The funniest part is beyond the typo, the complete lack of physical intuition from the analysts who circulated this. 500,000 tons is roughly the weight of 1.5 Empire State buildings. If your rack busbars weigh more than the structural steel of the facility housing them, you have a geotechnical engineering crisis on your hands. It is wild that we reached a point where financial modeling is so decoupled from physical reality that nobody paused to ask if the floor would collapse.
When I asked gemini it got it right away and as a source it provided a link to your comment... what a time to be alive, I hope you got the Empire State buildings calculations right because you're now part of the Truth ™
The irony is that I feel like everyone here is just copping off of WolframAlpha, which for the past 15 years has featured the ability to display references of similar magnitude for any query you give. And in fact, typing in "500,000 tons" into WolframAlpha tells you specifically that this is 1.4x the mass of the Empire State Building (and also 0.8x to 1.4x the mass of an Ultra Large Crude Carrier oil supertanker, and also 0.11x the mass converted to energy by the sun in one second).
The world of financial analysis and modeling is broad. It’s common to give these tasks to juniors and expect them to grind through it when the output doesn’t really matter.
In this case the output wasn’t actually used for financial modeling. If it had been, it would have been caught immediately when someone put it into a table where they calculated the price or the supply constraints or anything else.
Just to make it even more real: During covid I added a sub-panel and the wire (more like the sausage given the girth) between the sub-panel and main panel was aluminum because of cost. You just need to be a tad careful at the connection points with copper -- nothing a caring literate person can't handle
The incoming drop for a service panel is almost always aluminum (at least the residential stuff I've seen). It's cheaper and lighter than copper while still being ductile enough to work with. Just use the proper antioxidation paste at the connection point.
Did you use antioxidant paste on all of the aluminum terminations? If you didn’t, please pay an electrician to come fix it. If you did apply the paste, you are in the 0.01% of non-idiot non-electrician homeowners and can skip my rant about homeowners doing electrical work in the next paragraph. I sell electrical work and have seen some incredibly poor work done by homeowners.
Do not install aluminum wire if you are a homeowner unless you already know enough to use antioxidant paste and also to use a torque screwdriver or torque wrench for terminations and know where to find the torque values for the wiring devices you are using. If you don’t apply the paste, the surface of the aluminum will oxidize and could catch fire due to increased resistance.
I suggest not even touching copper as a homeowner, but it’s your choice and your house.
My first thought was, this is the kind of thing that an LLM writes and nobody checks. But then I realized, any decent LLM would have probably caught that inconsistency
Yeah, the good ones will use tools and give you a nice markdown table. Any claims like this anymore, I trust AI to go through it and give the numbers a reality check. With all the claims made about datacenter resources, power, and water usage, it's darkly funny how bad people are at understanding these things.
The math they do with their assumptions is usually pretty good, and you can tell when they put in the effort, but wow, the models and assumptions are all over the place.
Kinda like when someone said that instead of running for mayor, Bloomberg could have given everyone in the country $1M. A guest said it on NBC and Bryan Williams, nor the pundit had any intuition that it seemed grossly wrong. https://www.politifact.com/factchecks/2020/mar/06/msnbc/bad-...
I wouldn't be surprised if that part was not really reviewed by an expert. They have the unit mass correct but maybe an editor is like ok but what does this look like for a gw project? It doesn't take more than 3rd grade math and a pocket calculator to do it correctly but journalist hasn't had to fumble that ball before. An expert knows its all too easy for any person to make that mistake and would second guess their own work.
With regards to the copper market: it keeps surprising me that some people seem to assume copper is a hard requirement for conducting electricity.
In reality copper is just convenient. We use it because it's easy to work with, a great conductor, and (until recently) quite affordable. But for most applications there's no reason we couldn't use something else!
For example, a 1.5mm2 copper conductor is 0.0134kg/m, which at current prices is $0.17 / meter. A 2.4mm2 aluminum conductor has the same resistance, weighs 0.0065kg/m, which at current prices is $0.0195 / meter!
Sure, aluminum is a pain to work with, but with a price premium like that there's a massive incentive to find a way to make it work.
Copper can't get too expensive simply due to power demands because people will just switch to aluminum. The power grid itself had been using it for decades, after all - some internal datacenter busbars should be doable as well.
And you can browse that site for lugs, and they’re mostly rated for aluminum and copper. Copper-only lugs are actually rather unusual.
Sure, you can’t stick copper and aluminum wires into a wire nut, and finding terminations for smaller-gauge aluminum wire can be hard. But for larger wire, it’s really no problem.
Most wiring device terminations and lugs are Al/Cu rated. You’re supposed to use specific fittings for splicing together aluminum and copper wire (specific WAGO lever nuts are UL listed for this purpose along with some others), I assume that’s what you meant.
One good reason for moving from copper to aluminum is that copper contaminates recycled steel, but aluminum doesn't. Fortunately these days it's profitable to carefully extract most of the copper wire and such from scrap steel before it's melted down.
I am not an electricity/wiring guy so maybe you can help me understand. I thought aluminum is dangerous to wire with because it is a fire hazard (I bought a home this year and this was a prominent warning in my reading). Is that because it needs to be done very carefully? I imagine most data centers would not mess with a fire risk on such a scale.
Residential aluminum is a Really Bad Idea because DIY Dave will inevitably do something wrong - which then leads to a fire hazard. Copper is a lot more forgiving.
But a large scale datacenter, solar farm, or battery storage installation? Those will be installed and maintained by trained electricians, which means they actually know what a "torque wrench" is, and how to deal with scary words like "corrosion" and "oxidation".
Like I said: it's what's used for most of the power grid. With the right training it really isn't a big deal.
Aluminum got a bad rep due to a lot of poor installations made in the 1960 and 70'ies. Subsequently new alloys (AA-8000 series) for wiring, new installation and termination procedures etc. have been developed, and AFAIU the situation today is much better. It's still trickier than copper wiring, so probably not a good option for DIY, but for some industrial installation with competent personnel and where the savings can be substantial it could certainly be an attractive option.
Aluminum oxide has high resistance and if you mix aluminum wiring with copper outlets, etc the impedance mismatched is what causes fires. You need to either have special copper pigtails installed or use fixtures that are rated for aluminum wiring.
For commercial installs, it shouldn't be a problem as long as it's planned for.
It's because aluminium has a higher coefficient of thermal expansion. It expands and shrinks more as it heats, and as those cycles add up it tends to loosen electrical connections. Loose connections have higher resistance, heat up and can cause fires.
That said, there is no reason we can't design better connectors that can withstand the expansion and shrinkage cycles, like spring loaded or spring cage connectors.
Old aluminum wiring in your walls with cloth insulators, designed for a time where electricity consumption was a small fraction of today's electrified usage is dangerous because you're overloading an old, unprepared system.
Aluminum bus bars(solid, often exposed) would be designed for the required power levels and installation criteria.
Aluminum home wiring was from the 60s and 70s. It’s not the same as cloth covered knob and tube from earlier years. It has its own problems, but I’d take a house with knob and tube over a house with aluminum wiring.
Only true for lighting circuits though, and most household circuits are mixed.
