Sorry but I don't know the answer. I was trained in Engineering but my working life was in industrial control software. I'm skeptical about AI but I follow a chap called Sasha Yanshin who is a Russian financial advisor living in the UK. You might like to look at some of the first 8 of his videos about the big tech companies and how they are borrowing huge amounts of money to fund their AI ambitions and also how they are faking revenue in order to attract investors. I don't know how that ties in with the 25% margins you mentioned.
Our UK Prime Minister has spoken about his ambitions to make the UK a world leader in AI but seems to have no plan on how to achieve it especially since our electricity supply is increasingly dependent on renewables.
Thanks Steve, 25% is just the glossy investment portfolio number however I had heard that there may be a significant amount of of balance sheet financing which is what I think you’re pointing to. The cash cow could indeed be all smoke and mirrors.
When I was a moderator at PeakOil.com, I noticed a strange quirk that while geologists have a flawless understanding of the Earth’s crust, they occasionally harbor some truly wacky ideas about human ecology.
Petroleum geologist Art Berman attempted to soothe anxieties about the coming end of growth by redefining ‘collapse’. Relying on Joseph Tainter’s theory of the “loss of complexity”, Berman used the ancient city of Angkor Wat as his historical model.
According to Berman, when the empire fell, Angkor Wat didn’t instantly turn into a Mad Max wasteland. Instead, as the aqueducts silted up, the residents simply performed a “cost-benefit analysis”. Deciding the city was no longer worth the upkeep, they slowly packed up and trickled down the Mekong River to live in smaller, simpler communities.
Berman extrapolates this to the modern global crisis: Don’t worry, he implies, we aren’t facing a sudden cliff where billions die; we are just facing a slow, manageable downsizing where we voluntarily shed complexity.
This analogy is not just flawed; biophysically, it is insane. It suggests that 9 billion people are simply going to “leave” the industrial system—but where exactly are they going to go? Will their cost-benefit analysis conclude that they should simply stop eating?
Here is why the Angkor Wat analogy fails the test of modern thermodynamics.
1. The Closed System Problem (There is No ‘Outside’)
In the 12th century, the residents of Angkor Wat had a luxury that modern humanity does not possess: an ‘Outside’.
When the local urban carrying capacity failed, the population could physically walk a few dozen miles into a sparsely populated, fertile river delta and revert to subsistence foraging and agriculture. The surrounding ecosystem had an abundance of natural Exergy (solar energy, fertile soil, wild game) waiting to be tapped.
Today, the entire planet is Angkor Wat.
We operate in a hyper-optimised, closed global system. Every arable inch of topsoil is accounted for. If the global logistical grid (which delivers the diesel, the seeds, and the water) breaks down, the populations of megacities like Tokyo, Cairo, or London cannot simply “walk down the river” to farm. There is no empty, fertile frontier waiting to absorb 9 billion subsistence farmers.
2. The Haber-Bosch Cliff
Berman forgets that the modern human population is not a product of natural carrying capacity; it is an artifact of fossil fuels.
Roughly half the nitrogen in the bodies of the global population was synthesised in a factory using the Haber-Bosch process, which turns natural gas into ammonia fertiliser. We are literally eating natural gas.
When Angkor Wat’s aqueducts failed, the population lost a convenience. When the global fossil-fuel supply chain fails, the population loses its fundamental biological building block.
You cannot apply a “cost-benefit analysis” to starvation. If the Energy Return on Energy Invested (ERoEI) drops below the threshold required to manufacture and transport synthetic fertiliser, the Earth’s carrying capacity does not slowly “downsize” over a century. It drops off a cliff.
3. The Violence of Inelastic Demand
Berman assumes humans will rationally “shed complexity” like a household trimming its budget. This ignores the inelasticity of biological demand.
When a household is stressed, it cancels its Netflix subscription. When a biophysical system of 9 billion people is stressed by a lack of caloric input, it does not peacefully disperse; it turns kinetic. The loss of complexity is not a quiet migration; it is a violent scramble for the remaining low-entropy gradients (food, water, fuel).
