Organized by: BitpushNews

With the rapid growth in demand for artificial intelligence (AI) and high-performance computing (HPC) computing power, more and more publicly listed Bitcoin mining companies are beginning to explore shifting their data centers, electricity, and infrastructure towards AI computing. This trend has continuously raised a question in the market: Is AI changing or even reshaping the future of the Bitcoin mining industry?

In an interview broadcasted on the Blockspace Podcast on March 10, Liang Wang, Vice President of Canaan, a leading global Bitcoin mining machine manufacturer, shared his observations on this issue. He believes that Bitcoin, as an asset, is still likely to experience a new bull market cycle, but that does not necessarily mean that the Bitcoin mining industry itself will see a boom similar to what it experienced in the past. The more important reason is not AI or HPC, but rather that the economics of mining are gradually deteriorating over time.

The following is a summary formulated by the Bitpush team based on the interview, slightly condensed for ease of reading while attempting to preserve the original meaning.

“Bitcoin will have a bull market, but whether mining will is something I truly don’t know”

Host:

Many publicly listed mining companies are now shifting their computing resources towards AI. What do you think about this change? Will AI change or even replace the Bitcoin mining industry?

Liang Wang:

I have been thinking about this question. First, I believe we must embrace AI and AI HPC, as this is indeed a huge change that will alter our way of life and change many job structures, with many jobs being replaced by AI, which I think is obvious.

However, if the question is whether AI will replace the entire Bitcoin mining industry, I personally do not think so. Bitcoin as an asset class is still valuable, and it has its own cyclical nature. So if you ask me whether Bitcoin will see another bull market in the future, I think the answer is yes.

But if you ask another question: Will the Bitcoin mining industry see a bull market? Frankly, I do not know the answer.

Because right now, what is drawing attention away from this industry is not just AI HPC. The more important reason is that the economics of mining itself are gradually worsening over time.

“It is much harder to enter the mining industry now than five years ago”

Host:

Why do you think the economics of mining are getting worse?

Liang Wang:

Because today, this industry is completely different from five years ago. Five years ago, if you could get your hands on a mining machine, you were likely to make a lot of money; many people in the industry really viewed mining machines as a type of "money printing machine." But now, that is not the case; entering this industry has become very, very challenging.

Look at it now, Bitcoin prices have fallen to about $65,000 to $70,000, but the total network computing power has not significantly decreased, right? This alone indicates a problem. Logically, if the industry were that profitable and the logic were clear, or conversely, if the industry were completely replaced by AI, you should see more significant changes in computing power. But that is not the case. The reality is that everyone is still keeping their machines running.

Why? First, because they need income. Even if they are no longer profitable, as companies, they still need revenue, need to maintain operations, and need to provide jobs for their employees. Secondly, Bitcoin mining now plays an increasingly important role as a regulator of the power grid. HPC runs 24/7, you cannot turn it off lightly, so the power grid needs Bitcoin miners, which can flexibly start and stop, to absorb peaks and valleys and help the entire power system expand. Therefore, many companies, even if their profits are thin or they are not making money, will not easily shut down their computing power.

This is why I say the issue is not only that AI has stolen the spotlight from mining; rather, it is becoming increasingly difficult for new miners to make money in this industry.

After 2028, monetary returns may no longer be the key driving force in this industry

Host:

How do you see this industry evolving over the next two to three years, or even after 2028?

Liang Wang:

Bitcoin has a mechanism that everyone knows but many have not fully considered the consequences: the halving every four years. What does halving mean? It means that if the price of Bitcoin does not double, the entire industry’s economic returns from block rewards are declining.

So everyone knows in their hearts that there will be the next halving in 2028. The question arises: If by that time the price of Bitcoin has not risen to $300,000 or lacked a level sufficient to support miner profits, how will this industry continue to move forward? This is a question I have been thinking about.

My view is that Bitcoin mining will continue to exist; it will remain a part of the entire energy landscape, but I do not believe that after 2028, monetary returns will still be the core factor driving this industry. I think it is more likely that this industry will continue to exist around several directions, such as grid balancing, waste heat recovery, home scenarios, or utilizing resources that traditional energy systems cannot use efficiently. But if you ask me if there will be a boom period where everyone rushes in and miners generally make a huge profit, I am really not sure.

Of course, I hope I am wrong. I certainly hope Bitcoin rises to $500,000 each, and then everyone rushes back into this industry. But that part, no one can really predict; I do not have a crystal ball either.

