A New Understanding of Ethereum POS
The consensus mechanism of Ethereum has shifted from the previous POW to the current POS, which has always been a point I didn't quite understand.
The earliest explanation I saw for this transition came from Vitalik himself, who suggested that this change was for environmental protection and energy conservation…
Regarding this explanation, at that time, I thought:
I can't say this explanation is completely unreasonable, but it sounds more like a viewpoint that caters to a certain kind of Western leftist political correctness.
Therefore, I had a rather forced attitude towards it.
However, recently, with the exponential growth of energy consumption due to artificial intelligence applications, along with my reinforced recognition of Ethereum's "use value," I am increasingly inclined to believe that this transition of Ethereum is correct and necessary.
The exponential growth of energy consumption due to artificial intelligence applications is something I believe most readers have already read various concrete evidence from different information channels.
Among this evidence, what impressed me the most was a piece of information recently revealed by Microsoft's CEO in an interview: Microsoft currently has a large number of Nvidia GPUs piled up in warehouses and cannot be used, and the fundamental reason is that the power supply system cannot keep up.
Moreover, several major tech giants in the U.S. are now increasingly investing heavily in building small nuclear power plants to ensure the energy supply for data centers, which is an undeniable strong driving factor.
It is clear that the bottleneck for artificial intelligence applications has gradually shifted from GPUs to the energy sector.
In addition, there is another interesting piece of information circulating in the industry:
More and more Bitcoin mining farms are beginning to transition to data center operations.
I believe this business transition should not be understood merely as a commercial phenomenon, but rather we should pay attention to a potential risk hidden behind it:
If at some point in the future (or is it happening now?), the revenue from Bitcoin mining does not keep up with the income from data center operations, what does that mean?
That means more and more miners will convert their mining farms into data centers, and more and more miners will withdraw from maintaining the Bitcoin network, and the energy-dependent security barrier of Bitcoin will be eroded and weakened by artificial intelligence applications.
Such a risk did not exist during Satoshi Nakamoto's era because there were no widespread applications of artificial intelligence like today, nor the future exponential growth potential that artificial intelligence now exhibits.
Of course, this risk is merely my current speculation.
I do not have accurate data to compare the breakeven point and business switching threshold between mining farms and operating data centers, so I cannot definitively determine whether such a risk exists and, if so, under what circumstances it would arise.
However, I believe this risk can no longer be ignored. Moreover, I think that as we move forward, when artificial intelligence truly emerges with blockbuster applications like WeChat, this risk will definitely increase.
Therefore, from the perspective of competition for limited energy, future artificial intelligence applications are undoubtedly the biggest threat to POW mining. And the consensus mechanisms that can mitigate this threat at least include POS.
In the article the day before yesterday, I listed the "blue fox's" vision for the future of Ethereum, one of which is to run nodes on simple small devices—actually, continuously lowering hardware thresholds and continuously reducing energy consumption has always been what Ethereum has been striving to achieve.
Now it seems that after transitioning to POS, this approach is indeed very likely to mitigate the risks of Ethereum and artificial intelligence in energy competition.
This is one aspect.
Secondly, if we consider Ethereum from a broader perspective, since it is an application network, the ecosystem centered around it should ideally be as minimally constrained by the physical world as possible, as it is meant to support on-chain transactions for the entire virtual world.
And the limitations from energy are a significant constraint from the physical world.
Therefore, as an application protocol, it should strive to avoid such limitations, and quickly and early "detach from the physical" to become a "solid foundation" in the virtual world that relies as little on energy as possible.
In summary, whether by coincidence or historical chance, I now increasingly hold a supportive attitude towards Ethereum's transition from POW to POS.











