Edit: Just to clarify, I don't imply that this is public money to be spent. It will commission $500B worth of human and material resources for 5 years that can be much more productive if used for something else - i.e. high speed rail network instead of a machine that Chinese built for $5B.
If you want to compare the DeepSeek-R development costs to anything, you should be comparing it to what it cost OpenAI to develop GPT-o1 (not what they plan to spend to run it), but both numbers are somewhat irrelevant since they both build upon prior research.
Perhaps what's more relevant is that DeepSeek are not only open sourcing DeepSeek-R1, but have described in a fair bit of detail how they trained it, and how it's possible to use data generated by such a model to fine-tune a much smaller model (without needing RL) to much improve it's "reasoning" performance.
This is all raising the bar on the performance you can get for free, or run locally, which reduces what companies like OpenAI can charge for it.
The existing hyperscalers have already sunk ungodly amounts of money into literally hundreds of new data centers, millions of GPUs to fill them, chip manufacturing facilities, and even power plants with the impression that, due to the amount of compute required to train and run these models, there would be demand for these things that would pay for that investment. Literally hundreds of billions of dollars spent already on hardware that’s already half (or fully) built, and isn’t easily repurposed.
If all of the expected demand on that stuff completely falls through because it turns out the same model training can be done on a fraction of the compute power, we could be looking at a massive bubble pop.
Efficiency going up tends to increase demand by much more than the efficiency-induced supply increase.
Assuming that the world is hungry for as much AI as it can get. Which I think is true, we're nowhere near the peak of leveraging AI. We barely got started.
That's what's baffling with Deepseek's results: they spent very little on training (at least that's what they claim). If true, then it's a complete paradigm shift.
And even if it's false, the more wide AI usage is, the bigger the share of inference will be, and inference cost will be the main cost driver at some point anyway.
No, this is the change introduced by o1, what's different with R1 is that its use of RL is fundamentally different (and cheaper) that what OpenAI did.
It's just data centers full of devices optimized for fast linear algebra, right? These are extremely repurposeable.
The hardware can train LLM but also be used for vision, digital twin, signal detection, autonomous agents, etc.
Military uses seem important too.
Can the large GPU based data centers not be repurposed to that?
They aren't comparing the 500B investment to the cost of deepseek-R1 (allegedly 5 millions) they are comparing the cost of R1 to the one of o1 and extrapolating from that (we don't know exactly how much OpenAI spent to train it, but estimates put it around $100M, in which case deepseek would have been only 95% more cost-efficient, not 99%)
If new technology means we can get more for a dollar spent, then $500 billion gets more, not less.
The money is not spent. Deepseek published their methodology, incumbents can pivot and build on it. No one knows what the optimal path is, but we know it will cost more.
I can assure you that OpenAI won't continue to produce inferior models at 100x the cost.
What happens if that money is being actually spent, then some people constantly catch up but don't reveal that they are doing it for cheap? You think that it's a competition but what actually happening is that you bleed out of your resources at some point you can't continue but they can.
Like the star wars project that bankrupted the soviets.
Wasn't that a G.W Bush Jr thing?
- The hardware purchased for this initiate can be used for multiple architectures and new models. If DeepSeek means models are 100x as powerful, they will benefit
- Abstraction means one layer is protected from direct dependency on implementation details of another layer
- It’s normal to raise an investment fund without knowing how the top layers will play out
Hope that helps? If you can be more specific about your confusion I can be more specific in answering.
Then the Open Source world came out of the left and b*tch slapped all those head honchos and now its like this.
For tech like LLMs, it feels irresponsible to say 500 billion $$ investment and then place that into R&D. What if in 2026, we realize we can create it for 2 billion$, and let the 498 billion $ sitting in a few consumers.
It may still be flawed or misguided or whatever, but it’s not THAT bad.
It's such a weird question. You made it sound like 1) the $500B is already spent and wasted. 2) infrastructure can't be repurposed.
That compute can go to many things.