The quantity and (edit: aggregate) power draw of modern appliances is far greater now than 60 years ago, so the overall load on the old wires is much higher.
I'd bet that modern TVs are more efficient that CRT televisions. Do most people even have desktop computers anymore, or have they mostly been replaced by laptops, tables, and phones? I'd be interested to see the efficiency numbers for electric clothes dryers over time. I wouldn't be surprised if they are also slightly more efficient than older models, even if they are still using resistance heating. Due to smarter electronics that automatically turn the unit off after the clothes are dry (air humidity sensor). I think electric ranges, dish washers, toasters and coffee machines have been ubiquitous since the 1960s (but are probably about the same energy-consumption wise). Air conditioning units are one thing that I'd believe are much more common today than in the 1970s and 1980s. Household sizes are also smaller, so less electricity used for electric water heaters, and the oven, etc.. Electric vehicles are an up and coming user of electricity. What other appliances are likely to be using more now than before?
These are good points, but having worked on a few older houses, I usually see overextended and overloaded circuits, not the opposite.
Standard small-house service used to be 60A, sometimes as few as 4 circuits! It's now 100A minimum by code, with 200A common.
Ovens/ranges have gone from 30A to 50A (dedicated) circuits by code. Microwaves also require dedicated circuits now. Gaming computers with big GPUs are common. Air fryers and electric pressure cookers are newly-common countertop appliances. People definitely use resistive electric space heaters more now (very cheap, much safer than the older options). And there's a trend away from gas and to electric ranges and water heaters. Heat pumps are also increasingly common. You mentioned air conditioners and EV chargers. Kitchens and bathrooms are now required to have dedicated (and GFCI) circuits. Household sizes are smaller, but houses are larger.
So I guess I'd say that, properly expanded, individual circuits should carry less current than they used to. But very often, appliances (AC, microwave, gaming rig, air fryers), are just "plugged in" to an unexpanded system, with varying results.
If you're lucky, they pop a breaker and you call an electrician. If you're not lucky, they push the power draw into uncomfortable zones, esp for Al wire.
This is absolutely not true in areas where heating the air and water and cooking are done with natural gas. Every single appliance in a house is more efficient today than in 1970 due to advances in motor speed control, without exception. The only thing that didn’t get more efficient is electric resistive heat and it’s impossible to improve on that anyways.
I can’t think of a single appliance from 1970 that consumes less energy than its modern equivalent. Anything with a pump or fan is more efficient and so is lighting. LCD TVs use less energy than CRTs.
I also can’t think of an appliance that has become common in households that draws more than 100 watts of continuous load since the 1970 aside from just ‘computers’. An ancient 500W 80% efficiency PSU at max load only has 5.2A of current at 120V single-phase.
If you convert your natural gas furnace to a heat pump, you will use more electricity but excluding that and NG to electric HPWHs leaves only more efficient equipment.
Sure, but there are more appliances plugged in today than there were. The simplest evidence for this is that there are never enough outlets in an old (unrenovated) home.
In a renovated house, you won't have aluminum wire at all, so these concerns are null.
My original statement should be qualified. Since we were talking about aluminum wire it's relevant -- an updated house will have new (copper) circuits that can handle all this stuff. An NON updated house might have Al wire and be overloaded in a more severe way than it was in the 60s.
But FWIW, new >100W appliances:
- microwaves (1200+W)
- air fryers (1500W)
- electric pressure cookers
- rice cookers (mine claims 610W on the plate)
- stand mixers (old: 80W, new: 475W)
- desktop computers (esp gaming rigs)
- resistive space heaters (1500W)
- *bigger* TVs (compare 72" LCD to 19" CRT?)
- air purifiers (mine clocks 175W on high)
- towel warmers? :)
- and the ubiquity of 10-20W small stuff has of course exploded, and it all adds up
Unless I'm mistaken, the risk with aluminum is that it can expand and contract if it gets too hot. Aluminum sized properly with the correct connectors torqued to spec would be fine, aluminum wires in a residence with a DIYer working on it can be riskier and is why inspectors will always note it.
Aluminum wires became brittle over time(tens of years), fluid which requires some maintenance for screw terminals and inducts galvanic corrosion when coupled with copper without special care. If wiring was properly done and maintained, it is okayish.
If you apply antioxidant paste, torque the terminations to the correct value, and use wiring devices rated for Cu and Al conductors, there are zero problems.
Most homeowners know none of the requirements of aluminum wire installation so I don’t recommend using it, oxidation is bad and can cause fires. Just pay for copper NM cable, it’s dirt cheap anyways.
FWIW I sell and run electrical work for a living. I assume crote and amluto work in the electrical industry, since they’re the only other posters in this discussion who know what they’re talking about.
It’s a fire hazard in residential houses where people frequently do their own wiring, because it needs more expertise to wire correctly. Copper wiring is a lot more forgiving to being hooked up by amateurs.
The biggest reason is that aluminum oxidizes, and unlike copper, the oxide layer has high resistivity. In theory that shouldn’t be an issue in datacenters hiring expert technicians.
Yea, in a data center setup the cost difference can justify the things like soldiering connections rather than mechanical connectors with a higher resistance.
> In the Earth's crust, aluminium is the most abundant metallic element (8.23% by mass[68]) and the third most abundant of all elements (after oxygen and silicon).
TIL. I thought it would be relatively expensive due to the difficulty of extracting it.
The Washington Monument has an aluminum cap, which at the time was as expensive as silver. Two years later the Hall-Héroult process was invented, and as a result the price plummeted.
That's why my calculation example used a 1.5mm2 copper wire but a 2.4mm2 aluminum one.
Aluminum has a higher resistance, which means the same diameter will get hotter than copper. Make the cable thicker and its resistance drops, which means it gets less hot.
Want more amps at the same temperature? Ohm's law still applies: just use a thicker cable.
It depends on design, environment and maintenance. If it's well protected against oxidation and it's static, it's not really a problem. Heat issues are often at the joints, which are most vulnerable, but can be coated and encapsulated to mitigate. For some uses, aluminium is superior to copper all things considered. Obviously the diameter of the wire/cable depends on which conductor you use, AC/DC, current and service temp - for some applications, that may favour one material over the other.
If the two wires are the same gauge, yes. If you size up the aluminum, at the same resistance/current would mean the same amount of power over the length of the conductor and same heat.
Thicker cables or higher voltage(lower current) is the answer which is why it's used in power distribution networks where they can control the voltage by planning what to transform to.
> In reality copper is just convenient. We use it because it's easy to work with, a great conductor, and (until recently) quite affordable
It's convenient, it's easy to work with, great conductivity, and cheap enough all at the sametime... Dude, I think you just explained why cropper is used instead of anything else.
Really? In larger sizes, an equivalent ampacity aluminum cable is generally lighter and more flexible than copper. The main downside is that it’s thicker.
(Common terminations for larger wire sizes are often dual-rated for aluminum and copper. The engineering details for how to design lugs that work well for aluminum and copper were worked out long ago.)
I would imagine most large-scale data center construction projects will include electrical engineers to design the electrical subsystem. A rack's floor footprint is a few square feet. You can put several million dollars of hardware into that rack. A data center will have at least a few racks. It's a very reasonable investment to bring someone in to do electrical design.
