The shift of mining power to AI: potential risks for the crypto market in 2026+

[DrBeer]

Plus, the decline in the investment attractiveness of cryptocurrency itself (investors are actively transferring money from cryptocurrencies to AI) will lead to the devaluation of cryptocurrency,
 and this causes a chain reaction—low prices in the long term—mining becomes unprofitable or loss-making due to the same very expensive electricity for mining companies.

Most crypto investments are over the big corps ETF.
Mining is not really a target for Investors, they after values and a few follow Saylor's lead and buy bitcoin.
AI is not in competition to Crypto, only the hype followers believe the AI hype priest like the Nvidia, X, Zuck the oracle guy.
All have a lot to loose if AI does not succeed. The are also delusional with their AI plus robotic approach and sending data centers into space. None is talking about the inconvenience of service. Robots also need services and the market is rather small to get a household slave unable to fold a shirt.
So in my eyes there is not a competition. Crypto is a fair space you can only buy a coin if someone is selling one.
Futures and options are a different ballgame.

Again, I couldn't convey my thought correctly in English

Let me give you a slightly different example, which I hope will be clearer: when you invest in gold, you are not investing in mines. You are definitely not investing in mines! That is a fact. But when you buy gold, as others repeat, the appeal of research and exploration, and then the construction of gold mines, increases.  This is also a FACT! The result is an indirect but quite obvious relationship: by buying a product and creating demand for it, you influence its production and growth.
Now let's imagine that the locations of gold deposits are being given away, for example, for construction... let's say for solar energy fields, because 1 dollar invested in gold mining will give you, very roughly, 10 dollars in profit over 10 years, while 1 dollar invested in solar energy will give you 1,000 dollars over the same period. The question is, where will there be more investment, and where will it come from? Especially if the key resource is land, and priority is given to gold miners? And as a result, what will develop and what will become less profitable and lose its appeal and demand in the market?

[gibrab16]

Plus, the decline in the investment attractiveness of cryptocurrency itself (investors are actively transferring money from cryptocurrencies to AI) will lead to the devaluation of cryptocurrency,
 and this causes a chain reaction—low prices in the long term—mining becomes unprofitable or loss-making due to the same very expensive electricity for mining companies.

Most crypto investments are over the big corps ETF.
Mining is not really a target for Investors, they after values and a few follow Saylor's lead and buy bitcoin.
AI is not in competition to Crypto, only the hype followers believe the AI hype priest like the Nvidia, X, Zuck the oracle guy.
All have a lot to loose if AI does not succeed. The are also delusional with their AI plus robotic approach and sending data centers into space. None is talking about the inconvenience of service. Robots also need services and the market is rather small to get a household slave unable to fold a shirt.
So in my eyes there is not a competition. Crypto is a fair space you can only buy a coin if someone is selling one.
Futures and options are a different ballgame.

Again, I couldn't convey my thought correctly in English

Let me give you a slightly different example, which I hope will be clearer: when you invest in gold, you are not investing in mines. You are definitely not investing in mines! That is a fact. But when you buy gold, as others repeat, the appeal of research and exploration, and then the construction of gold mines, increases.  This is also a FACT! The result is an indirect but quite obvious relationship: by buying a product and creating demand for it, you influence its production and growth.
Now let's imagine that the locations of gold deposits are being given away, for example, for construction... let's say for solar energy fields, because 1 dollar invested in gold mining will give you, very roughly, 10 dollars in profit over 10 years, while 1 dollar invested in solar energy will give you 1,000 dollars over the same period. The question is, where will there be more investment, and where will it come from? Especially if the key resource is land, and priority is given to gold miners? And as a result, what will develop and what will become less profitable and lose its appeal and demand in the market?

When people buy gold, they’re not funding mines directly. No one buying a gold bar is thinking about excavators and drilling permits. But demand still sends a loud signal. High demand makes mining more attractive, exploration ramps up, investors step in, and suddenly more land and resources get locked into gold production. That’s the indirect link.

