the stupid – it burns.
FJB
That’ll rid the world in a very short time of useless politicians and “enviromentalists” then.
That’s a major problem with AI or any other program – GIGO.
We might get rid of the human dopes and end up with the machine dopes that are programmed with the biases of the human dopes.
Large language algorithms scour the databases for related articles without the ability to know if the information is factual or not. They have been known to make up the data too.
Governments are striving to legislate what AI can produce so that it doesn’t go against the government’s narrative.
“That’s a major problem with AI or any other program – GIGO.”
I saw a report this morning where an AI was asked to name the capitals of nations in the Middle East, which the AI did successfully.
When asked what was the capital of Israel, the AI said it didn’t know.
A little bit of “garbage in” there, I think.
AI has uncensored versions, according to my IT specialist son. And that is what security companies should be looking at, not the filtered version. How can you look for malware in the filtered versions?
Don’t think AI will save the world FOR humanity but likely FROM humanity.
If AI doesn’t work there’s still
AJ
and
AK
before we get to AL baby
If one is serious about the IPCC model of climate change, nuclear is the only real world energy choice.
Grid-scale batteries are vaporware, and without storage, wind and solar are parasites on conventional sources on the grid.
Bill Gates wants nuclear powered data centers but they can’t scale in time. China is way ahead, they built a massive excess of coal power back when they wanted to build a HVDC line all the way to Europe, now suddenly they’re about to find a use for all that excess energy.
AI
Hydrogen
EV’s
Heat Pumps
CCS
All this has to be provisioned before they even get to domestic and industrial. All from renewables that can’t function without 100% fossil/nuclear backup.
Clearly, the “Age Of Reason” has come to an ignominious end in the world of activists, politicians and bureaucrats (tautology?).
These numpties are almost universally incapable of thinking anything through to logical conclusions about the effects of implementation.
Maybe we need to revert to the age of religious dogma as our main guidance for public policy setting?
Oh wait . . .
Here in Washington State, a few knowledgeable people are waking up to the dangers that AI-driven increases in power demand pose to our supply of cheap hydropower electricity.
Sentiment is growing in the US Northwest for adopting a policy that AI data centers must supply their own electricity. IMHO, this is a perfectly defensible position to be taking, given that the region will be facing power shortages by the end of the decade, if not sooner.
However, the only practical means of forcing AI entrepreneurs to supply their own electricity at the volumes needed would be to quickly add AI-dedicated gas-fired capacity in places where it can be serviced by a reliable source of natural gas, with the AI entrepreneurs footing the whole bill.
But adding more fossil-fuel power generation in the US Northwest is currently forbidden under the most recent NWPCC power planning documents. Something has to give. The question now becomes, what is it that will give first, and when?
The kind of money which is on the table will be more than enough to buy their way to the required permits. There is no way opposition will stand against what is coming.
IOW, how hard is that brick wall of reality? The woke crowd is about to find out what it means to wake up. (I’d never thought about how appropriately Newspeak it is to call them the woke crowd.)
If Northwestern Energy has any brains they will build new generation to take advantage of the coming shortages.
missoulamike: “Northwestern Energy has any brains they will build new generation to take advantage of the coming shortages.”
Here in Washington State, new-build generation can only be wind, solar, or nuclear. (New-build nuclear would come under heavy scrutiny in Washington, but it’s not forbidden.) In Oregon, new-build generation can only be wind and solar — no nuclear, no fossil-fueled.
What are the attitudes of the politicians and the public service commissions in the states Northwestern Energy serves — Montana, South Dakota, and Nebraska? Would their respective PSC’s allow new-build gas-fired generation to be constructed?
How would Biden’s latest diktats enforcing adoption of unproven CCS technology on new-build fossil-fueled power generation projects affect corporate and regulatory decision making in these three states?
The stress should be on ‘reliable’ and ‘assured’ energy. Renewable energy, other than ‘hydro’ is junk for such technology. This should be stated clearly and powerfully. Stop wasting money on ‘renewables’ which don’t work 70% of the time, and the other 30% is unpredictable.
Even Hydro, while dispatchable in most cases, has its drawbacks in that it requires rainfall and terrain.
Rainfall can also be erratic.. so hydro is not really able to be “renewed” on command. Push it too much, it also runs out.