The goop is generally considered to be unnecessary. What you need is a termination that is rated for aluminum wire, and these are very, very common as long as the wire in question is fairly large (8AWG or so and larger).
I only put it in the comment since the commenters here seemed uninformed.
It's better to be safe and follow specs; it's more when you're doing some compound splice where it makes sense. I figure it'll make more people feel easy about it if there's a preventative measure
This sort of mistake is easy to make when you're mixing up your units; if they kept to one system of measure, it would've been trivial to catch, before or after release.
We need to standardize on using Earth circumferences as the unit of length. Or better, football fields! (the type of football of course being implied by the website's ccTLD)
We _have_ standardized on Earth circumferences for length, only we divide by 40 million to make the numbers more sane, and got the measurement slightly wrong!
How hard would it be to fix this? Could we theoretically add or subtract enough material or make whole thing slightly more dense or less dense to compensate?
Per Wikipedia, the discrepancy is approximately 74 km, so digging a ditch with an average depth of approximately 74/2π ≅ 12 km around the circumference of the Earth would theoretically fix the problem.
Feasibility and geological implications are left as exercises for the reader.
Regardless, I suspect a more cost-effective fix would be to redefine the meter to be a couple "legacy" millimeters longer.
You jest, but times around the Earth is the actual origin of the Meter. Kinda.
The history is quite interesting and well worth checking out.
I can't recommend a book on the subject, but I do heartily recommend "Longitude", which is about the challenges of inventing the first maritime chronometers for the purpose of accurately measuring longitude.
> If the "half a million tons" figure were accurate, a single 1 GW data center would consume 1.7% of the world's annual copper supply. If we built 30 GW of capacity—a reasonable projection for the AI build-out—that sector alone would theoretically absorb almost half of all the copper mined on Earth.
Quickly doing such "back of an envelope" calculations, and calling out things that seem outlandish, could be a useful function of an AI assistant.
Using your brain is so vastly more energy efficient, we might just only need half of that 30 GW capacity if fewer people had these leftpad-style knee-jerk reactions.
While I don't know how more or less efficient it is, WolframAlpha works well for these sorts of approximations, and shows its work more clearly than the AI chatbots I've used.
A Gemini query uses about a kilojoule. The brain runs at 20 W (though the whole human costs 100 W). So, the human is less energy if you can get it done in under 50 seconds.
Each person uses about 100W (2000kcal/24h=96W). Running all of humanity takes about 775GW.
Sure, using or not using your brain is a negligible energy difference, so if you aren't using it you really should, for energy efficiency's sake. But I don't think the claim that our brains are more energy efficient is obviously true on its own. The issue is more about induced demand from having all this external "thinking" capacity on your fingertips
Is there an AI system with functionality at or equal to a human brain that operates on less than 100W? Its currently the most efficient model we have. You compare all of humanity's energy expenditure, but to make the comparison, you need to consider the cost of replicating all that compute with AI (assuming we had an AGI at human level in all regards, or a set of AIs that when operated together could replace all human intelligence).
So, this is rather complex because you can turn AI energy usage to nearly zero when not in use. Humans have this problem of needing to consume a large amount of resources for 18-24 years with very little useful output during that time, and have to be kept running 24/7 otherwise you lose your investment. And even then there is a lot of risk they are going to be gibbering idiots and represent a net loss of your resource expenditure.
For this I have a modern Modest Proposal they we use young children as feed stock for biofuel generation before they become a resource sink. Not only do you save the child from a life of being a wage slave, you can now power your AI data center. I propose we call this the Matrix Efficiency Saving System (MESS).
No one will ever agree on when AI systems have equivalent functionality to a human brain. But lots of jobs consist of things a computer can now do for less than 100W.
Also, while a body itself uses only 100W, a normal urban lifestyle uses a few thousand watts for heat, light, cooking, and transportation.
> Also, while a body itself uses only 100W, a normal urban lifestyle uses a few thousand watts for heat, light, cooking, and transportation.
Add to that the tier-n dependencies this urban lifestyle has—massive supply chains sprawling across the planet, for example involving thousands upon thousands of people and goods involved in making your morning coffee happen.
Wikipedia quoted global primary energy production at 19.6 TW, or about 2400W/person. Which is obviously not even close to equally distributed. Per-country it gets complicated quickly, but naively taking the total from [1] brings the US to 9kW per person.
And that's ignoring sources like food from agriculture, including the food we feed our food.
To be fair, AI servers also use a lot more energy than their raw power demand if we use the same metrics. But after accounting for everything, an American and an 8xH100 server might end up in about the same ballpark
Which is not meant as an argument for replacing Americans with AI servers, but it puts AI power demand into context
Obviously we don't have AGI so we can't compare many tasks. But on tasks where AI does perform at comparable levels (certain subsets of writing, greenfield coding and art) it performs fairly well. They use more power but are also much faster, and that about cancels out. There are plenty of studies that try to put numbers on the exact tradeoff, usually focused more on CO2. Plenty that find AI better by some absurd degree (800 times more efficient at 3d modelling, 130 to 1500 times more efficient at writing, or 300 to 3000 times more efficient at illustrating [1]). The one I'd trust the most is [2] where GPT4 was 5-19 times less CO2 efficient than humans at solving coding challenges
I did some math for this particular case by asking Google’s Gemini Pro 3 (via AI studio) to evaluate the press release. Nvidia has since edited the release to remove the “tons of copper” claim, but it evaluated the other numbers at a reported API cost of about 3.8 cents. If the stated pricing just recovers energy cost, that implies 1500kJ of energy as a maximum (less if other costs are recovered in the pricing). A human thinking for 10 minutes would use sbout 6kJ of direct energy.
I agree with your point about induced demand. The “win” wouldn’t be looking at a single press release with already-suspect numbers, but rather looking at essentially all press releases of note, a task not generally valuable enough to devote people towards.
That being said, we normally consider it progress when we can use mechanical or electrical energy to replace or augment human work.
Nobody on HN is a bigger AI stan than I am -- well, maybe that SimonW guy, I guess -- but the truth is that problems involving unit conversions are among the riskiest things you can ask an LLM to handle for you.
It's not hard to imagine why, as the embedding vectors for terms like pounds/kilograms and feet/yards/meters are not going to be far from each other. Extreme caution is called for.
I edited the post with a speculation, but it's just a guess, really. In the training data, different units are going to share near-identical grammatical roles and positions in sentences. Unless some care is taken to force the embedding vectors for units like "pounds" and "kilograms" to point in different directions, their tokens may end up being sampled more or less interchangeably.
Gas-law calculations were where I first encountered this bit of scariness. It was quite a while ago, and I imagine the behavior has been RLHF'ed or otherwise tweaked to be less of a problem by now. Still, worth watching out for.
> In the training data, different units are going to share near-identical grammatical roles and positions in sentences.
Yes, but I would also expect the training data to include tons of examples of students doing unit-conversion homework, resources explaining the concept, etc. (So I would expect the embedding space to naturally include dimensions that represent some kind of metric-system-ness, because of data talking about the metric system.) And I understand the LLMs can somehow do arithmetic reasonably well (though it matters for some reason how big the numbers are, so presumably the internal logic is rather different from textbook algorithms), even without tool use.
