Discuss potential risks in the AI industry chain, the U.S. tech sector, and the crypto market with frontier tech investor Didier and secondary market researcher Minta.
Editor: Wu Says Blockchain
This issue of "Wu Says Not Crypto" invites frontier tech investor Didier and secondary market researcher Minta to discuss controversies related to Jensen Huang and rumors of management changes at Nvidia, focusing on the AI industry chain, the U.S. tech sector, and potential risks in the crypto market. Didier believes Nvidia is not just a chip supplier but also influences the entire AI infrastructure ecosystem through investments, procurement, and credit support. If its expansion strategy retracts, it could lead to financing difficulties for emerging cloud service providers, a slowdown in data center construction, and a shock to the entire hardware industry chain.
Regarding market outlook, Didier judges that the investment cycle for AI hardware may have peaked, with some emerging cloud service providers facing severe funding gaps, potentially leading to a reshuffle in the industry. Minta is relatively mild, believing that a slowdown in capital expenditure, rising financing costs, and restrictions on data center construction will continue to suppress the market, but there may still be structural opportunities in AI software, computational efficiency optimization, and semiconductor alternative supply chains. Both agree that Bitcoin is currently impacted by insufficient liquidity and declining risk appetite, but the long-term development logic has not fundamentally changed.
The discussion also touches on the potential impact of changes in U.S. politics on AI and crypto. Didier is optimistic about the development space for compliant perpetual contracts, believing that the AI and crypto industries need to shift from aggressive expansion to sustainable growth. Minta emphasizes that crypto cannot solely rely on policy expectations; attracting real users and new capital with on-chain stocks, derivatives, and yield products is key to long-term growth.
Special note: Content involving Jensen Huang's personal controversies, Nvidia management changes, political connections, and related investigations are all rumors, discussions, or personal speculations from the program's guests and have not been independently verified, and should not be seen as confirmed facts. Market forecasts and corporate financial assessments in the text do not represent confirmed results.
Guest speeches do not represent Wu Says' viewpoints and do not constitute any investment advice. Please strictly follow local laws and regulations. Audio transcription is completed by AI and may contain errors. Please listen to the full podcast:
Xiaoyuzhou:
https://www.xiaoyuzhoufm.com/episode/6af894a53caef3b52f55c0b5
Behind the controversy surrounding Jensen Huang, are there American political factors involved?
Cat Brother: Recently, controversies related to Jensen Huang and rumors about Nvidia's management changes have sparked discussions. Is this matter related to the political changes after the midterm elections in the U.S.?
Didier: I believe there may be political factors. After the shift in the U.S. political landscape, the Democratic Party may push for a re-evaluation of the Trump administration's policies regarding AI chip exports, cross-border computing power, and related regulations. Nvidia is at the core of the global AI industry chain, making it naturally a focus of attention.
I even speculate that the personal controversies may relate to Jensen Huang's desire to stay away from political whirlwinds, but this is purely personal speculation without direct evidence. If investigations arise in the future targeting chip flow, export compliance, or management, the impact could ripple through the entire AI industry chain.
Minta: Compared to the rumors themselves, I'm more focused on the potential industry impact. Jensen Huang is deeply involved in Nvidia's technology direction and business decisions. If there really is a significant change in management, the pace of advancing next-generation GPUs, the Rubin platform, and CPO (co-packaged optics) technologies could be affected.
At the same time, if the U.S. tightens approvals for data centers and chip export restrictions, previously aggressive computing power construction plans may also be delayed. In the short term, we mainly look at market liquidity, while in the medium to long term, we need to observe whether these changes genuinely impact corporate operations.
With the simultaneous decline of AI and Crypto, has the hardware bull market ended?
Cat Brother: Recently, AI stocks, the U.S. tech sector, and crypto have declined simultaneously. What is the main reason for this?
Minta: I think it cannot be simply attributed to the rumors about Nvidia. More importantly, the market is beginning to question whether the massive capital expenditures on AI can be sustained, and whether future profits will be sufficient to cover financing costs.