You want to invest $500B to a high speed rail network which the Chinese could build for $50B?
https://www.businessinsider.com/french-california-high-speed...
Doubly delicious since the French have a long and not very nice colonial history in North Africa, sowing long-lasting suspicion and grudges, and still found it easier to operate there.
The problem is loose vs strong property rights.
We don't have the political will in the US to use eminent domain like we did to build the interstates. High speed rail ultimately needs a straight path but if you can't make property acquisitions to build the straight rail path then this is all a non-starter in the US.
Edit: asked Deepseek about it. I was kinda spot on =)
Cost Breakdown
Solar Panels $13.4–20.1 trillion (13,400 GW × $1–1.5M/GW)
Battery Storage $16–24 trillion (80 TWh × $200–300/kWh)
Grid/Transmission $1–2 trillion
Land, Installation, Misc. $1–3 trillion
Total $30–50 trillion
The most common idea is to spend 3-5% of GDP per year for the transition (750-1250 bn USD per year for the US) over the next 30 years. Certainly a significant sum, but also not too much to shoulder.
I guess the power plants are salvageable.
If your rich spend all their money on building pyramids you end up with pyramids instead of something else. They could have chosen to make irrigation systems and have a productive output that makes the whole society more prosperous. Either way the workers get their money, on the Pyramid option their money ends up buying much less food though.
It’s smart on their part.
https://fortune.com/2025/01/23/saudi-crown-prince-mbs-trump-...
Since the Stargate Initiative is a private sector deal, this may have been a perfect shakedown of Saudi Arabia. SA has always been irrationally attracted to "AI", so perhaps it was easy. I mean that part of the $600 billion will go to "AI".
And if you don’t want to look that far just lookup what his #1 donor Musk said…there is no actual $500Bn.
There was an amusing interview with MSFT CEO Satya Nadella at Davos where he was asked about this, and his response was "I don't know, but I know I'm good for my $80B [that I'm investing to expand Azure]".
Either that or its an excuse for everyone involved to inflate the prices.
Hopefully the datacenters are useful for other stuff as well. But also I saw a FT report that it's going to be exclusive to openai?
Also as I understand it these types of deals are usually all done with speculative assets. And many think the current AI investments are a bubble waiting to pop.
So it will still remain true that if jack falls down and breaks his crown, jill will be tumbling after.
1. Stargate is just another strategic deception like Star Wars. It aims to mislead China into diverting vast resources into an unattainable, low-return arms race, thereby hindering its ability to focus on other critical areas.
2. We must keep producing more and more GPUs. We must eat GPUs at breakfast, lunch, and dinner — otherwise, the bubble will burst, and the consequences will be unbearable.
3. Maybe it's just a good time to let the bubble burst. That's why Wall Street media only noticed DeepSeek-R1 but not V3/V2, and how medias ignored the LLM price war which has been raging in China throughout 2024.
If you dig into 10-Ks of MSFT and NVDA, it’s very likely the AI industry was already overcapacity even before Stargate. So in my opinion, I think #3 is the most likely.
Just some nonsense — don't take my words seriously.
Well, this is a private initiative, not a government one, so it seems not, and anyways trying to bankrupt China, whose GDP is about the same as that of the USA doesn't seem very achievable. The USSR was a much smaller economy, and less technologically advanced.
OpenAI appear to genuinely believe that there is going to be a massive market for what they have built, and with the Microsoft relationship cooling off are trying to line up new partners to bankroll the endeavor. It's really more "data center capacity expansion as has become usual" than some new strategic initiative. The hyperscalars are all investing heavily, and OpenAI are now having to do so themselves as well. The splashy Trump photo-op and announcement (for something they already started under Biden) is more about OpenAI manipulating the US government than manipulating China! They have got Trump to tear up Biden's AI safety order, and will no doubt have his help in removing all regulatory obstacles to building new data centers and the accompanying power station builds.
I was thinking the same thing...how much is that investment mostly grift?
1: https://www.chinatalk.media/p/deepseek-ceo-interview-with-ch...