It's almost always the engineers, analysts and MBA spreadsheet pushers and other people removed from the physical consequences outputting these mistakes because it's way easier to not notice a misplaced decimal or incorrect value when you deal in pure numbers and know what they "should" be than you are the person actually figuring out how to make it happen the difference between needing 26666666.667 and 266666666.667 <units> of <widget> is pretty meaningful. Engineers don't output these mistakes as often as analysts or whatever because they work in organizations that invest more in catching them, not because they make them all that much less.
Whether talking weight or bulk a decimal place is approximately the difference between needing a wheelbarrow, a truck, a semi truck, a freight train and a ship.
Around here, asking "does this number make sense?" when coming across a figure is second nature, reinforced since early in engineering school. The couple of engineers from the US that I know behave similarly, which makes sense because when your job is to solve practical problems and design stuff, precision matters.
> difference between needing 26666666.667 and 266666666.667 <units> of <widget> is pretty meaningful
To be fair, that’s why we’d use 2.6666666667e7 and 2.66666666667e8, which makes it easier to think about orders of magnitude. Processes, tools and methods must be adapted to reduce the risk of making a mistake.
Even among engineering fields routine handling of diverse and messy unit systems (e.g. chemical engineering) are relatively uncommon. If you work in one of these domains, there is a practiced discipline to detect unit conversion mistakes. You can do it in your head well enough to notice when something seems off but it requires encyclopedic knowledge that the average person is unlikely to have.
A common form of this is a press release that suggests a prototype process can scale up to solve some planetary problem. In many cases you can quickly estimate that planetary scale would require some part of the upstream inputs to be orders of magnitude larger than exists or is feasible. The media doesn't notice this part and runs with the "save the planet" story.
This is the industrial chemistry version of the "in mice" press releases in medicine. It is an analogue to the Gell-Mann amnesia effect.
With a higher voltage you can reduce your copper needs by a substantial amount. Seems if copper cost was a concern this would be what these data centers would do.
Agreed. I was kind of surprised to see 54 VDC mentioned. I am assuming this is low enough to meet some threshold for some kind of safety regulation. In other words, it doesn't shock you just 220 VAC would. I'm not entirely convinced of that however as it turns out bus bars are really dangerous in general. A 54 VDC bus bar won't shock you, but if you drop even a paperclip between the bus bar and a metal part that is grounded it basically disappears instantly in a small blast of plasma. The injury from that can be far worse than any shock you'd receive.
My experience has been that SELV (safety extra low voltage, less than 60V peak and less than 240VA) is considered safe and anything exceeding that needs certain levels of protection.
But bus bars generally should be protected regardless of voltage as they carry currents from high current capacity sources so even a lower voltage can be a safety concern.
Many server power supplies can take AC or DC input, with the DC input in the 300-500V range as this is comparable to the boost voltage for the AC power factor correction circuit. I just assumed most data centers using DC would be distributing around 400V within each rack.
Checking the arithmetic in every paper published seems like an good use case for LLMs. Has someone built a better version than uploading a PDF to ChatGPT and asking it to check the arithmetic?
Modern reasoning models are actually pretty good at arithmetic and almost certainly would have caught this error if asked.
Source: we benchmark this sort of stuff at my company and for the past year or so frontier models with a modest reasoning budget typically succeed at arithmetic problems (except for multiplication/division problems with many decimal places, which this isn't).
ChatGPT 5.2 has recently been churning through unsolved Erdös problems.
I think right now one is partially validated by a pro and the other one I know of is "ai-solved" but not verified. As in: we're the ones who can't quite keep up.
You can feed it the Hodge Conjecture for all I care,
the current algorithms are a joke and without real breakthroughs your just generating left to right text with billions in hardware.
Yes, yes. We’ve all seen the same screenshots. Very funny.
Those of us who don’t base our technical understandings on memes are well aware of the tooling at the disposal of all modern reasoning models gives them the capability to do such things.
The funniest part is beyond the typo, the complete lack of physical intuition from the analysts who circulated this. 500,000 tons is roughly the weight of 1.5 Empire State buildings. If your rack busbars weigh more than the structural steel of the facility housing them, you have a geotechnical engineering crisis on your hands. It is wild that we reached a point where financial modeling is so decoupled from physical reality that nobody paused to ask if the floor would collapse.
When I asked gemini it got it right away and as a source it provided a link to your comment... what a time to be alive, I hope you got the Empire State buildings calculations right because you're now part of the Truth ™
https://imgur.com/a/NaTfvtS
The irony is that I feel like everyone here is just copping off of WolframAlpha, which for the past 15 years has featured the ability to display references of similar magnitude for any query you give. And in fact, typing in "500,000 tons" into WolframAlpha tells you specifically that this is 1.4x the mass of the Empire State Building (and also 0.8x to 1.4x the mass of an Ultra Large Crude Carrier oil supertanker, and also 0.11x the mass converted to energy by the sun in one second).
https://www.wolframalpha.com/input?i=500%2C000+tons
The world of financial analysis and modeling is broad. It’s common to give these tasks to juniors and expect them to grind through it when the output doesn’t really matter.
In this case the output wasn’t actually used for financial modeling. If it had been, it would have been caught immediately when someone put it into a table where they calculated the price or the supply constraints or anything else.
Just to make it even more real: During covid I added a sub-panel and the wire (more like the sausage given the girth) between the sub-panel and main panel was aluminum because of cost. You just need to be a tad careful at the connection points with copper -- nothing a caring literate person can't handle
The incoming drop for a service panel is almost always aluminum (at least the residential stuff I've seen). It's cheaper and lighter than copper while still being ductile enough to work with. Just use the proper antioxidation paste at the connection point.
Did you use antioxidant paste on all of the aluminum terminations? If you didn’t, please pay an electrician to come fix it. If you did apply the paste, you are in the 0.01% of non-idiot non-electrician homeowners and can skip my rant about homeowners doing electrical work in the next paragraph. I sell electrical work and have seen some incredibly poor work done by homeowners.
Do not install aluminum wire if you are a homeowner unless you already know enough to use antioxidant paste and also to use a torque screwdriver or torque wrench for terminations and know where to find the torque values for the wiring devices you are using. If you don’t apply the paste, the surface of the aluminum will oxidize and could catch fire due to increased resistance.
I suggest not even touching copper as a homeowner, but it’s your choice and your house.
My first thought was, this is the kind of thing that an LLM writes and nobody checks. But then I realized, any decent LLM would have probably caught that inconsistency
Yeah, the good ones will use tools and give you a nice markdown table. Any claims like this anymore, I trust AI to go through it and give the numbers a reality check. With all the claims made about datacenter resources, power, and water usage, it's darkly funny how bad people are at understanding these things.
The math they do with their assumptions is usually pretty good, and you can tell when they put in the effort, but wow, the models and assumptions are all over the place.
2008 would like a word.
Kinda like when someone said that instead of running for mayor, Bloomberg could have given everyone in the country $1M. A guest said it on NBC and Bryan Williams, nor the pundit had any intuition that it seemed grossly wrong. https://www.politifact.com/factchecks/2020/mar/06/msnbc/bad-...