4. The Comforting Delusion
Berman’s use of the Angkor Wat analogy is a psychological defence mechanism. It is an attempt to look at the terrifying mathematical reality of the Resource Entropy Singularity and make it palatable.
He wants to believe that because collapse was slow in a pre-industrial, low-population agrarian society, it will be slow in a hyper-connected, artificially sustained, high-population industrial society.
But physics does not negotiate. When you pull the plug on an artificial life-support system, the patient doesn’t slowly reConclusionvert to a hunter-gatherer lifestyle. The patient crashes. Nine billion people cannot simply ‘walk away’ from the only machine keeping them alive.
Control of the data is what the powerful desire. Supplying cheap power for other people to own/process data puts NZ right where it is with eg the aluminium smelter, milk powder supply, delivering commodity product which is bought at the cheapest possible price. Rio Tinto’s bargaining for power at Bluff makes it 100% clear how that future looks. Not to mention we need the power for our own people! It’s not like we have a lot of spare. Data centres to manage our own data - yes. For others - no.
For a country that has a significant shortfall in electricity generation capacity as has been discussed at length it is shear folly to even consider building AI datacentres in New Zealand. The energy requirements are absolutely massive plus they require large quantities of water for their cooling.
Do we really want to add a huge consumption asset/s to the mix considering we are already losing industries as a result of energy supply and price issues?
At least with Tiwai Point, and any other aluminium smelter or steel mill, there are significant direct and indirect jobs created, money injected into the local community and a tangible value added product produced, data centres don't tick any of these boxes.
Insightful commentary. We're in a situation where we're losing industries ostensibly because of the cost of energy and looking at how to ration electricity by shedding demand yet we see these data centers as saviors? Also these companies have a clear track record of exporting any profits. IMO we're fools if we think we're winners in this deal.
Really enjoy how you couple energy consumption, industrialization and standard of living, a point missing in the public debate.
We don’t need a hyperscaler for NZ’s data needs, it would be an electricity export business and low margin. However, it would give us a bargaining position with Rio Tinto once we built and demonstrated capability with the first one.
Presumably the amount of data needs fibre links to other countries. Security of these links may constraint NZ as a international centre. I think there are currently 3 or 4 main fibre cables to NZ. In a conflict these would be cut fairly quickly.
We should control the storage of our own data, perhaps share with Australia and Pacific islands.
Thank you, Larry, this is a first class poser: Do we just want the electricity bill (as State revenue), or do we want to go further, and what does that require? One thing I do know is that it takes bucketloads of funding, up front (but is that viable if we’re borrowing every squillion), can we access the physical technology, and do we have the people who know how to build and operate this technology?
This reminds me of a time, way back in the early 80’s, when a brand new full-scale replica power plant simulator was delivered to a new power plant project in SE Queensland. It was just a smidge past the dawn of something called “the computer age”, and this new machine ran on a series of giant electro-mechanical processors - and lots of electricity to run the machine, to provide cooling for the computer equipment, and for the human environment.
When introduced to this machine, it was explained that complex scientific mathematics was the foundation for its programming and operation, and that the operators would need to quickly become proficient. That was us, some of whom still hadn’t been exposed to any form of computer or basic forms of programming.
Cutting a longish story short, and after hundreds of man-hours devoted by the simulator builder’s large team in futile efforts to commission the machine, they declared defeat… and walked away!
Over a few months, a small group of very determined, computer illiterate operators and a couple of control systems engineers, managed to not only get this behemoth working but also to refine most of its programmes. I last used that simulator in the late-2000’s, and it was an admirable training and engineering tool. Most of us moved on to newer technologies, in the same field.
My point: If we humans are capable of looking forwards and putting our best efforts into making complex technology work, then New Zealand surely is capable of extracting the best outcomes from a State-funded investment in this new field of “data refining”. Throw out the slide-rules and spanners, and get going.
