GuShiio.com DeepSeek’s success proves to us that reforming the past is still the most effective way to break through the dilemma of poverty.
Author: YBB Capital Researcher Zeke
1. AI is inevitable, but Crypto x AI?
At the beginning of 2025, DeepSeek, created by magic square quantification, has dropped a “nuclear bomb” on the AI circle. A China AI model that uses only 2048 NVIDIA H800 GPUs and has a training cost of US$5.58 million (about one-tenth of Meta) has directly achieved the level of hand-wrestling with GPT-4o and Llama 3.1 in benchmark tests such as MMLU and GPQA. It even slightly surpasses these top large models from Silicon Valley in terms of complex reasoning and Chinese semantic understanding. The United States has imposed a chip blockade on China for many years, but it encountered dramatic deconstruction on DeepSeek’s computing power sand table. The Chinese AI under the siege has finally developed a technological path that suits national conditions but rivals cutting-edge. A set of open source, low-cost, and homogeneous combination of punches directly penetrated the United States ‘computing power moat.
China’s technology products, which are always slightly inferior to the United States in terms of performance, always give me a stereotype of cheap and plagiarism. I believe this is also the intuitive perception of most people about the products of China Internet companies. But this time DeepSeek is really different. I will not talk about whether the experience is higher than the subjective feeling of ChatGPT. Judging from the tense reactions of American politics and technology giants frequently mentioning this matter, China is no longer playing the role of a follower of technology this time, and the global shock caused by this is also very strong.
Although the first thing affected was my wallet, which to some extent was indeed caused by a misjudgment of the development of traditional AI, I still want to talk about my opinion, that is, DeepSeek’s impact on Crypto.
1. NVIDIA was the most seriously injured in this incident. First, its AI computing power requirements were questioned. Second, NVIDIA’s unified computing architecture “CUDA” was bypassed. Friends familiar with the AI field should be familiar with CUDA. It is one of the important cornerstones for promoting the development of modern AI. When large model developers use NVIDIA GPUs, they generally do research and development based on CUDA. Using CUDA has lower requirements for developers, because some functions are already encapsulated in CUDA, so there is no need to pay attention to too many details when using it, but execution efficiency will definitely be lost.
Because CUDA is a universal programming framework, it results in a loss of flexibility when training models. DeepSeek’s approach is to directly use PTX (NVIDIA’s intermediate instruction set framework for GPUs) to circumvent hardware limitations on training speed and shorten training time. Other models take 10 days to train, but DeepSeek can complete it in 5 days. This also means that if DeepSeek intends to adapt to China’s domestic GPUs in the future, it will be more convenient in terms of hardware adaptation, and Nvidia’s AI chip throne will also be shaken. (This paragraph comes from a report by Korea Future Asset Securities on DeepSeek’s training session)
In addition to the strong impact that a fall in stock prices will have on the currency circle, which is strongly correlated with U.S. stocks, I personally think it will be a good thing for decentralized computing power projects in the long run. First of all, more personal GPUs can also take advantage of their waste in the future. Secondly, if DeepSeek’s small and beautiful open source model path is successful, it will force a large number of AI companies to be forced to open source. In the future, the demand for computing power required for local deployment and secondary development will be more and more vigorous. Judging from the hardware requirements of DeepSeek R1 with a minimum parameter of 1.5B to a maximum parameter of 70B, from the lowest NVIDIA GeForce GTX 1660 Super to 40-series and 50-series, to professional-grade A100, The H800 GPU has the opportunity to contribute extra computing power again. For decentralized computing power projects that currently have a relatively transparent sense of existence and a somewhat weak supply of computing power, it may be an opportunity to turn around (of course, if the delay is low enough).
2. The AI framework project was the hottest emerging track in Crypto before DeepSeek dropped the “nuclear bomb”, and it was also the last track I wrote about before the Spring Festival. Nowadays, under the bombardment of DeepSeek, almost everyone is moving closer to zero. After all, others can compete with OpenAI at a cost of less than 6 million yuan, and our FDV’s leading projects of more than one billion have yet to produce some that can be called an AI Agent with practical value.