The system incentivizes ‘sounds good’ over ‘is correct’.
If news agencies needed to pay out fines for false info, you better believe they’d be checking more.
I wouldn't be surprised if that part was not really reviewed by an expert. They have the unit mass correct but maybe an editor is like ok but what does this look like for a gw project? It doesn't take more than 3rd grade math and a pocket calculator to do it correctly but journalist hasn't had to fumble that ball before. An expert knows its all too easy for any person to make that mistake and would second guess their own work.
With regards to the copper market: it keeps surprising me that some people seem to assume copper is a hard requirement for conducting electricity.
In reality copper is just convenient. We use it because it's easy to work with, a great conductor, and (until recently) quite affordable. But for most applications there's no reason we couldn't use something else!
For example, a 1.5mm2 copper conductor is 0.0134kg/m, which at current prices is $0.17 / meter. A 2.4mm2 aluminum conductor has the same resistance, weighs 0.0065kg/m, which at current prices is $0.0195 / meter!
Sure, aluminum is a pain to work with, but with a price premium like that there's a massive incentive to find a way to make it work.
Copper can't get too expensive simply due to power demands because people will just switch to aluminum. The power grid itself had been using it for decades, after all - some internal datacenter busbars should be doable as well.
You just have to not mix and match. If you mix them the two need special bonding at every single connector or they can cause arching and fires.
If by “special” you mean ordinary low-cost lugs, then sure.
Check it out:
https://lugsdirect.com/WhyAluminumOverCopperFAQ.htm
And you can browse that site for lugs, and they’re mostly rated for aluminum and copper. Copper-only lugs are actually rather unusual.
Sure, you can’t stick copper and aluminum wires into a wire nut, and finding terminations for smaller-gauge aluminum wire can be hard. But for larger wire, it’s really no problem.
Most wiring device terminations and lugs are Al/Cu rated. You’re supposed to use specific fittings for splicing together aluminum and copper wire (specific WAGO lever nuts are UL listed for this purpose along with some others), I assume that’s what you meant.
One good reason for moving from copper to aluminum is that copper contaminates recycled steel, but aluminum doesn't. Fortunately these days it's profitable to carefully extract most of the copper wire and such from scrap steel before it's melted down.
Most residential 200A electrical panels use tin plated aluminum busbars. https://www.eaton.com/cr/en-us/catalog/power-connections/bus...
I am not an electricity/wiring guy so maybe you can help me understand. I thought aluminum is dangerous to wire with because it is a fire hazard (I bought a home this year and this was a prominent warning in my reading). Is that because it needs to be done very carefully? I imagine most data centers would not mess with a fire risk on such a scale.
That's the "a pain to work with" part.
Residential aluminum is a Really Bad Idea because DIY Dave will inevitably do something wrong - which then leads to a fire hazard. Copper is a lot more forgiving.
But a large scale datacenter, solar farm, or battery storage installation? Those will be installed and maintained by trained electricians, which means they actually know what a "torque wrench" is, and how to deal with scary words like "corrosion" and "oxidation".
Like I said: it's what's used for most of the power grid. With the right training it really isn't a big deal.
Aluminum got a bad rep due to a lot of poor installations made in the 1960 and 70'ies. Subsequently new alloys (AA-8000 series) for wiring, new installation and termination procedures etc. have been developed, and AFAIU the situation today is much better. It's still trickier than copper wiring, so probably not a good option for DIY, but for some industrial installation with competent personnel and where the savings can be substantial it could certainly be an attractive option.
https://en.wikipedia.org/wiki/Aluminum_building_wiring
Aluminum oxide has high resistance and if you mix aluminum wiring with copper outlets, etc the impedance mismatched is what causes fires. You need to either have special copper pigtails installed or use fixtures that are rated for aluminum wiring.
For commercial installs, it shouldn't be a problem as long as it's planned for.
It's because aluminium has a higher coefficient of thermal expansion. It expands and shrinks more as it heats, and as those cycles add up it tends to loosen electrical connections. Loose connections have higher resistance, heat up and can cause fires.
That said, there is no reason we can't design better connectors that can withstand the expansion and shrinkage cycles, like spring loaded or spring cage connectors.
Old aluminum wiring in your walls with cloth insulators, designed for a time where electricity consumption was a small fraction of today's electrified usage is dangerous because you're overloading an old, unprepared system.
Aluminum bus bars(solid, often exposed) would be designed for the required power levels and installation criteria.
Aluminum home wiring was from the 60s and 70s. It’s not the same as cloth covered knob and tube from earlier years. It has its own problems, but I’d take a house with knob and tube over a house with aluminum wiring.
Household electrical usage has decreased, not increased, since that time. Your scaling is backwards.
Old aluminum wires in your walls were designed for a time when you lit your home with 100 watt incandescent lamps rather than 12 watt LEDs.
Only true for lighting circuits though, and most household circuits are mixed.
The quantity and (edit: aggregate) power draw of modern appliances is far greater now than 60 years ago, so the overall load on the old wires is much higher.
> power draw of modern appliances
Here's a article that claims that refrigerator energy efficiency has improved dramatically from 1972 to 2012.
https://appliance-standards.org/blog/how-your-refrigerator-h...
I'd bet that modern TVs are more efficient that CRT televisions. Do most people even have desktop computers anymore, or have they mostly been replaced by laptops, tables, and phones? I'd be interested to see the efficiency numbers for electric clothes dryers over time. I wouldn't be surprised if they are also slightly more efficient than older models, even if they are still using resistance heating. Due to smarter electronics that automatically turn the unit off after the clothes are dry (air humidity sensor). I think electric ranges, dish washers, toasters and coffee machines have been ubiquitous since the 1960s (but are probably about the same energy-consumption wise). Air conditioning units are one thing that I'd believe are much more common today than in the 1970s and 1980s. Household sizes are also smaller, so less electricity used for electric water heaters, and the oven, etc.. Electric vehicles are an up and coming user of electricity. What other appliances are likely to be using more now than before?
These are good points, but having worked on a few older houses, I usually see overextended and overloaded circuits, not the opposite.
Standard small-house service used to be 60A, sometimes as few as 4 circuits! It's now 100A minimum by code, with 200A common.
Ovens/ranges have gone from 30A to 50A (dedicated) circuits by code. Microwaves also require dedicated circuits now. Gaming computers with big GPUs are common. Air fryers and electric pressure cookers are newly-common countertop appliances. People definitely use resistive electric space heaters more now (very cheap, much safer than the older options). And there's a trend away from gas and to electric ranges and water heaters. Heat pumps are also increasingly common. You mentioned air conditioners and EV chargers. Kitchens and bathrooms are now required to have dedicated (and GFCI) circuits. Household sizes are smaller, but houses are larger.
So I guess I'd say that, properly expanded, individual circuits should carry less current than they used to. But very often, appliances (AC, microwave, gaming rig, air fryers), are just "plugged in" to an unexpanded system, with varying results.
If you're lucky, they pop a breaker and you call an electrician. If you're not lucky, they push the power draw into uncomfortable zones, esp for Al wire.
>Gaming computers with big GPUs are common.
Is there a good way to quantify this?
My best guess would be the size of the GPU market, before the AI crush.