But, first take giant steps to assure an electricity system which is capable of advancing the Nation’s prosperity. Larry’s efforts have been providing the awareness, knowledge and understanding that is needed to do what’s best for everyone.
If you mention ‘Peak Oil’ today in polite company—or even among serious economists—you are immediately filed away in the same cabinet as Y2K survivalists and Mayan Calendar enthusiasts.
It is viewed as a failed prophecy. The narrative goes like this: “Around 2005, a bunch of doomers said we were running out of oil. They predicted Mad Max by 2010. Instead, we got the shale revolution, the iPhone, and a stock market that only goes up. Technology saved us. Malthus was wrong. Again.”
We do have a need for local secure data centres - probably public - if we want all the data the government gathers to be held in Aotearoa, not overseas.
Agreed re the need to think about value chains if claiming to be investing for externalities from economic development. I've written elsewhere about this in respect of our investments in R&D https://www.linkedin.com/pulse/nzs-public-investment-rd-doing-better-simon-arnold-ka4oc/. In the case of Marsden Pt there is perhaps bio refining to fill out the chain, if it is going anywhere.
On the electricity-to-data prospect it strikes me as commodities all the way up, and we have limited control over the end market without a bit of regulatory intervention.
It has to be likely there will be higher value opportunities for what energy surpluses we might have, perhaps adding value to what we're already good at. I've recently been looking at the prospects for quantum technologies for NZ and rather proving this last point the big potential wins for NZ appear to be in using quantum sensors (sourced from wherever) to our primary sector.
Sorry but I don't know the answer. I was trained in Engineering but my working life was in industrial control software. I'm skeptical about AI but I follow a chap called Sasha Yanshin who is a Russian financial advisor living in the UK. You might like to look at some of the first 8 of his videos about the big tech companies and how they are borrowing huge amounts of money to fund their AI ambitions and also how they are faking revenue in order to attract investors. I don't know how that ties in with the 25% margins you mentioned.
https://www.youtube.com/@SashaYanshin/videos
Our UK Prime Minister has spoken about his ambitions to make the UK a world leader in AI but seems to have no plan on how to achieve it especially since our electricity supply is increasingly dependent on renewables.
Thanks Steve, 25% is just the glossy investment portfolio number however I had heard that there may be a significant amount of of balance sheet financing which is what I think you’re pointing to. The cash cow could indeed be all smoke and mirrors.
REALITY CHECK TIME
The Angkor Wat Fallacy
When I was a moderator at PeakOil.com, I noticed a strange quirk that while geologists have a flawless understanding of the Earth’s crust, they occasionally harbor some truly wacky ideas about human ecology.
Petroleum geologist Art Berman attempted to soothe anxieties about the coming end of growth by redefining ‘collapse’. Relying on Joseph Tainter’s theory of the “loss of complexity”, Berman used the ancient city of Angkor Wat as his historical model.
According to Berman, when the empire fell, Angkor Wat didn’t instantly turn into a Mad Max wasteland. Instead, as the aqueducts silted up, the residents simply performed a “cost-benefit analysis”. Deciding the city was no longer worth the upkeep, they slowly packed up and trickled down the Mekong River to live in smaller, simpler communities.
Berman extrapolates this to the modern global crisis: Don’t worry, he implies, we aren’t facing a sudden cliff where billions die; we are just facing a slow, manageable downsizing where we voluntarily shed complexity.
This analogy is not just flawed; biophysically, it is insane. It suggests that 9 billion people are simply going to “leave” the industrial system—but where exactly are they going to go? Will their cost-benefit analysis conclude that they should simply stop eating?
Here is why the Angkor Wat analogy fails the test of modern thermodynamics.
1. The Closed System Problem (There is No ‘Outside’)
In the 12th century, the residents of Angkor Wat had a luxury that modern humanity does not possess: an ‘Outside’.