[bitgolden]

While GPU mining is still a thing, it's much smaller than what it used to be, because GPU's are going towards AI instead. SO we have asic miners, which has nothing to do with AI, and we have companies that use these ASIC miners to mine bitcoin, doubt really it would be that much of a overlap between asic mining and AI world because the only thing common between them is that they both use energy, but so does your fridge, doesn't destroy fridge market suddenly.

You can't just open up an AI company and have bunch of servers, so instead people do ASIC mining, which only requires a capital to start.

[DrBeer]

Plus, the decline in the investment attractiveness of cryptocurrency itself (investors are actively transferring money from cryptocurrencies to AI) will lead to the devaluation of cryptocurrency,
 and this causes a chain reaction—low prices in the long term—mining becomes unprofitable or loss-making due to the same very expensive electricity for mining companies.

Most crypto investments are over the big corps ETF.
Mining is not really a target for Investors, they after values and a few follow Saylor's lead and buy bitcoin.
AI is not in competition to Crypto, only the hype followers believe the AI hype priest like the Nvidia, X, Zuck the oracle guy.
All have a lot to loose if AI does not succeed. The are also delusional with their AI plus robotic approach and sending data centers into space. None is talking about the inconvenience of service. Robots also need services and the market is rather small to get a household slave unable to fold a shirt.
So in my eyes there is not a competition. Crypto is a fair space you can only buy a coin if someone is selling one.
Futures and options are a different ballgame.

Again, I couldn't convey my thought correctly in English

Let me give you a slightly different example, which I hope will be clearer: when you invest in gold, you are not investing in mines. You are definitely not investing in mines! That is a fact. But when you buy gold, as others repeat, the appeal of research and exploration, and then the construction of gold mines, increases.  This is also a FACT! The result is an indirect but quite obvious relationship: by buying a product and creating demand for it, you influence its production and growth.
Now let's imagine that the locations of gold deposits are being given away, for example, for construction... let's say for solar energy fields, because 1 dollar invested in gold mining will give you, very roughly, 10 dollars in profit over 10 years, while 1 dollar invested in solar energy will give you 1,000 dollars over the same period. The question is, where will there be more investment, and where will it come from? Especially if the key resource is land, and priority is given to gold miners? And as a result, what will develop and what will become less profitable and lose its appeal and demand in the market?

When people buy gold, they’re not funding mines directly. No one buying a gold bar is thinking about excavators and drilling permits. But demand still sends a loud signal. High demand makes mining more attractive, exploration ramps up, investors step in, and suddenly more land and resources get locked into gold production. That’s the indirect link.

Well, here I would argue that demand for a resource and its price increase lead to investments in the development of new deposits. But when it comes to mining and AI, the situation is slightly different—there is demand for cryptocurrencies, but their investment attractiveness has declined, as if the price of gold began to fall and demand also began to fall.
 However, another resource has appeared that requires land where there is potential for gold, let it be... let it be, for example, fields for solar stations, which (let's assume) generate significantly more profit than gold.
 Investors will look for land, and governments, for example, will easily sell land with potential gold deposits for giant solar power plant projects, as they guarantee higher profits at lower costs.
 This means that gold mining will lose resources for gold extraction. This is exactly what could potentially happen with mining resources—space, electricity, and, for example, water for cooling. When AI becomes less profitable, these resources may return to mining. But right now, there is a wild surge in demand for AI, and it is difficult to say how long it will last. I do not rule out that it will be virtually eternal.
 

[DrBeer]

Some news on the topic, with specific examples of changing priorities even among miners themselves:
Crypto mining companies are changing their business model and focusing on artificial intelligence. The reason is that Bitcoin mining is becoming more difficult and less profitable, while demand for AI computing power continues to grow rapidly.
Miners already have key assets that are particularly valuable today: large data centers, advanced cooling systems, land, and long-term electricity contracts. This is exactly the kind of infrastructure needed by tech giants such as Amazon, Microsoft, Alphabet, and Meta*, which are actively expanding their AI operations.
The transition to AI is not easy. Working with AI loads requires more complex cooling systems and network infrastructure. However, cooperation with former miners allows AI companies to launch new capacities faster and cheaper than building data centers from scratch.
Some companies are going even further. For example, Core Scientific has announced plans to completely abandon Bitcoin mining by 2028 and focus exclusively on serving AI clients.