Some AI companies have seen widening credit default swap (CDS) spreads, reflecting heightened market sensitivity to credit risk. At the same time, inflation, interest rate expectations, and energy costs continue to affect liquidity. Previously, the rise in tech stocks attracted significant momentum and leveraged funds, and once the market weakens, concentrated deleveraging could amplify losses.
The uncertainty related to Nvidia further exacerbates market concerns about GPU, optical module, and data center construction delays. Crypto is also impacted by declining risk appetite, ETF fund outflows, and leverage liquidation, but these factors do not necessarily mean that the market will continue to fall.
Didier: My judgment is more pessimistic: this round of AI hardware bull market may have already ended. Over the past few years, the industry's capital expenditure expansion has been too rapid, but the power access, approvals, and construction progress of data centers may not keep pace.
If the data centers originally scheduled to be operational next year are delayed, GPU, storage, and optical module procurement will also decline accordingly, and the entire hardware industry chain may face demand revisions.
More importantly, Nvidia plays a role similar to that of a "central bank" in this ecosystem. It not only sells chips but also supports emerging cloud service providers (Neocloud) like CoreWeave and Nebius through equity investments, procurement agreements, and business cooperation, indirectly enhancing their financing capabilities.
These companies need continuous financing to purchase GPUs, construct data centers, and then recover costs through computing power leasing. Once Nvidia reduces investments or business support, the financing chain may come under pressure.
Taking CoreWeave as an example, based on my scenario analysis, even considering GPU mortgage financing and prepayments from customers, it may still face an equity financing gap of about 10 billion dollars by 2027, and in a pessimistic scenario, this could be close to 15 billion dollars. This is not confirmed data from the company, but my estimation.
If companies are forced to finance through ATMs (market price issuance), it could severely dilute existing shareholders' equity. Some emerging cloud service providers may even be unable to maintain expansion, ultimately leading to industry consolidation.
However, I do not believe the long-term trend of AI has ended. What may truly end is the first round of the investment cycle driven by aggressive hardware capital expenditures, and the next stage of value creation may come more from software and practical applications.
Bitcoin drops below key moving averages; has the long-term bull market logic changed?
Cat Brother: Previously, you were quite optimistic about Bitcoin. Now that BTC has dropped below key technical indicators, does the market cycle need to be reassessed?
Didier: I believe the short-term drop is mainly due to insufficient liquidity, adjustments in risk assets, and shifts in market narratives.
The crypto market often forms narratives around specific events, but what truly affects prices could be sustained capital flow. For example, Strategy's purchase of BTC through financing tools like STRC was one significant source of demand previously.
The current problem is weak liquidity at the end of the year, a lack of sufficient strong new buying, and rising risks in the AI industry chain, causing investors to overall reduce their risk exposure. Bitcoin's drop below the 200-day moving average further exacerbated market panic.
I remain optimistic about Bitcoin's long-term prospects, but I'm more cautious in the short term. If the AI hardware industry faces widespread financing difficulties and business clearances, risk assets may continue to be under pressure. However, extreme declines could also present new long-term investment opportunities.
After the slowdown in AI hardware investments, what opportunities remain in software and semiconductors?
Cat Brother: Does Minta agree that the AI hardware bull market has already ended? What directions are still worth focusing on?
Minta: I am generally cautious, but not as pessimistic as Didier. A decline in the growth rate of AI capital expenditures does not mean that all related companies will see an immediate drop in revenue and profits.
I am primarily focusing on three areas. First is AI software. If the growth of computing power supply slows, the cost of inference per unit token may not decrease as quickly as previously expected, and software companies that can improve computing power utilization and maintain stable customer demand may still benefit, such as Palantir.
Second is the semiconductor alternative supply chain. If China and the U.S. further tighten restrictions on GPUs, models, and key materials, some domestic alternatives and U.S. local supply chains may find opportunities, including materials like indium phosphide for optical communications, as well as semiconductor manufacturing and testing processes.
