bitcoin mining difficulty keeps increasing, will more miners shift to AI?

[Ziskinberg]

According to the 2026 Difficulty Thread by @philipma1957, Bitcoin mining difficulty has already increased, and the latest adjustment was +5.9237%, which is the third-largest percentage increase this year.

So I found this article in our local board and the news is about Genesis Digital Assets, which was one of the biggest Bitcoin mining companies before, but is now slowly turning its mining infrastructure into AI data centers through SWI Group.

So as mining difficulty continues to increase, do you think more Bitcoin mining companies will move to AI data centers?

[SquirrelJulietGarden]

According to the 2026 Difficulty Thread by @philipma1957, Bitcoin mining difficulty has already increased, and the latest adjustment was +5.9237%, which is the third-largest percentage increase this year.

So I found this article in our local board and the news is about Genesis Digital Assets, which was one of the biggest Bitcoin mining companies before, but is now slowly turning its mining infrastructure into AI data centers through SWI Group.

So as mining difficulty continues to increase, do you think more Bitcoin mining companies will move to AI data centers?

Some Bitcoin miners will move to work for AI data centers but some will stay and I believe that if people or companies behind Bitcoin mining farms are well prepared their finance for operations in long term, mining and holding most of their bitcoins will be better than working for AI data centers. With time, Bitcoin will have very good ROI that brings great profit for Bitcoin miners which manage can hold most of their bitcoins from mining.

It's not bad to work for AI data centers if currently and in the near future, it brings better income than from mining Bitcoin but because of better ROI from Bitcoin in long term, I would like to use most of my cash or fiat currency received from AI data center payments for purchasing bitcoin, and holding bitcoin for long term investment.

[Ziskinberg]

Some Bitcoin miners will move to work for AI data centers but some will stay and I believe that if people or companies behind Bitcoin mining farms are well prepared their finance for operations in long term, mining and holding most of their bitcoins will be better than working for AI data centers. With time, Bitcoin will have very good ROI that brings great profit for Bitcoin miners which manage can hold most of their bitcoins from mining.

That's if all operators are looking for long-term profit.

But the reality is, they don't look at it as a long-term investment, at least not all of them they are running a business with expensive operations so they want to ensure that it stays profitable, with the high difficulty of mining some of them will likely try AI data centers as the demand is bigger than Bitcoin mining.

Though it's not a concern as if they leave, the competition will drop, resulting in better rewards for the remaining miners.

It's not bad to work for AI data centers if currently and in the near future, it brings better income than from mining Bitcoin but because of better ROI from Bitcoin in long term, I would like to use most of my cash or fiat currency received from AI data center payments for purchasing bitcoin, and holding bitcoin for long term investment.

again, another investment recommendation. can we stay at the topic?
Let us not consider the holding part, let's compare the profitability of these 2 industries.

[Filicius]

According to the 2026 Difficulty Thread by @philipma1957, Bitcoin mining difficulty has already increased, and the latest adjustment was +5.9237%, which is the third-largest percentage increase this year.

So I found this article in our local board and the news is about Genesis Digital Assets, which was one of the biggest Bitcoin mining companies before, but is now slowly turning its mining infrastructure into AI data centers through SWI Group.

So as mining difficulty continues to increase, do you think more Bitcoin mining companies will move to AI data centers?

Partly yes, but keep in mind that the drop in hashrate would affect the rate at which blocks are produced, which in turn would reduce the mining difficulty as a consequence. Not immediately or truly proportionally, but it would act as a mechanism of self-regulation.

As for the fact that in the long run it would not compensate for the expected appreciation of Bitcoin, we all wish that to happen, but most large businesses, apart from Strategy and few others, seek immediate profitability, and I don't know if the owners of the big mining infrastructure will be willing to take such a gamble when the ROI of the AI alternative is better.

Edit: It seems that you have addressed precisely these two points in your previous post, while writing.

[Alvin_talk]

So as mining difficulty continues to increase, do you think more Bitcoin mining companies will move to AI data centers?

Difficulty alone is likely not something that will make mining companies change their focus to ai, since it can go up and down, depending on network conditions. Personally, I think companies will consider halving to be more of a significant challange that will make them to consider transitioning into other industries (including ai) than difficulty. Because if bitcoin price does not go as espected in the coming years, mining companies might not be making enough profits because of reduction in block rewards and in such a situation, some will definitely transition to other industries to remain operational and more profitable.