[WillyAp]

Miners already have key assets that are particularly valuable today: large data centers, advanced cooling systems, land, and long-term electricity contracts. This is exactly the kind of infrastructure needed by tech giants such as Amazon, Microsoft, Alphabet, and Meta*, which are actively expanding their AI operations.

What they don't have is the hardware. But if that is not an issue the danger lies in less networking power.

[DrBeer]

Miners already have key assets that are particularly valuable today: large data centers, advanced cooling systems, land, and long-term electricity contracts. This is exactly the kind of infrastructure needed by tech giants such as Amazon, Microsoft, Alphabet, and Meta*, which are actively expanding their AI operations.

What they don't have is the hardware. But if that is not an issue the danger lies in less networking power.

To train neural networks, you need:
- equipment: mainly specialized accelerators and RAM
- to power all of this, a very high amount of electricity
- for all of the above, huge areas for data centers, with supporting systems such as cooling

[WillyAp]

To train neural networks, you need:
- equipment: mainly specialized accelerators and RAM
- to power all of this, a very high amount of electricity
- for all of the above, huge areas for data centers, with supporting systems such as cooling

Here is someone who did a test:

"I've got a bunch of GPU mining rigs, surely I can do something useful with them like train my own AI or provide hashing power to astronomy or genome folding operations?"

https://ai.stackexchange.com/questions/40511/can-gpu-crypto-mining-rigs-be-converted-to-train-or-run-ai

[DrBeer]

To train neural networks, you need:
- equipment: mainly specialized accelerators and RAM
- to power all of this, a very high amount of electricity
- for all of the above, huge areas for data centers, with supporting systems such as cooling

Here is someone who did a test:

"I've got a bunch of GPU mining rigs, surely I can do something useful with them like train my own AI or provide hashing power to astronomy or genome folding operations?"

https://ai.stackexchange.com/questions/40511/can-gpu-crypto-mining-rigs-be-converted-to-train-or-run-ai

Although I started this topic by giving the AI sector the benefit of the doubt, I increasingly have a feeling of déjà vu I suppose that the mega-boom will probably not end in a “dot-com crash,” but there will be a noticeable decline in the INVESTMENT attractiveness of AI.
 Today's COSTS for AI, in terms of mass consumption, will not meet investors' expectations, in my opinion, which will lead to an outflow of capital. This will not happen tomorrow, but soon enough, I would guess in 3-5 years.
 Therefore, the market will separate into “mass AI,” with relatively inexpensive systems for everyday use, and narrow-profile, already really expensive systems that actually bring significant profits because they will provide solutions in areas where AI can generate high added value, such as pharmaceuticals and the military-industrial complex.
 

[EarnOnVictor]

Crypto mining and AI development are two different subjects, in my opinion, which do not relate in any wide sense of it. Is it the fear of electricity, or any share/distribution of resources, I believe the two industries will thrive independently without hurting themselves.

What I still do not know is how you believe the two industries relate. Perhaps you need to shed more light because, even miners could co-own it and AI development projects. They do not harm each other, in my opinion. Besides, AI is being used in crypto mining now, it's all about how you view it.

[DrBeer]

Crypto mining and AI development are two different subjects, in my opinion, which do not relate in any wide sense of it. Is it the fear of electricity, or any share/distribution of resources, I believe the two industries will thrive independently without hurting themselves.

What I still do not know is how you believe the two industries relate. Perhaps you need to shed more light because, even miners could co-own it and AI development projects. They do not harm each other, in my opinion. Besides, AI is being used in crypto mining now, it's all about how you view it.

The collapse of the price of Bitcoin makes mining less attractive and much less profitable than investing in data centers for AI. It's simple and needs no explanation—the cost of Bitcoin is mainly determined by the cost of mining farms and their expenses—equipment + electricity consumption. Low Bitcoin price = abandonment of mining. Equipment downtime combined with mandatory payments (rent, obligations to suppliers, etc.) results in losses. And no one wants losses. This means they will look for a way out. What are mining farms very similar to? That's right—data centers for AI training....
       