The third area is storage and computing efficiency optimization. Even if data center construction slows, companies will still need to improve the utilization of existing hardware. Areas like CXL (Compute Express Link), storage management software, and Pure Storage are worth exploring.
In the future, AI investments may shift from universally rising hardware sectors to more differentiated structural trends. These opportunities are more phase-specific and do not indicate that the entire industry will immediately enter a new round of comprehensive bull market.
How will changes in U.S. politics affect the AI industry and capital expenditure?
Cat Brother: If the political landscape in the U.S. changes, how will the development pace of the AI industry be affected?
Didier: I believe the long-term trend of AI remains unchanged, but the excessively aggressive capital expenditure model of the past few years may not be sustainable.
In the future, U.S. policies may pay more attention to balancing AI development with employment and social benefits. While AI improves productivity, it may also disrupt a large number of white-collar jobs, and the government needs to consider employment security, income distribution, and even universal basic income (UBI) issues.
If regulations tighten, chip exports, cross-border computing power, and data center construction may face more restrictions. However, this does not mean that the U.S. will abandon AI. On the contrary, if large cloud service providers spread their originally concentrated capital expenditures over a longer cycle, it could help improve cash flow.
I believe AI may undergo an industrial transformation similar to that after the internet bubble. In the early stages, a large amount of funding was invested in fiber optics and data centers, but what truly creates long-term value later are software and internet applications. The next stage of AI may also belong to general models, vertical AI agents, and software companies that can translate technology into actual revenue.
Changes in crypto regulation: which has more opportunities, RWA or compliant perpetual contracts?
Cat Brother: What impact might changes in U.S. regulatory direction have on Bitcoin, RWA, and perpetual contracts?
Didier: I think it's necessary to distinguish between different businesses. The long-term performance of Bitcoin does not entirely depend on whether a political party supports crypto; rather, the more important driving factors are the U.S. fiscal deficit, monetary policy, and global liquidity.
RWA and on-chain stocks are more dependent on the SEC's regulatory stance. If the U.S. intensifies scrutiny of securities tokenization, some projects may be affected, but the specific outcomes still depend on future legislation and regulatory rules.
In contrast, I am more optimistic about compliant perpetual contracts under the CFTC regulatory framework. There is already a huge demand for derivatives in traditional financial markets, while the crypto market's perpetual contracts have advantages in trading time, leverage management, and product design. If the U.S. allows compliant platforms to offer related products to domestic users, the market could further expand.
Furthermore, stablecoins, on-chain payments, and asset tokenization could also become important financial infrastructures for future AI agents and the machine economy.
Thus, I believe the long-term direction of crypto has not changed, but the regulatory environments and development opportunities faced by different sectors may significantly diverge.
How can on-chain stocks and derivatives bring new capital to crypto?
Cat Brother: What is Minta's view on the next round of growth opportunities for crypto?
Minta: I am more focused on real product demand rather than mere policy benefits. Whether it's RWA, stablecoins, or on-chain stocks, the ultimate question is whether they can create new use cases to attract real users and capital.
I am particularly interested in on-chain stocks and derivatives. Some platforms already support users trading assets related to U.S. stocks and provide additional yields through liquidity incentives. This creates certain differences from traditional securities accounts, but whether these related yields can be sustained in the long term still needs observation.
Platforms in the Solana ecosystem, such as Raydium and Meteora, are also exploring RWA and stock-related trading to improve liquidity and user experience. The Hyperliquid ecosystem is expanding perpetual contracts for traditional assets like stocks, gold, and silver.
These products might offer users more flexible trading times, leverage tools, and cross-asset trading experiences, thus attracting users from the traditional financial market.
However, I believe that what truly determines whether crypto can continue to grow is not policy expectations or short-term token incentives, but whether users are willing to continue using the products after subsidies end. Only when on-chain stocks, derivatives, and yield products create sustainable real demand can the industry gain long-term growth momentum.
免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到[email protected],本平台相关工作人员将会进行核查。