AI and mining are not a zero-sum relationship

Host:

A very popular saying in the market now is that AI will directly take away the electrical resources of Bitcoin mining. Do you agree?

Liang Wang:

I do not quite agree with understanding this as a zero-sum game. Because AI HPC and Bitcoin mining machines are essentially not the same type of load. AI HPC must run continuously; it needs to be online 24 hours a day, and most of the time, you cannot turn it off. But Bitcoin mining machines are different; one of their biggest advantages is that they can be turned off when needed and can rapidly start up when there is excess electricity.

This is also why I have always felt that in places like Texas, Bitcoin mining machines are actually welcomed by the power grid and energy suppliers. Because they can help the grid manage imbalance. When needed, they can yield electrical resources, and they can also "consume" excess electricity. Battery storage is certainly also a solution, but from a cost perspective, it is much more expensive than Bitcoin mining. So I have always viewed the position of Bitcoin mining machines in the energy system optimistically. I think they are not only not being replaced by AI; rather, they may become more important in the AI era in a different way.

So in my view, AI and Bitcoin mining can often coexist and even complement each other, rather than one forcing the other out.

North America remains one of the few markets with long-term predictability

Host:

From a regional perspective, where is mining growth most likely to occur in the future?

Liang Wang:

We have certainly looked at many places and conducted many trials. For instance, in Kazakhstan, we have had quite a lot of experience in the past. But the problem is that many countries will initially welcome miners because you help them digest idle electricity, but once local power usage tightens or the political environment changes, mining will quickly go from being "welcomed" to being "targeted."

This is why we place more importance on North America, especially the United States and Canada. It is not that there are no problems here, but overall it is more predictable. When you do business in the United States, at least you know what the rules are, you know what the differences are among different states, and you also know that if you truly bring employment, tax revenue, or help the grid, local levels may understand and support you. But in many other countries, the uncertainty is too high; sometimes just one sentence, one policy, or even a change in the attitude of a local authority figure can result in your entire business disappearing.

For a publicly listed company, this kind of long-term predictability is very important. You cannot use shareholders’ money to gamble on a market whose operability remains uncertain ten years down the line.

“China's situation is very complicated”

Host:

The outside world has always been very concerned about the situation in China's mining industry. What is your view on the current situation of the remaining mining sites in China?

Liang Wang:

The situation in China has always been very complicated, and I do not think this issue is a simple, completely top-down, unified process. China is vast, and considerations differ at various levels and regions. On a national level, there is more concern for financial stability, especially not wanting to achieve capital outflow through mining or Bitcoin trading, which has always been an important background. Therefore, since 2021, Bitcoin mining and cryptocurrency trading on an official level have indeed been prohibited, and this has not changed.

On the other hand, local levels have their own realities. For some regions, Bitcoin mining was once helpful; it could create jobs, generate tax revenue, and digest electricity that would otherwise go unused, especially during economic downturns. Many were willing to see miners as a way to monetize electricity infrastructure. So why do you still see some activity today? Because there is indeed a market demand.

But I want to say, we have never understood this situation as "China will welcome Bitcoin mining again." We do not see it this way. We do not think it is reasonable to bet our resources on self-operated mining in China. Because as a publicly listed company, you cannot make long-term capital investments based on speculation about whether a policy might change. This is why we have always focused on North America instead of gambling on whether China will come up with a new cryptocurrency strategy.

Has AI taken away mining machine production capacity?

Host:

Another question in the market is, with the surge in demand for AI chips, will it become increasingly difficult for mining machine manufacturers to secure chip production capacity?

Liang Wang:

I do not think production capacity will be completely taken away by AI. In the short term, there will definitely be periods of chip production tightness, as well as periods of easing, but stretched over five to ten years, I think supply and demand will return to a relatively balanced state.

Foundries like TSMC and Samsung are engaging in long-term investment plans worth tens of billions or even over a hundred billion dollars. They are looking at the next ten years, not just the next year or two. Mining machine chips are certainly a part of their business, but they have never been the whole, nor necessarily the most critical part.

In fact, what is more crucial is whether you have a long-term cooperative relationship with the foundry, whether you have successful experience in tape-out, and whether you can actually create a competitive product in a given process. Because with each generation of technology, the research and development investment is astronomical, and the cost of trial and error is also extremely high. New players lacking these accumulations will find it hard to squeeze in, even if they have money.

This industry is not about "I want to create the most advanced chip," but about whether you can make it and whether anyone will buy it once you make it. Therefore, I tend to see this as a long-term technical partnership issue rather than a simple understanding that AI has taken away the production capacity of mining machines.