[Oluwa-btc]

To clarify, I am not talking about the “end of mining,” I am talking about a tangible redistribution of resources and capital in the context of a rapidly growing AI market and demand for these services.
If there are miners out there, it would be very interesting to hear their opinion!

How realistic can it be that the shifting of mining power to AI, basically mining and AI operation operates separately some may have it that AI can have so much influence on the market if there's an opportunity for adoption in the crypto community. But then I'm trying to fix in my understanding to what exactly you're heading at.But regardless I think the growth of AI will definitely affect the economic and markets positively.

[DrBeer]

To clarify, I am not talking about the “end of mining,” I am talking about a tangible redistribution of resources and capital in the context of a rapidly growing AI market and demand for these services.
If there are miners out there, it would be very interesting to hear their opinion!

How realistic can it be that the shifting of mining power to AI, basically mining and AI operation operates separately some may have it that AI can have so much influence on the market if there's an opportunity for adoption in the crypto community. But then I'm trying to fix in my understanding to what exactly you're heading at.But regardless I think the growth of AI will definitely affect the economic and markets positively.

I am not exaggerating; I am openly stating the risks that could “distract” investors from the crypto world and force them to shift their focus to AI, at least for a certain period of time.
 As we can see, cryptocurrencies are currently in a less than favorable situation, while the AI market is attracting astronomical sums and buying up everything it needs, which often overlaps with the needs of the crypto market—chips, electricity, land plots. The simplest example is electricity. Here, AI companies are willing to buy up electricity for virtually “any price,” which could simply leave mining without the necessary power.

[STT]

I think AI is already separate in its designs from Bitcoin, the two might have started off parallel is some way via GPU relation though BTC was CPU to start but its not true now.
  I wish it were true, I miss my GPU mining days if only I had carried on without profit and some optimism.  BTC is ASIC now and AI is still closer to the GPU market so its the game or graphic designers who might be troubled by AI interference.

I agree on the competition idea economically and finance related perhaps, money is always in competition with alternatives.   I have heard that China harvests chips from desktop computers to repurpose them into AI data centres due to being blocked in their direct orders by politics based bans.
   The other element on that point would be that the industry itself wants people to stream their games not generate the results locally.  In this way the GPU could be used for both games and AI on demand when needed, probably making alot of money that way in a flexible service.   BTC itself may see a lower mining cost is something to consider possibly?

[shinratensei_]

It's all rainbows and sunshine until AI data center doesn't become profitable anymore and there is too much competition. The same way hashrate difficulty of bitcoin adjusted and make mining become more profitable than having AI data center.
This is called the game of rotating capital, if companies shifted focus to AI, soon enough the market become over competitive and their margin profit become thin.

Other concern is lack of GPU supply because semiconductor production is limited due to complexity and their capital needed to start AI data center will exponentially become more expensive making ROI longer.

[WillyAp]

I think AI is already separate in its designs from Bitcoin, the two might have started off parallel is some way via GPU relation though BTC was CPU to start but its not true now.

Litecoin still uses CPU, so does monero.
There must be more. 

[justdimin]

It's all rainbows and sunshine until AI data center doesn't become profitable anymore and there is too much competition. The same way hashrate difficulty of bitcoin adjusted and make mining become more profitable than having AI data center.
This is called the game of rotating capital, if companies shifted focus to AI, soon enough the market become over competitive and their margin profit become thin.

Other concern is lack of GPU supply because semiconductor production is limited due to complexity and their capital needed to start AI data center will exponentially become more expensive making ROI longer.

That is what is happening right now, many companies are looking the other way now. Most companies "promised" a lot of investment into AI, and now they are realizing it is not that profitable to keep doing that and this is why they are not putting more money into it.