We have been almost paranoid in pursuing capitalization since the era of inscriptions. Nowadays, the currency circle is quite open to inclusiveness of capitalization. AI framework projects that are not linked at all only need an open source Github code library and a social account to issue coins. The price of “using the library to convert coins” is to be able to accept the dimensionality reduction blow thrown by traditional AI companies one day.
In the golden age of AI development, DeepSeek will not be the only big killer that traditional Internet companies can resort to. The development of AI between the United States and the United States will only get faster and faster. What direction does Crypto need to combine with in the upstream and downstream aspects of AI to highlight the advantages of decentralization and not be killed by an inexplicable AOE is the key to the issue. Roughly dividing the technology stack of Crypto x AI into computing power layer, data layer, middleware layer, and application layer. In today’s layers, I have never felt the need for Crypto to’s existence. It may be a good perspective to consider privacy and security from a future perspective. After all, it has become a reality for AI Agents to replace or assist human work today. How to ensure the privacy of AI processed working data and private data may be an issue that traditional Internet companies cannot solve. Furthermore, if an AI Agent can have payment authority, how to ensure the security of the wallet will also become a problem. Using blockchain as the compliance audit layer of the AI model should be our main development direction in the future.
On the other hand, incentives should be given in which direction. In addition to promoting computing power layers and model sharing, incentives can also allow humans to teach AI how to interact with the virtual world. Unlike LLM training, which naturally has decades of world Internet corpus knowledge, teaching AI how to act correctly actually requires humans to constantly mark the correct position. Just like teaching visual models to recognize what animals are and what cars are, which cannot be accomplished by an outsourcing company finding a group of college students, the birth of an AI Agent that can interact with the virtual world requires a complex decentralized network. Teaching AI through countless human individuals is also a direction. This has been discussed in more detail in my past articles, but what else can be done around motivation? Combined with DePin, it teaches agents to interact with reality, encourages AI to gain attention, stimulates secondary innovation about AI (such as Bittensor’s incentive method is a good example), and the Token incentive mechanism is automatically adjusted by AI (This is derived from a question I had in a previous article, the original words are as follows: When a decentralized project becomes extremely large and enters the mainstream, what should we do about deflation and inflation? Do we rely on simple code with rules, or do we listen to a project party with only a few people or dozens of people? Or are they those soul figures? Oh, by the way, we also have governance tokens. It’s just that governance tokens are meaningless until the witch problem is solved. Democratic voting can never be reflected in governance proposals. After all, a16z can veto the approval vote of a large community with only a few wallets. What is the point of voting? Wait.
We really can’t gather a group of high-tech AI talents like traditional Internet companies and buy or rent a large-scale GPU cluster for training. Rebuilding a DeepSeek in the blockchain is naturally a fantasy. Crypto exists to give another field irreplaceable decentralization characteristics, just as we once gave financial freedom. AI is an inevitable narrative for mankind, but what role can Crypto play in it?
3. This is the first time I have talked about Wordcoin in an article. Even now, I find this crypto-utopia project initiated by Sam Altman a bit ridiculous. Whether to enter the iris means that you need to make a choice between state surveillance and corporate surveillance. This kind of thing is like choosing to take the red pill or the blue pill in the matrix.
However, the concept of universal income or inclusive finance may no longer be a joke at this stage. DeepSeek, which can be deployed locally to rival cutting-edge large models, has made agents appear in hospitals and government agencies in China. According to McKinsey’s 2024 forecast report, 50% of jobs will be replaced by AI as soon as the next six years. Future versions of Wordcoin may even be issued uniformly by the government. If this process continues to intensify, tokens related to the concept of inclusive finance may also appear in large numbers and be repeatedly hyped. According to the window period of five to six years, it also coincides with Trump’s term of office. So will the encryption president issue a similar Token? I think it’s very possible.
4. According to Elon Musk’s recent remarks, AI may win the Nobel Prize for 25 years. So I think promoting AI scientific research through blockchain funding (and even contributing computing power, storage, methods and other resources) is more interesting and effective than the current DeSci. Maybe I can call it Decentralized AI Science? That is DeAIS.
2. Meme Coin is no longer Meme
When we analyzed Meme Coin, we talked about subculture, community consensus, and communication effects. Now, sitting in front of GMGN, I analyze the cabal, the locomotive address, and Dev’s rat warehouse every day. When a CA is sent to each dog group, it is the time to charge. Today’s Meme is more ridiculous than at any time in the past. At this stage of Pump.fun, let alone finding a Token that can sleep peacefully, you may just go to the toilet, and the K line will fall directly off the cliff.