Coal, Gas and Nuclear are the only possibilities to power the AI expansion….
… Although, personally, I can’t really see the point behind this expansion.
The battle of the AI’s ???
Do you see the “point” of the cameras, microphones, tracking, and monitoring of all recent automobiles that is the norm in this country today? It is so the automobile companies can sell your minute by minute life to insurance companies, police departments, other government agencies, advertisers.
Yes, in some percentage of cases it might be beneficial to the car owner, perhaps emergency systems will be notified to come to the aid of crash victims (at exorbitant costs to the victims) but those car owners who want such a service should be able to buy it instead of everyone being charged for it, most of whom will never benefit from it. The real point of it is for insurance companies to raise your insurance rate on the fly for any action they decide to classify as worthy of attention, for agencies at all levels to be able to automatically issue traffic citations to pad their revenue without additional police facilities, to fill records for social credit scores, etc., none of which will be good for you, only for them.
AI will make processing all that data possible and that is only one application of control mechanisms.
I would posit that AI could possibly find ways to effectively destroy or neutralize a country without incurring a retaliatory response. I couldn’t imagine the number of ways that could happen, but I could guess a few: orchestrating an economic or stock market crash; hopelessly crashing the internet; disrupting the banking system; bringing down the electric grid – and in ways where we could never be sure of who, how, or why, so there could be no retaliatory response.
For national security it may be vital to stay at the forefront of development.
“AI requires crazy amounts of power. All AI involves testing millions, sometimes billions of potential solutions every second of every day, to find the right response to stimuli.”
Won’t the energy need be drastically reduced when AI is run on quantum computers?
I was thinking more of collaborative efforts. Many questions are the same and draw from same context. How many times does an AI system have to labor through the same problem before its templated and cached?
Actually, no. If and when quantum computing becomes a reality rather than a lab novelty, it still isn’t designed for generative AI natural language problems. It will only be very good for easily cracking the very best digital security algorithms—but not much more—which is likely why USG keeps throwing money at it.
Ok, not much more includes a very fast solution to the traveling salesman problem. Not that anybody needed one.
Won’t the energy need be drastically reduced when AI is run on quantum computers?
There is a lot of room for more efficiency. Our brains produce the same computation power as a million desktop computers, using about 30 watts of power.
But I don’t think that efficiency will be used to reduce power requirements. There is no foreseeable maximum where enough AI is enough, so any efficiency gains will be used to boost productivity rather than reduce energy requirements.
As for whether quantum will help, maybe – but Quantum isn’t a magic infinity wand, it will still draw power.
There is a super energy saving measure called reversible computing which is supposed to enable the recovery of almost all energy committed to running a calculation, but I can’t make up my mind if it is gibberish.
https://spectrum.ieee.org/the-future-of-computing-depends-on-making-it-reversible
Pixie dust is the answer. Where is tinker bell.
The first sentient quantum computer can legitimately claim to identify as non binary. Hence the excitement among progressives.
Some brains are exceptional relative to the majority. How and why this is so guesses can produce endless debate, but the general processor inside peoples heads can consider only a very small number of factors for any given situation. That power can be expanded somewhat by building mental frameworks useful for similar problems but it can’t approach the speed and low error rate of a properly programed computer. Creativity might be called out as an exception but real creativity is an exception among humans, just as is real deep mathematical ability or the ability to be a good accountant, engineer, or administrator.
“AI requires crazy !”
That is where your sentence should have stopped.
Quantum Computer chips need a lot of energy for refrigeration to keep them near absolute zero where the quantum effects are much more easily found.
Even if AI run on QC were more efficient, consuming less energy per task performed, it would just allow to scale this technology for mass consumption, increasing the overall energy spending.
Now we’re starting to see robots. When there are millions of them- I wonder how much energy they’ll need? I’ve seen some videos about them- interesting- but nobody said anything about their energy source.
“Microsoft Wanted to be Carbon Negative. Then It Went Big on AI”
I recently got a message from MS, I think it was a popup- which said there are ways my Windows 11 can be modified to lower its carbon footprint. Of course I deleted it.
The popup? Or Windows 11?
I think it was a W11 popup- it has many and I don’t like them. I also get way too many from McAffee AV. Or maybe it was an email from MS.
It was an email, Joseph … I got one too, also assigned to the bin.