[coinlary]

Difficulty isn't  really the problem here and the fact that we are experiencing increase in difficulty  means there's  is also an increase in hashrate and that can only happen when there's increase in more people willing to mine or willing to contribute more mining power. If that should decrease the difficulty will adjust. 

Imo, we will just have bigger pools overtime but definitely we will lose some miners to AI, that doesn't mean there won't be new miners.

[gracreavix]

I don't think it's because the mining difficulty has increased then we will see more mining centres moving to AI data centres. I mean nothing worth hard would be easy. They could be a lot of reasons why they are switching, what if the money they are making is not as much as if they had an AI data centre.

If you even see the world now, AI is moving fast and the demand is at all time high. This is just basic economics at play. When Bitcoin was on everyone's lips, people where becoming miners. Now it is the AI era. Doesn't mean that there isn't profit for miners. They are just following the trend of economics.

[Antidote47k]

 I think currently running an AI facilities looks more profitable and to some extent more stable for people to have a long term plan around, in fact some AI contracts may not immediately translate to revenue but since it can offer better risk adjusted returns for operators, they are starting to consider it for long term.
Coinshares’ Q2 2026 reports shows estimate of $1.5m per MW for AI when compared to $0.5m per MW for bitcoin mining, I don’t know how substantial this research was and I don’t necessarily think it’s more profitable for everyday miner but we have to acknowledge that Bitcoin mining depends on it’s price and other factors including halving while AI facilities can secure long term contracts which makes future revenue more predictable and relatively safer to plan around but contracts alone don’t guarantee profitability.
Then again if it’s long term security operators are after bitcoin mining shouldn’t be any less profitable than AI facilities if miners are holding some portion of their btc, the value of bitcoin is likely to become even more valuable, in the long term I can’t really say who will end up with more profit. AI facilities offers a more predictable Cashflow while bitcoin mining offers the potential upside when btc appreciates.

[joniboini]

while AI facilities can secure long term contracts which makes future revenue more predictable and relatively safer to plan around but contracts alone don’t guarantee profitability.

I'm not sure how exactly those contract works, but some news suggest AI companies are basically drowning in debt to scale their operations. I'm not sure where the money comes from and how long until they can pay that off, though I'm sure more people will definitely use AI (cloud-based or not). I'm still skeptical the demand is that big though.

Speaking of miners, I'm sure I saw similar news a few months ago. It's easy to make headlines so media tend to focus on this side, even though there are more to discuss about mining and why the number of players fluctuate.

[bitbollo]

It has already happened to one of the biggest mining company last year https://www.mara.com/ This has included also sales of over 15K btc to finance their operation.
The news at that time was both interesting for AI investors (it has confirmed the interest by market operators) and for bitcoiners (that have seen a bigger player changing their business).

IMHO mining is always a "decentralized" process. It's really hard to make prediction about the market choice of a businesses.
The rise of difficulty could be seen also as "efficient" operator - an increasing competition. Meanwhile AI could just be solved by more and more efficient systems/models.

[X-ray]

I think a lot of us underestimate how costly it is to convert mining facilities to AI data center.

Besides the ever increasing HBM memory cost, it needs to have high redundancy networking, denser power distribution, and good uptime.

Nobody is going to throw away their ASICs and replace it with GB2000 NVL72 on a whim because they see 6% difficulty increase.

It's more realistic to say that miners stopped expanding their mining farm and decided to allocate their money into building a new AI data center.

[Hypnosis00]

I think a lot of us underestimate how costly it is to convert mining facilities to AI data center.

Besides the ever increasing HBM memory cost, it needs to have high redundancy networking, denser power distribution, and good uptime.

If they have a complete analysis and see that moving to AI data center servers is more profitable, they will follow that path.

That's how businesses work they are risk takers in the first place, so they take calculated risks that would help them.

Nobody is going to throw away their ASICs

Throwing away is pure stupidity... why not sell it?

[Questat]

Difficulty isn't  really the problem here and the fact that we are experiencing increase in difficulty  means there's  is also an increase in hashrate and that can only happen when there's increase in more people willing to mine or willing to contribute more mining power. If that should decrease the difficulty will adjust. 