I don't track that market (not a gamer), but it seemed substantial enough that I was aware of it. I do travel in geeky circles though.
Also, plasma TVs were big for a decade or so, and they run hotter than the CRTs that preceded them, or the LCDs/etc of today.
This is absolutely not true in areas where heating the air and water and cooking are done with natural gas. Every single appliance in a house is more efficient today than in 1970 due to advances in motor speed control, without exception. The only thing that didn’t get more efficient is electric resistive heat and it’s impossible to improve on that anyways.
I can’t think of a single appliance from 1970 that consumes less energy than its modern equivalent. Anything with a pump or fan is more efficient and so is lighting. LCD TVs use less energy than CRTs.
I also can’t think of an appliance that has become common in households that draws more than 100 watts of continuous load since the 1970 aside from just ‘computers’. An ancient 500W 80% efficiency PSU at max load only has 5.2A of current at 120V single-phase.
If you convert your natural gas furnace to a heat pump, you will use more electricity but excluding that and NG to electric HPWHs leaves only more efficient equipment.
Sure, but there are more appliances plugged in today than there were. The simplest evidence for this is that there are never enough outlets in an old (unrenovated) home.
In a renovated house, you won't have aluminum wire at all, so these concerns are null.
My original statement should be qualified. Since we were talking about aluminum wire it's relevant -- an updated house will have new (copper) circuits that can handle all this stuff. An NON updated house might have Al wire and be overloaded in a more severe way than it was in the 60s.
But FWIW, new >100W appliances:
Unless I'm mistaken, the risk with aluminum is that it can expand and contract if it gets too hot. Aluminum sized properly with the correct connectors torqued to spec would be fine, aluminum wires in a residence with a DIYer working on it can be riskier and is why inspectors will always note it.
Aluminum wires became brittle over time(tens of years), fluid which requires some maintenance for screw terminals and inducts galvanic corrosion when coupled with copper without special care. If wiring was properly done and maintained, it is okayish.
Old aluminum is dangerous. Modern aa-8000 series aluminum is fine. Bigger pipes though.
If you apply antioxidant paste, torque the terminations to the correct value, and use wiring devices rated for Cu and Al conductors, there are zero problems.
Most homeowners know none of the requirements of aluminum wire installation so I don’t recommend using it, oxidation is bad and can cause fires. Just pay for copper NM cable, it’s dirt cheap anyways.
FWIW I sell and run electrical work for a living. I assume crote and amluto work in the electrical industry, since they’re the only other posters in this discussion who know what they’re talking about.
It’s a fire hazard in residential houses where people frequently do their own wiring, because it needs more expertise to wire correctly. Copper wiring is a lot more forgiving to being hooked up by amateurs.
The biggest reason is that aluminum oxidizes, and unlike copper, the oxide layer has high resistivity. In theory that shouldn’t be an issue in datacenters hiring expert technicians.
Yea, in a data center setup the cost difference can justify the things like soldiering connections rather than mechanical connectors with a higher resistance.
> In the Earth's crust, aluminium is the most abundant metallic element (8.23% by mass[68]) and the third most abundant of all elements (after oxygen and silicon).
TIL. I thought it would be relatively expensive due to the difficulty of extracting it.
(Iron is much cheaper than I thought, too.)
It used to be extremely expensive!
The Washington Monument has an aluminum cap, which at the time was as expensive as silver. Two years later the Hall-Héroult process was invented, and as a result the price plummeted.
Aluminium essentially reflects electricity prices and increasingly recycled
AFAIK Aluminum wires will become a heat liability, especially at higher amps
That's why my calculation example used a 1.5mm2 copper wire but a 2.4mm2 aluminum one.
Aluminum has a higher resistance, which means the same diameter will get hotter than copper. Make the cable thicker and its resistance drops, which means it gets less hot.
Want more amps at the same temperature? Ohm's law still applies: just use a thicker cable.
Aluminum has lower resistance per kg, or per $. This is why transmission lines use aluminum (around a steel core).
Any wire---of any non superconducting material---will be hotter at higher current flows. You size the wire to the application.
It depends on design, environment and maintenance. If it's well protected against oxidation and it's static, it's not really a problem. Heat issues are often at the joints, which are most vulnerable, but can be coated and encapsulated to mitigate. For some uses, aluminium is superior to copper all things considered. Obviously the diameter of the wire/cable depends on which conductor you use, AC/DC, current and service temp - for some applications, that may favour one material over the other.
If the two wires are the same gauge, yes. If you size up the aluminum, at the same resistance/current would mean the same amount of power over the length of the conductor and same heat.
Thicker cables or higher voltage(lower current) is the answer which is why it's used in power distribution networks where they can control the voltage by planning what to transform to.
> In reality copper is just convenient. We use it because it's easy to work with, a great conductor, and (until recently) quite affordable
It's convenient, it's easy to work with, great conductivity, and cheap enough all at the sametime... Dude, I think you just explained why cropper is used instead of anything else.
> and cheap enough
Not anymore. :(
> Sure, aluminum is a pain to work with
Really? In larger sizes, an equivalent ampacity aluminum cable is generally lighter and more flexible than copper. The main downside is that it’s thicker.
(Common terminations for larger wire sizes are often dual-rated for aluminum and copper. The engineering details for how to design lugs that work well for aluminum and copper were worked out long ago.)
interesting, thanks!
Aluminum conductors are dangerous unless the entire system is designed for it. It is not a case of switching to something cheaper.
Look at the electrical fires of the 1950’s and 1960’s as an example, and that was at household levels of current.
Aluminum is used, but everything accounts for the insane coefficient of linear expansion and other annoying properties.
I would imagine most large-scale data center construction projects will include electrical engineers to design the electrical subsystem. A rack's floor footprint is a few square feet. You can put several million dollars of hardware into that rack. A data center will have at least a few racks. It's a very reasonable investment to bring someone in to do electrical design.
This isn't true at all unless you narrowly dwfine "the entire system".
Each feeder can be aluminum if you put special goop on any copper connections. Breakers accept it just fine, etc.
You should avoid it for smaller wiring, though. There's special 8000 series aluminum if you're trying to be serious with Al feeders
The goop is generally considered to be unnecessary. What you need is a termination that is rated for aluminum wire, and these are very, very common as long as the wire in question is fairly large (8AWG or so and larger).
I only put it in the comment since the commenters here seemed uninformed.
It's better to be safe and follow specs; it's more when you're doing some compound splice where it makes sense. I figure it'll make more people feel easy about it if there's a preventative measure
Amazingly ironic that in an article about a typo, there is a typo.
> "Tat sounds like the ultimate catalyst for the commodities market and copper has been hitting records."
"Tat" should be "That", imo.
you should've made a typo as well, as per tradition
I interpreted the typo as a meta commentary on the state of journalism in the 21st century.
I saw this too, awesome to report on a typo when you yourself have a typo. What are the odds?
Muphry's Law
https://en.wikipedia.org/wiki/Muphry%27s_law
but one weighs half a million ton ¯\_(ツ)_/¯
This sort of mistake is easy to make when you're mixing up your units; if they kept to one system of measure, it would've been trivial to catch, before or after release.