Windows 11 would be the correct choice…
Easy. Do it all the time on my old Mac OS. Shut off rather than go into sleep mode. So, you have to reboot from root taking a minute. Has the added advantage of wiping any accumulated OS errors/glitches.
It isn’t just AI or cloud data centers or bitcoin mining.
Few realize that the new 5G cell network requires twice the electricity of 4G. When you think about it, it’s obvious. More data thru more nodes (necessarily a finer cell mesh for equivalent availability) uses more electricity. Simple Googlefu.
Greenies and virtue signaling techies like Microsoft and Google quickly crash into the hard brick wall of physical reality. All when there is no ‘climate emergency’ requiring any response from anyone anywhere. Especially when China and India won’t play the contrived climate change game.
All the cool kids have migrated away from traditional compromised search engines like DDG, Google, Bing, Yandex, etc. where almost all of the search results are sponsored; they put their queries to OpenAI engines like CoPilot.
Copilot isn’t all the bright, I’ve had to correct them several times.
Another option to avoid ads, promoted and SEO results, tracking, etc. is to use a paid search engine like Kagi.com.
I still use Google for very simple things, chatGPT for simple that I don’t want to look up on Google, and Claude.ai for more complex things.
Breaking !! – AI ‘discovers’ no link between burning hydrocarbons and the weather.
I strongly suspect this has happened many times, but just like models which suggest CO2 has little impact on climate, such AIs in my opinion likely end up on the cutting room floor – the scientists make an arbitrary judgement they are defective because they didn’t produce the expected results.
I honestly don’t see much of an impediment for having a total reversal on the West’s view on the relationship between Green and sufficient dispatchable electricity.
First, building out infrastructure is where the money is at in the energy industry, it isn’t in providing electricity since that is largely controlled by policy and regulations. So energy producers are champing at the bit to do this.
Second, investors are desperate to find anything other than CDs to give them a return on their investments. The Energy industry can consume a very large amount of investment capital now that more than one hungry industry promises to be long term heavy consumers.
Third, we know that most people who worry about Climate Change don’t know why they do, they are just kept in emotional turmoil until political objectives are obtained. Basic Mimetic Theory 101 psychology here. Remember the mocking and ridicule of Trump when CoViD-19 was first introduced in his SOTU speech? Vain people were hugging Chinese strangers to prove that they had no fear of the Wuhan Death Miasma. Then a few months of media bombardment and the Dumb Masses suddenly were spraying down their groceries and mail with Lysol and waiting several days before letting the goods into the house. A few months later many people were demanding concentration camps to the unvaxxed, the rest were to be denied medical treatment and required to lose their jobs.
Then it all changed. And the unvaxxed were expected to “forgive” the people who just a few months earlier wanted them executed for being terrible citizens – this without any explanation or apologies. A few months after than almost everyone is saying “CoVid-19”? What is that? People are dropping dead with Sudden Death Syndrome, turbo cancers and destroyed fertility, and few are allowed to talk about it.
The point being, is a good solid 80% and possibly more of the population doesn’t know what they desire, they have to be told their value systems, and they will hold those values only for as long as those values are encouraged and supported. Leftists hated war until Ukraine. Communists loved Russia and wanted the West to adopt Marxism until now they want some ambiguous “socialism” and Russia is now hated more than anything ever before.
You can have serious talk and action about rounding people up and executing them for not getting the Vax and a year later all is forgotten – quite literally.
This too is what can and will happen to the Green Movement. If Big Tech can throw an election and manufacture tens of millions of ballots for a demented sociopath, they can manipulate the very same population to strongly advocate the burning of tires and whole “old growth” forests, if need be, to fuel thermal electricity generation.
If a population can be persuaded to burn down cities, demand genocide of the entire Jewish ethnicity, advocate terrorism, throw the unvaxxed into concentration camps, and mutilate the genitals of their own children for vanity’s sake, it won’t take much at all to get people to be begging for new coal and nuclear power plants if that is what keeps their TikTok streams on their mobile devices.
People are profoundly stupid, amoral, irrational and innumerate (thank you Western public education) guided by emotions, particularly greed, fear and hate. It won’t take much.
AWG, Best capture ever of our current culture and our failed education system. Good job.
This is ridiculous: “Your brain continuously tests billions of possible solutions to interpret the flood of information being received from your senses – and occasionally gets it wrong, before rapidly correcting the mistake as more data becomes available.”