The continuous simplification of asset issuance thresholds and the high anonymity of blockchain have contributed to the intensification of this crazy casino culture. An unknown grassroots team can use the currency circle as a cash machine. And Meme’s evolution direction is becoming more and more arbitrary. Not only is it possible to monetize the library as described above, but any event, anyone or even any AI can monetize it. Without a cultural core and consensus, it only takes a few weeks for the so-called leader of founding a sect to be completely forgotten. The celebrity coin trend that started with Trump only lasted for a month. Under a tweet from President Millay, hundreds of millions of funds flowed from Sol to the circle again, and Meme began to ebb. President Millay’s reaction to this was also very simple. He deleted the tweet and replied to everyone: I didn’t know.
The rapid development of AI has taken away too much attention from the world, and the development of technologists has been unable to move forward. Individual investors who give up value investment can only bet on being a lucky few in the pig killing plate. The increasingly scarce liquidity has been repeatedly extracted, reflecting in reality the day-to-day red K lines in Cex and Dex, as well as the disdain of traditional capital and outsiders for copycats.
3. You cannot ask for a sword from a carved ship.
The law of cycle has obviously failed, and all ideas of crafting a boat to seek a sword are difficult to find. Just because BTC loses cattle does not mean that the shanzhai loses cattle, but BTC loses bears and the shanzhai will definitely win bears. Our understanding of shanzhai must be reformed. The shanzhai market is no longer a market that can be supported by a white paper. A large project launched in the head Cex must be mature enough to support its own currency price.
Looking back at the token growth in the past seven years, there were less than 2000 tokens listed on the market in 2017, while nearly 25000 tokens listed in 2024 (data comes from CoinGeko, including removed tokens), the exponential expansion of the number of tokens is essentially the irreversible evolution of a blockchain low-entropy value system to a high-entropy noise system. While each token in 2017 still carries the ideal of subverting the world, the tokens in 2024 have evolved into a bargaining chip for liquidity exit. The emergence of more tokens has not led to more innovation and implementation, but the high valuation of star projects has exponentially doubled the demand for market liquidity.
As mentioned above, retail investors cannot afford the valuation of these projects without recognition from the outside world. The listings of most copycat projects tend to be at an all-time high, and Binance is the last stop. The currency circle needs an innovation, and star projects should be worthy of their huge financing. The public disclosure of the financial report of Bybit’s pilot project may be a solution. But in my humble opinion, the market needs a deep bear to reshape the valuation system and listing standards for today’s copycat projects.
4. Confusion
I once saw a ray of light on Ton and thought that the beginning of Crypto consumer applications had arrived, but this short-lived light died as the Tap to Earn trend receded. Five years ago, the liquidity mining derived from DeFi caused the currency circle to experience a peak moment of great glory. Five years later, our only successful area is DeFi.
The topic I communicate with insiders nowadays is also very simple. Have you bought BTC? Are you short? Give it a CA. Everyone is very confused. We can’t find the right direction. Except for BTC, buying any kind of token cannot sleep well is the current situation in the coin circle. Diamond Hands is not a positive word today. If you don’t buy BTC, then it is more like a synonym for fool.
When I turned on various media links on my mobile phone, it was as if I were reading the New York Times and gossip media. All kinds of phenomena reflected that most of the hopes of this circle had been pinned on policies and eyeballs. From a VC’s perspective, we should probably only invest in tool products in the future, and the asset issuance platform should become a seller and rent collector in order to survive.
conclusion
Obviously, this is not the status quo we want to see. Although Crypto seems to be in the fog of disorientation at present, DeepSeek’s success proves to us that reforming the past is still the most effective way to break through the dilemma of poverty. Crypto currently has the best policy coverage, attention, capital amount and solid infrastructure in history. In the near future, there will be many counterfeit ETFs that can inject another liquidity into the currency circle. We are clearly in the mainstream, but we are trapped under our own siege. There may be a turning point behind the ebb tide of Meme Coin today. AI may not be the only one in the future of mankind.