We need to standardize on using Earth circumferences as the unit of length. Or better, football fields! (the type of football of course being implied by the website's ccTLD)
We _have_ standardized on Earth circumferences for length, only we divide by 40 million to make the numbers more sane, and got the measurement slightly wrong!
How hard would it be to fix this? Could we theoretically add or subtract enough material or make whole thing slightly more dense or less dense to compensate?
Per Wikipedia, the discrepancy is approximately 74 km, so digging a ditch with an average depth of approximately 74/2π ≅ 12 km around the circumference of the Earth would theoretically fix the problem.
Feasibility and geological implications are left as exercises for the reader.
Regardless, I suspect a more cost-effective fix would be to redefine the meter to be a couple "legacy" millimeters longer.
We should just redefine it to make the speed of light a nice even 300.000 km/s - we are so close already!
Why stop there when we could just as easily redefine it to be 1 (new base unit of length)/second or 1 meter/(new base unit of time)?
But it's already a nice round 10m/s in base 299,792,458.
If only there were some kind of international system of standard units.
Olympic swimming pools for liquids, times around the the earth for length and number of double decker busses for height.
Don't forget packs of cigarettes as a more convenient unit for measuring volumes significantly smaller than Olympic-sized pools.
There is, of course, no more need to standardize on a specific brand or style of cigarette than on a specific depth of Olympic-sized pool.
You jest, but times around the Earth is the actual origin of the Meter. Kinda.
The history is quite interesting and well worth checking out.
I can't recommend a book on the subject, but I do heartily recommend "Longitude", which is about the challenges of inventing the first maritime chronometers for the purpose of accurately measuring longitude.
The original meter (1790s France) was defined as 1/10,000,000 of the distance from the equator to the North Pole along a meridian.
Not sure if you're correcting me, but yes, that is "a" path around the Earth.
It's not the most aesthetic one, but it was at the time the most able to be measured.
For smaller lengths and radiation bananas are also acceptable.
A good physicist can calculate banana equivalent dose other head. Always ask for it when dealing with radiation.
Don’t forget cheetahs for velocity and elephants for weight.
A horse as a measure of power and a crocodile bite as a measure of compression strength
I thought all measurements in data centers were in US football fields.
For the floor area or length/width it is, but if you want the height then that's in Empire State Buildings.
There are standard units, yes.
It’s interestingly related that Amazon is also buying a copper mine: https://www.wsj.com/finance/commodities-futures/amazon-is-bu...
They don't buy a mine they buy the output.
The error in the original article has been corrected.
https://developer.nvidia.com/blog/nvidia-800-v-hvdc-architec...
> If the "half a million tons" figure were accurate, a single 1 GW data center would consume 1.7% of the world's annual copper supply. If we built 30 GW of capacity—a reasonable projection for the AI build-out—that sector alone would theoretically absorb almost half of all the copper mined on Earth.
Quickly doing such "back of an envelope" calculations, and calling out things that seem outlandish, could be a useful function of an AI assistant.
Using your brain is so vastly more energy efficient, we might just only need half of that 30 GW capacity if fewer people had these leftpad-style knee-jerk reactions.
While I don't know how more or less efficient it is, WolframAlpha works well for these sorts of approximations, and shows its work more clearly than the AI chatbots I've used.
A Gemini query uses about a kilojoule. The brain runs at 20 W (though the whole human costs 100 W). So, the human is less energy if you can get it done in under 50 seconds.
Where does that number come from? Does it factor in the energy required to build the servers used to train the model? Does it factor in… the training?
There is no end to this energy comparison...
For example does it factor in the 18-24 years needed to train a human and the energy used for that?
Hopefully we aren’t doing too much AI training to work out 200 * 1000. If a computer is involved at all it’s disappointing, if AI is used, more so.
It doesn't matter what the model is actually doing at the end of the day, when training and hosting it involves massive amounts of energy.
If humans aren’t more efficient the energy is still used, as they remain alive. Maybe the AI will notice this?
Each person uses about 100W (2000kcal/24h=96W). Running all of humanity takes about 775GW.
Sure, using or not using your brain is a negligible energy difference, so if you aren't using it you really should, for energy efficiency's sake. But I don't think the claim that our brains are more energy efficient is obviously true on its own. The issue is more about induced demand from having all this external "thinking" capacity on your fingertips
Is there an AI system with functionality at or equal to a human brain that operates on less than 100W? Its currently the most efficient model we have. You compare all of humanity's energy expenditure, but to make the comparison, you need to consider the cost of replicating all that compute with AI (assuming we had an AGI at human level in all regards, or a set of AIs that when operated together could replace all human intelligence).
>all human intelligence
So, this is rather complex because you can turn AI energy usage to nearly zero when not in use. Humans have this problem of needing to consume a large amount of resources for 18-24 years with very little useful output during that time, and have to be kept running 24/7 otherwise you lose your investment. And even then there is a lot of risk they are going to be gibbering idiots and represent a net loss of your resource expenditure.
For this I have a modern Modest Proposal they we use young children as feed stock for biofuel generation before they become a resource sink. Not only do you save the child from a life of being a wage slave, you can now power your AI data center. I propose we call this the Matrix Efficiency Saving System (MESS).
>Is there an AI system with functionality at or equal to a human brain that operates on less than 100W?
Obviously not equal to a human brain, but my GPU takes about 150W and can draw an image in a minute that would take me forever to replicate.
No one will ever agree on when AI systems have equivalent functionality to a human brain. But lots of jobs consist of things a computer can now do for less than 100W.
Also, while a body itself uses only 100W, a normal urban lifestyle uses a few thousand watts for heat, light, cooking, and transportation.
> Also, while a body itself uses only 100W, a normal urban lifestyle uses a few thousand watts for heat, light, cooking, and transportation.
Add to that the tier-n dependencies this urban lifestyle has—massive supply chains sprawling across the planet, for example involving thousands upon thousands of people and goods involved in making your morning coffee happen.
Wikipedia quoted global primary energy production at 19.6 TW, or about 2400W/person. Which is obviously not even close to equally distributed. Per-country it gets complicated quickly, but naively taking the total from [1] brings the US to 9kW per person.
And that's ignoring sources like food from agriculture, including the food we feed our food.
To be fair, AI servers also use a lot more energy than their raw power demand if we use the same metrics. But after accounting for everything, an American and an 8xH100 server might end up in about the same ballpark
Which is not meant as an argument for replacing Americans with AI servers, but it puts AI power demand into context
https://www.eia.gov/energyexplained/us-energy-facts/
Obviously we don't have AGI so we can't compare many tasks. But on tasks where AI does perform at comparable levels (certain subsets of writing, greenfield coding and art) it performs fairly well. They use more power but are also much faster, and that about cancels out. There are plenty of studies that try to put numbers on the exact tradeoff, usually focused more on CO2. Plenty that find AI better by some absurd degree (800 times more efficient at 3d modelling, 130 to 1500 times more efficient at writing, or 300 to 3000 times more efficient at illustrating [1]). The one I'd trust the most is [2] where GPT4 was 5-19 times less CO2 efficient than humans at solving coding challenges
1: https://www.nature.com/articles/s41598-024-54271-x?fromPaywa...