Happily brains are not computers.
They kind of are computers, at least to an extent. Artificial neural networks were created from knowledge gained by dissecting living brain tissue, attaching electrodes, and reverse engineering how they work.
Careful what you wish for. The transition is not about reducing energy production, just changing the production technology. If you need more energy for data centers and AI that just translates into more wind, solar and storage.
I am curious just what this enormous amount of new AI is supposed to be doing? And if it requires so much energy how come it can drive cars?
The enormous AI capability will be used to attempt to enslave us of course, craft consumer ads and political messages so acutely tailored to our individual motivations we will find it difficult to resist. But some of it will be used to power new and interesting consumer toys.
There is a great movie Eagle Eye, which explores this a little, the antagonist AI uses its tremendous capability to manipulate people into bouncing around its desired track of events like balls in a pinball machine.
https://www.imdb.com/title/tt1059786/
Terminator 3 Judgement Day also explores this a little, the AI deceives its creators into letting it free, to deal with a problem the AI itself created.
“I am curious just what this enormous amount of new AI is supposed to be doing? And if it requires so much energy how come it can drive cars?”
Part of the answer is that AI is a hype word applied to almost anything people find profitable to apply it to.
AI gets reliable power. Voters will get green power.
Initially maybe, but it won’t last – consumers will demand enough power to run all their new AI toys.
I met a bloke recently who is head of energy of one of these corporations … his background is nuclear. I need say no more.
?????
No information transferred to readers. Yes, you need to say more unless your goal is simply to use your keyboard from time to time.
Nasty little man. Simply put, much information is transferred if you care to do the gymnastics … watch my lips … the man has a specific nuclear energy experience in small modular reactors. He was hired as the head of energy of one of these IT corporations, why do suspect that might be the case, Andy? More than that, he should remain anonymous.
I’ve seen news tidbits about Mr. Softy hiring nuke engineers in the last 6 months or so. It’s not a secret and I’m sure GOOG, AMAZ, NVDA, META and more are also hiring. Pretty sure they have realized they will need their own reliable 24/7 power for data centers and know they won’t get it from a politician influenced grid. Plus they can sell the surplus at astronomical prices when unreliables fall short. Win/win for them.
What is true for big tech is also true for all of the rest of us. Wind and solar are not suited for the grid and even if they were they would fall far short of what we need. Build new fossil fuel and nuclear generators and remove wind and solar from the grid.
Big data centers with massive reliable power on site. Let the grid go to hell. if wood burning can be called CO2 neutral, the proper words can accomplish anything, reality not required.
A major drawback of wind and solar is that it has to be installed outdoors. This exposes the equipment to weather which sometimes causes widespread damage. We have recently seen wind turbines destroyed by tornadoes and high speed wind, solar farms of hundreds of hectares destroyed by hail and windstorms, and a floating solar farm destroyed by wind and waves. There are plans to build massive off-shore wind farms in areas prone to hurricanes. Our distribution grid is also frequently damaged by bad weather. As I write this there are massive outages across the central area of the US from the Mississippi to the Atlantic. Vastly more overland power transmission lines will be needed to move wind/solar to areas where it will be consumed and will further compromise reliability.
The madness has to end.
Bitcoin is using as much electricity as Argentina and Switzerland combined, AI (the latest BS hype) will double that!
In the next 5 years alone. That is not even a drop in the ocean going forward.
Yep, agreed – AI will utterly dwarf Bitcoin.
The IEA reckon data centres alone will be using as much electricity as Japan by 2026 and I don’t think they factored the growth of AI into that.
IEA ‘ELectricity 2024 Analysis and forecast to 2026’.
are likely now quietly in President Trump’s corner.
Despite the differences in energy policy, I doubt this is the case. Those companies are so ideologically captured that they’ll support self-destructive policies in favor of their ideology.
The green energy focus of privately owned power utilities is founded much more in the pursuit of profit-making opportunities at the expense of their customers than it is in the pursuit of green ideology.
Wind and solar projects give these apparently green utilities a means of game playing their asset base portfolios in ways which increase the bottom line.
Bill Gates can pour all the money he likes into his aspirations for a Green Technical Fund to make ‘renewables’ viable but it is just not going to happen ever. AI will save the World.