When the local urban carrying capacity failed, the population could physically walk a few dozen miles into a sparsely populated, fertile river delta and revert to subsistence foraging and agriculture. The surrounding ecosystem had an abundance of natural Exergy (solar energy, fertile soil, wild game) waiting to be tapped.
Today, the entire planet is Angkor Wat.
We operate in a hyper-optimised, closed global system. Every arable inch of topsoil is accounted for. If the global logistical grid (which delivers the diesel, the seeds, and the water) breaks down, the populations of megacities like Tokyo, Cairo, or London cannot simply “walk down the river” to farm. There is no empty, fertile frontier waiting to absorb 9 billion subsistence farmers.
2. The Haber-Bosch Cliff
Berman forgets that the modern human population is not a product of natural carrying capacity; it is an artifact of fossil fuels.
Roughly half the nitrogen in the bodies of the global population was synthesised in a factory using the Haber-Bosch process, which turns natural gas into ammonia fertiliser. We are literally eating natural gas.
When Angkor Wat’s aqueducts failed, the population lost a convenience. When the global fossil-fuel supply chain fails, the population loses its fundamental biological building block.
You cannot apply a “cost-benefit analysis” to starvation. If the Energy Return on Energy Invested (ERoEI) drops below the threshold required to manufacture and transport synthetic fertiliser, the Earth’s carrying capacity does not slowly “downsize” over a century. It drops off a cliff.
3. The Violence of Inelastic Demand
Berman assumes humans will rationally “shed complexity” like a household trimming its budget. This ignores the inelasticity of biological demand.
When a household is stressed, it cancels its Netflix subscription. When a biophysical system of 9 billion people is stressed by a lack of caloric input, it does not peacefully disperse; it turns kinetic. The loss of complexity is not a quiet migration; it is a violent scramble for the remaining low-entropy gradients (food, water, fuel).
4. The Comforting Delusion
Berman’s use of the Angkor Wat analogy is a psychological defence mechanism. It is an attempt to look at the terrifying mathematical reality of the Resource Entropy Singularity and make it palatable.
He wants to believe that because collapse was slow in a pre-industrial, low-population agrarian society, it will be slow in a hyper-connected, artificially sustained, high-population industrial society.
But physics does not negotiate. When you pull the plug on an artificial life-support system, the patient doesn’t slowly reConclusionvert to a hunter-gatherer lifestyle. The patient crashes. Nine billion people cannot simply ‘walk away’ from the only machine keeping them alive.
https://theuaob.substack.com/p/the-compartmental-fracture-of-art
Control of the data is what the powerful desire. Supplying cheap power for other people to own/process data puts NZ right where it is with eg the aluminium smelter, milk powder supply, delivering commodity product which is bought at the cheapest possible price. Rio Tinto’s bargaining for power at Bluff makes it 100% clear how that future looks. Not to mention we need the power for our own people! It’s not like we have a lot of spare. Data centres to manage our own data - yes. For others - no.
For a country that has a significant shortfall in electricity generation capacity as has been discussed at length it is shear folly to even consider building AI datacentres in New Zealand. The energy requirements are absolutely massive plus they require large quantities of water for their cooling.
Do we really want to add a huge consumption asset/s to the mix considering we are already losing industries as a result of energy supply and price issues?
Data centre energy demands well detailed here: https://youtu.be/dhqoTku-HAA
The arguments seem rather analogous to those amount aluminium smelters which for a very long time bargain for cut price energy to keep going.
At least with Tiwai Point, and any other aluminium smelter or steel mill, there are significant direct and indirect jobs created, money injected into the local community and a tangible value added product produced, data centres don't tick any of these boxes.
Insightful commentary. We're in a situation where we're losing industries ostensibly because of the cost of energy and looking at how to ration electricity by shedding demand yet we see these data centers as saviors? Also these companies have a clear track record of exporting any profits. IMO we're fools if we think we're winners in this deal.