2: https://www.nature.com/articles/s41598-025-24658-5
I did some math for this particular case by asking Google’s Gemini Pro 3 (via AI studio) to evaluate the press release. Nvidia has since edited the release to remove the “tons of copper” claim, but it evaluated the other numbers at a reported API cost of about 3.8 cents. If the stated pricing just recovers energy cost, that implies 1500kJ of energy as a maximum (less if other costs are recovered in the pricing). A human thinking for 10 minutes would use sbout 6kJ of direct energy.
I agree with your point about induced demand. The “win” wouldn’t be looking at a single press release with already-suspect numbers, but rather looking at essentially all press releases of note, a task not generally valuable enough to devote people towards.
That being said, we normally consider it progress when we can use mechanical or electrical energy to replace or augment human work.
I think this is exactly the thing that should be done by a person without AI, to check what AI is writing.
That can be true while also seeing value in using an AI to sanity check human-generated claims.
Nah just spin up an agentic ai proofreading agent on the cloud.
Nobody on HN is a bigger AI stan than I am -- well, maybe that SimonW guy, I guess -- but the truth is that problems involving unit conversions are among the riskiest things you can ask an LLM to handle for you.
It's not hard to imagine why, as the embedding vectors for terms like pounds/kilograms and feet/yards/meters are not going to be far from each other. Extreme caution is called for.
That sounds like the sort of thing I'd expect them to be good at. What goes wrong?
I edited the post with a speculation, but it's just a guess, really. In the training data, different units are going to share near-identical grammatical roles and positions in sentences. Unless some care is taken to force the embedding vectors for units like "pounds" and "kilograms" to point in different directions, their tokens may end up being sampled more or less interchangeably.
Gas-law calculations were where I first encountered this bit of scariness. It was quite a while ago, and I imagine the behavior has been RLHF'ed or otherwise tweaked to be less of a problem by now. Still, worth watching out for.
> In the training data, different units are going to share near-identical grammatical roles and positions in sentences.
Yes, but I would also expect the training data to include tons of examples of students doing unit-conversion homework, resources explaining the concept, etc. (So I would expect the embedding space to naturally include dimensions that represent some kind of metric-system-ness, because of data talking about the metric system.) And I understand the LLMs can somehow do arithmetic reasonably well (though it matters for some reason how big the numbers are, so presumably the internal logic is rather different from textbook algorithms), even without tool use.
It’s one of the useful functions of an engineer.
It's almost always the engineers, analysts and MBA spreadsheet pushers and other people removed from the physical consequences outputting these mistakes because it's way easier to not notice a misplaced decimal or incorrect value when you deal in pure numbers and know what they "should" be than you are the person actually figuring out how to make it happen the difference between needing 26666666.667 and 266666666.667 <units> of <widget> is pretty meaningful. Engineers don't output these mistakes as often as analysts or whatever because they work in organizations that invest more in catching them, not because they make them all that much less.
Whether talking weight or bulk a decimal place is approximately the difference between needing a wheelbarrow, a truck, a semi truck, a freight train and a ship.
Around here, asking "does this number make sense?" when coming across a figure is second nature, reinforced since early in engineering school. The couple of engineers from the US that I know behave similarly, which makes sense because when your job is to solve practical problems and design stuff, precision matters.
> difference between needing 26666666.667 and 266666666.667 <units> of <widget> is pretty meaningful
To be fair, that’s why we’d use 2.6666666667e7 and 2.66666666667e8, which makes it easier to think about orders of magnitude. Processes, tools and methods must be adapted to reduce the risk of making a mistake.
>why we’d use 2.6666666667e7 and 2.66666666667e8
Engineers prefer (wait for it) "Engineering Notation": 26.67e6 and 266.7e6.
https://en.wikipedia.org/wiki/Engineering_notation
These kinds of errors are not rare in the media.
Even among engineering fields routine handling of diverse and messy unit systems (e.g. chemical engineering) are relatively uncommon. If you work in one of these domains, there is a practiced discipline to detect unit conversion mistakes. You can do it in your head well enough to notice when something seems off but it requires encyclopedic knowledge that the average person is unlikely to have.
A common form of this is a press release that suggests a prototype process can scale up to solve some planetary problem. In many cases you can quickly estimate that planetary scale would require some part of the upstream inputs to be orders of magnitude larger than exists or is feasible. The media doesn't notice this part and runs with the "save the planet" story.
This is the industrial chemistry version of the "in mice" press releases in medicine. It is an analogue to the Gell-Mann amnesia effect.
With a higher voltage you can reduce your copper needs by a substantial amount. Seems if copper cost was a concern this would be what these data centers would do.
Agreed. I was kind of surprised to see 54 VDC mentioned. I am assuming this is low enough to meet some threshold for some kind of safety regulation. In other words, it doesn't shock you just 220 VAC would. I'm not entirely convinced of that however as it turns out bus bars are really dangerous in general. A 54 VDC bus bar won't shock you, but if you drop even a paperclip between the bus bar and a metal part that is grounded it basically disappears instantly in a small blast of plasma. The injury from that can be far worse than any shock you'd receive.
My experience has been that SELV (safety extra low voltage, less than 60V peak and less than 240VA) is considered safe and anything exceeding that needs certain levels of protection.
But bus bars generally should be protected regardless of voltage as they carry currents from high current capacity sources so even a lower voltage can be a safety concern.
Many server power supplies can take AC or DC input, with the DC input in the 300-500V range as this is comparable to the boost voltage for the AC power factor correction circuit. I just assumed most data centers using DC would be distributing around 400V within each rack.
I wonder how many 'innocent' inaccurate data is taken by journalists to use for their agenda. I think this happens a lot in the age of AI.
Efficient markets!
Checking the arithmetic in every paper published seems like an good use case for LLMs. Has someone built a better version than uploading a PDF to ChatGPT and asking it to check the arithmetic?
LLM's are why we're in this mess, they can't do math or count r's
Modern reasoning models are actually pretty good at arithmetic and almost certainly would have caught this error if asked.
Source: we benchmark this sort of stuff at my company and for the past year or so frontier models with a modest reasoning budget typically succeed at arithmetic problems (except for multiplication/division problems with many decimal places, which this isn't).
Interesting, how have you found they have been performing at more complex things like calculus and analysis?
It’s on the front page of HN once in a while.
They can't do math?
ChatGPT 5.2 has recently been churning through unsolved Erdös problems.
I think right now one is partially validated by a pro and the other one I know of is "ai-solved" but not verified. As in: we're the ones who can't quite keep up.
https://arxiv.org/abs/2601.07421
And the only reason they can't count Rs is that we don't show them Rs due to a performance optimization.
You can feed it the Hodge Conjecture for all I care, the current algorithms are a joke and without real breakthroughs your just generating left to right text with billions in hardware.
Guess frontier math and programming are just left to right text then.
An LLM usually has a powerful digital computer right in its disposal, and could use it as a tool to do precise calculations.
More accurate to say they can’t see r’s. They process language but not letters.
Yes, yes. We’ve all seen the same screenshots. Very funny.
Those of us who don’t base our technical understandings on memes are well aware of the tooling at the disposal of all modern reasoning models gives them the capability to do such things.
Please don’t bring the culture war here.