Imo, we will just have bigger pools overtime but definitely we will lose some miners to AI, that doesn't mean there won't be new miners.

That's possible mate but we also have to understand that not all miners have the same operational costs and profitability, well there are some who will continue (those who are profitable, but those who feel they are not will leave and look for better opportunities.

And considering what you said that new miners will come in, it still does not guarantee that the existing miners will remain profitable, especially if they are using older equipment which we know what would result and of course the higher electricity rates, those can change, so in the end it's all business as we can say that new miners will come but it's not about that, it's about whether mining is still more profitable than AI data centers.

[Outhue]

while AI facilities can secure long term contracts which makes future revenue more predictable and relatively safer to plan around but contracts alone don’t guarantee profitability.

I'm not sure how exactly those contract works, but some news suggest AI companies are basically drowning in debt to scale their operations. I'm not sure where the money comes from and how long until they can pay that off, though I'm sure more people will definitely use AI (cloud-based or not). I'm still skeptical the demand is that big though.

Speaking of miners, I'm sure I saw similar news a few months ago. It's easy to make headlines so media tend to focus on this side, even though there are more to discuss about mining and why the number of players fluctuate.

Basically drowning in debt how? I thought running a data center is far more profitable than Bitcoin mining? Ai data center is still very early and to get into this very early is why many companies are going for it, the last time I looked into this I saw that Bitcoin mining is way more harder to run than AI data center.

[Antidote47k]

while AI facilities can secure long term contracts which makes future revenue more predictable and relatively safer to plan around but contracts alone don’t guarantee profitability.

I'm not sure how exactly those contract works, but some news suggest AI companies are basically drowning in debt to scale their operations. I'm not sure where the money comes from and how long until they can pay that off, though I'm sure more people will definitely use AI (cloud-based or not). I'm still skeptical the demand is that big though.

Speaking of miners, I'm sure I saw similar news a few months ago. It's easy to make headlines so media tend to focus on this side, even though there are more to discuss about mining and why the number of players fluctuate.

That is true, I think it’s more about the predictability of it with regards to revenue without making the underlying investment less risky in every aspect. AI is the future, there is no doubt about that, as for where this technology will take us, we’ll just have to wait to find out.
The debt issue is not unfounded but I think across the contract’s life the revenue likely to be generated will be sufficient to cover the investment and associated costs or atleast that’s the projection, anything can happen along the line, who knows…. For these projections to hold up operating costs, financing over time and most importantly demand has to be taken into consideration.
The part I’m curious about most is the contract structure itself, are mining companies the ones bearing the financial and utilization risk or the Ai customer bears most of the demand risk??

[pooya87]

It is more complicated than that.

As you most probably know bitcoin miners own specialized hardware called ASIC that is only capable of doing one thing and that is to compute double SHA256 hash of a block header. Meanwhile the computation required by AI is different (Matrix Calculations) and requires different hardware (special GPUs).

So for a bitcoin miner to switch to AI "miner", they'll have to get rid of the hardware they own and have made a significant investment in and replace it with other hardware. To make such a switch, these businesses require 2 things:
1. Massive incentive. The profit in AI sector has to high enough to incentivize them to make that switch.
2. Massive insurance. Switching hardware is a big commitment, it is not something they can undo if it didn't work out. So the AI sector has to show a solid future for them to make that commitment.

I don't think these two conditions are being met for bitcoin miners to make that kind of switch at large scale. AI sector is not yet that profitable, in fact they are losing money for the most part! Its future is also not guaranteed. Remember what happened the day China revealed its AI sector and crashed the US AI market? That can repeat.
Meanwhile Bitcoin's future is guaranteed better than AI future and the profit is also there. The miners aren't losing money like AI companies do either.

Also keep in mind that while bitcoin's difficulty is rising, its price is rising as well... in the long run.
For example ~8 years ago (2018) price was $3k, ~4 years ago (2022) price was $15k but today (2026) it is $80k+. How much will the price be in 4 years from now in 2030? Does AI provide the same growth potential? I don't think so. AI is currently in the over-hyped bubble mode and that is not going to last long while bitcoin's adoption is going to increase and alongside its capped supply ensures profitability.