Sure building the AI into being the best was a race, and every company wanted to have the best one, but they soon realized it costs more to try to beat your opponent than the money people are willing to pay for AI and that is why there is no reason why they should invest into it. The best part is that we are talking about something that will not be getting better at the rate it has been, but we can safely say that it will not stand where it is neither and may get better but slowly.

[shinratensei_]

That is what is happening right now, many companies are looking the other way now. Most companies "promised" a lot of investment into AI, and now they are realizing it is not that profitable to keep doing that and this is why they are not putting more money into it.

Sure building the AI into being the best was a race, and every company wanted to have the best one, but they soon realized it costs more to try to beat your opponent than the money people are willing to pay for AI and that is why there is no reason why they should invest into it. The best part is that we are talking about something that will not be getting better at the rate it has been, but we can safely say that it will not stand where it is neither and may get better but slowly.

Yeah, AI data center is a whale game. It requires billions injected to make it work. Not to mention most popular AI is closed source. Not every data center can host it nor commercialize it.
I just find the idea of switching to AI data center to be lacking of fundamental strategy to make it profitable.

Are they really trying to switch to AI because they found some alpha or they swithced due to FOMO because apparently not every AI data center is profitable.

[SquallLeonhart]

AI data center is a whale game. It requires billions injected to make it work. Not to mention most popular AI is closed source. Not every data center can host it nor commercialize it.
I just find the idea of switching to AI data center to be lacking of fundamental strategy to make it profitable.

Are they really trying to switch to AI because they found some alpha or they swithced due to FOMO because apparently not every AI data center is profitable.

I would say that it became a problem because of how many people wanted to use it and they couldn't turn a profit out of that. Everyone wants to use AI for free, nobody wants to pay for it, and while there are some people who are willing to do it, most don't.

So for example if you have a server and 90% of it are free users and only 10% pay for it, the server will not pay for itself and makes no sense to keep that going. I get that it is not a big deal, but it has to be something that will cause them to close up. The more paying customers they eventually get, the better they will be.

[philipma1957]

Microsoft and Meta are purchasing nuclear power plants for the purpose of developing and training AI networks.

MS, Google, Oracle, XAI all have very little to do with mining.
Halving cuts their earnings but they know well in advance.
There was also a mining rig in New York buying a power-plant, 
The Hash power is useless in AI
https://medium.com/@market-news24.com/who-knew-your-precious-crypto-mining-chips-are-useless-for-ai-427a8ca2df38

You probably misunderstood me, or perhaps I didn't express my thoughts very clearly. Let me repeat: I'm not saying that AI will take away miners' mining equipment! It's completely useless to AI, that's a fact!
I'm talking about resources and investments “changing course,” and among these, there are resources that can create problems for mining. More precisely, the lack or reduction of their availability will be a problem for mining. First and foremost, this is electricity. And as we can see, “all doors are open” for AI companies, which is not the case for mining companies. Plus, the decline in the investment attractiveness of cryptocurrency itself (investors are actively transferring money from cryptocurrencies to AI) will lead to the devaluation of cryptocurrency,
 and this causes a chain reaction—low prices in the long term—mining becomes unprofitable or loss-making due to the same very expensive electricity for mining companies.

you are spot on and here we are about six weeks later with mining camps using less hash and less power.

to all just read coinwarz btc diff charts and see the slow steady decline in diff and hashrate.

the last jump and this jump toss out since the cold snap in texas shutdown over 12% of the network last jump and it went back online now.

so basically the down 11 last jump and up 12 reflect that.

what has happened since November is a slow-steady drop in the diff some is due to ai grabbing power.


ie i have 10 megawatts of power at 3 cents
mining makes me 5 cents a kwatt

or i sell half my power contract for 4.5 cents to an ai company

zero risk and 1.5 cent profit

or take risk with mining of price goes down
or gear breaks
and certain labor cost.


well since I have 10 megawatts for five years I middle the risk
selling ½ to ai to get certain 1.5 cent profit


this means btc mining will shrink

now all of the above would be wrong if btc paid 150k say 9 cents or 12 cents a kwatt.


so right now we are going down.

it could change if ai nukes a city by accident and people say fuck ai.