Really enjoy how you couple energy consumption, industrialization and standard of living, a point missing in the public debate.
We don’t need a hyperscaler for NZ’s data needs, it would be an electricity export business and low margin. However, it would give us a bargaining position with Rio Tinto once we built and demonstrated capability with the first one.
Presumably the amount of data needs fibre links to other countries. Security of these links may constraint NZ as a international centre. I think there are currently 3 or 4 main fibre cables to NZ. In a conflict these would be cut fairly quickly.
We should control the storage of our own data, perhaps share with Australia and Pacific islands.
Plus develop other means of transmission.
Thank you, Larry, this is a first class poser: Do we just want the electricity bill (as State revenue), or do we want to go further, and what does that require? One thing I do know is that it takes bucketloads of funding, up front (but is that viable if we’re borrowing every squillion), can we access the physical technology, and do we have the people who know how to build and operate this technology?
This reminds me of a time, way back in the early 80’s, when a brand new full-scale replica power plant simulator was delivered to a new power plant project in SE Queensland. It was just a smidge past the dawn of something called “the computer age”, and this new machine ran on a series of giant electro-mechanical processors - and lots of electricity to run the machine, to provide cooling for the computer equipment, and for the human environment.
When introduced to this machine, it was explained that complex scientific mathematics was the foundation for its programming and operation, and that the operators would need to quickly become proficient. That was us, some of whom still hadn’t been exposed to any form of computer or basic forms of programming.
Cutting a longish story short, and after hundreds of man-hours devoted by the simulator builder’s large team in futile efforts to commission the machine, they declared defeat… and walked away!
Over a few months, a small group of very determined, computer illiterate operators and a couple of control systems engineers, managed to not only get this behemoth working but also to refine most of its programmes. I last used that simulator in the late-2000’s, and it was an admirable training and engineering tool. Most of us moved on to newer technologies, in the same field.
My point: If we humans are capable of looking forwards and putting our best efforts into making complex technology work, then New Zealand surely is capable of extracting the best outcomes from a State-funded investment in this new field of “data refining”. Throw out the slide-rules and spanners, and get going.
But, first take giant steps to assure an electricity system which is capable of advancing the Nation’s prosperity. Larry’s efforts have been providing the awareness, knowledge and understanding that is needed to do what’s best for everyone.
If you mention ‘Peak Oil’ today in polite company—or even among serious economists—you are immediately filed away in the same cabinet as Y2K survivalists and Mayan Calendar enthusiasts.
It is viewed as a failed prophecy. The narrative goes like this: “Around 2005, a bunch of doomers said we were running out of oil. They predicted Mad Max by 2010. Instead, we got the shale revolution, the iPhone, and a stock market that only goes up. Technology saved us. Malthus was wrong. Again.”
https://theuaob.substack.com/p/the-thermodynamic-blind-spot-why
We don’t need a hyperscaler for own needs
We do have a need for local secure data centres - probably public - if we want all the data the government gathers to be held in Aotearoa, not overseas.
That is already the case. NZ and other governments have data sovereignty requirements for local only data storage. Many corporations require the same.
We already have many data centers and more planned.
I understand that there is government data held in Sydney.
Agreed re the need to think about value chains if claiming to be investing for externalities from economic development. I've written elsewhere about this in respect of our investments in R&D https://www.linkedin.com/pulse/nzs-public-investment-rd-doing-better-simon-arnold-ka4oc/. In the case of Marsden Pt there is perhaps bio refining to fill out the chain, if it is going anywhere.
On the electricity-to-data prospect it strikes me as commodities all the way up, and we have limited control over the end market without a bit of regulatory intervention.
It has to be likely there will be higher value opportunities for what energy surpluses we might have, perhaps adding value to what we're already good at. I've recently been looking at the prospects for quantum technologies for NZ and rather proving this last point the big potential wins for NZ appear to be in using quantum sensors (sourced from wherever) to our primary sector.