[SquirrelJulietGarden]

That's if all operators are looking for long-term profit.

But the reality is, they don't look at it as a long-term investment, at least not all of them they are running a business with expensive operations so they want to ensure that it stays profitable, with the high difficulty of mining some of them will likely try AI data centers as the demand is bigger than Bitcoin mining.

Though it's not a concern as if they leave, the competition will drop, resulting in better rewards for the remaining miners.

again, another investment recommendation. can we stay at the topic?
Let us not consider the holding part, let's compare the profitability of these 2 industries.

If they can get much more considerable profit by working for AI data centers, they can enjoy profit and use part of profit for buying bitcoin. This practice will at least keep them staying in Bitcoin market and still accumulate bitcoin with time. If their fund for bitcoin accumulation is from profit with AI data mining, it's not big problem because we know Bitcoin with a long time of holding generates profit passively for Bitcoin holders.

Profitability of two industries include many things like short term profit and long term profit. Long term with Bitcoin relates to holding.

So for a bitcoin miner to switch to AI "miner", they'll have to get rid of the hardware they own and have made a significant investment in and replace it with other hardware. To make such a switch, these businesses require 2 things:
1. Massive incentive. The profit in AI sector has to high enough to incentivize them to make that switch.
2. Massive insurance. Switching hardware is a big commitment, it is not something they can undo if it didn't work out. So the AI sector has to show a solid future for them to make that commitment.

I don't think these two conditions are being met for bitcoin miners to make that kind of switch at large scale. AI sector is not yet that profitable, in fact they are losing money for the most part! Its future is also not guaranteed. Remember what happened the day China revealed its AI sector and crashed the US AI market? That can repeat.
Meanwhile Bitcoin's future is guaranteed better than AI future and the profit is also there. The miners aren't losing money like AI companies do either.

I don't know about challenge for switching hardware from Bitcoin mining to AI data centers but you enlightened me about that. I only thought that Bitcoin miners can simply switch their ASICs to work for AI data centers, remotely and no need of physical migrations to AI data center locations. If the switch requires this physical migration, it's too cumbersome with a lot of cost.

[Sammysmart001]

So for a bitcoin miner to switch to AI "miner", they'll have to get rid of the hardware they own and have made a significant investment in and replace it with other hardware.

@pooya87 you just said it all. One can not just update a software to turn a Bitcoin ASIC miner into an AI processor cause the hardware aren’t same, they are completely different. And again AI data centers do requires expensive cooling and 100 percent continuous power. When electricity gets so expensive Bitcoin miners can easily shut down. What a really see is that when mining do say that they switching to AI they are simply leasing their power contracts and also their land to a big tech companies.

[Darker45]

This has been a persistent topic for years already.

Bitcoin miners become source of compute for AI?
The shift of mining power to AI: potential risks for the crypto market in 2026+
https://bitcointalk.org/index.php?topic=5579146.0
AI data centers taking over?

You mentioned Genesis Digital. Apart from that, there's also Bitfarms which announced their decision this year that "we are no longer a Bitcoin company". The only difference is that Genesis' pivot is mainly due to a new ownership while Bitfarms' decision is basically coming from within. That's like a legit change of heart. And then there's also TeraWulf and IREN Limited.

The rest like MARA, Core Scientific, and Hut 8 aren't entirely shifting focus. They're like having a hybrid approach.

Notwithstanding, it's apparent that mining difficulty has continuously recovered since its ATH a year ago. It's possible further recovery of Bitcoin's price will push the difficulty even higher.


https://www.coinwarz.com/bitcoin-difficulty

[viljy]

I read the first post and at first I didn’t understand how ASICs could be used instead of chips for AI. Then it dawned on me - we’re talking, so to speak, about repurposing a business. So what’s so extraordinary about that? It’s a common phenomenon in business. You could also call it expanding the business’s scope.

Although people talk about the AI bubble, in the end, the computing infrastructure will remain and will be in demand even if, say, the bubble bursts in the coming years. Computing power will, in any case, be a good investment for the future. So, I consider this to be a well‑considered decision, and not just because the difficulty of mining Bitcoin has increased. Besides, the cost of electricity has probably risen recently (due to the increase in the cost of energy resources).