News Square
Global Crypto News · Real-time Aggregation
📚 Deep Analysis
More →Ken Griffin Says AI Is 'Almost Certainly' in a Bubble—And Compares GPUs to Airline Seats
Ken Griffin, founder of the Wall Street giant Citadel, <a href="https://www.marketwatch.com/story/ai-stocks-are-probably-in-a-bubble-but-might-be-worth-holding-onto-says-citadels-ken-griffin-87f535ce" target="_blank" class="sc-adb616fe-0 dJryYS" rel="nofollow">said</a> Tuesday that investors are “almost certainly” in an AI bubble—though he isn’t telling them to dump their AI stocks. “What’s left in the wake of these bubbles, if we want to use that word, is often game changing for the world. And the companies that are on the other side of that bubble are the market leaders for decades to come,” Griffin said at Citadel’s Future of Global Markets Conference in New York on Oct. 20 This reaches beyond tech investors, since the AI buildout has <a href="https://decrypt.co/359175/nvidia-earnings-steady-markets-ai-spending-debate-intensifies" target="_blank" class="sc-adb616fe-0 dJryYS" rel="nofollow">underpinned equity markets</a> over the past year, so many index funds and retirement accounts hold a slice of it. A bubble is what happens when prices climb far faster than the businesses behind them can justify, until reality catches up. Citadel is a hedge fund, a private firm that invests money for wealthy clients and institutions. But the fact that it seems like a bubble isn’t <i>that</i> big a deal if you look at the big picture, according to Griffin. “Any radical transformation of our economy is generally intertwined with a bubble,” he said, recalling the internet boom and adding that he wishes he had kept his Amazon stock. The companies that emerge on the other side of a bubble tend to lead markets for decades, he argued, though he stopped short of predicting a collapse. His own firm is a case study: Griffin said AI agents, software that carries out multistep tasks on its own, cut some high-level Citadel research from weeks to a couple of hours. Griffin compared GPUs, the graphics chips that many companies use to train and run AI models, to airline seats. Both have a very low marginal cost to run, meaning the extra expense of serving one more customer is tiny, so prices can collapse once supply outstrips demand. He pointed to the telecom boom around 2000, when companies laid more fiber-optic cable than demand required and communication prices fell. He also doubted that chips five or six years old, which still earn rental rates he called extraordinary, will keep doing so for another five or six years. Buyers are getting pickier, Griffin said, sorting which jobs justify premium models from OpenAI and Anthropic and which can run on cheaper open-source alternatives, meaning models anyone can download and run on their own computers. Corporate America has been uneven at extracting value from AI, he added, with companies studying their bills from the labs and asking where the productivity is. Gartner forecasts worldwide<a href="https://www.gartner.com/en/newsroom/press-releases/2025-09-17-gartner-says-worldwide-ai-spending-will-total-1-point-5-trilli
Tesla's Bitcoin holdings remained unchanged in the third quarter, while the value of its digital assets rose sharply.
CoinWorld data: According to Benzinga, Tesla kept its Bitcoin ($BTC) holdings unchanged in the third quarter, with the value of its digital assets surging to $962 million, up 42% from the previous quarter. The company also recorded $288 million in unrealized gains, ending three consecutive quarters of paper losses. Arkham Intelligence estimates that Tesla still holds 11,509 bitcoins, with the value matching its balance sheet. Tesla is the 11th-largest corporate holder of Bitcoin, ahead of companies such as Block and Riot Platforms. SpaceX ($SPCX) also kept its Bitcoin holdings unchanged in the third quarter. Tesla reported third-quarter revenue of $28.05 billion and adjusted earnings per share of 48 cents, beating market expectations of 43 cents. Tesla shares fell 3.7% in after-hours trading.
NVIDIA launches Minerva, an AI factory operations platform that can analyze billions of data points per hour.
Nvidia launched its AI operations platform Minerva on October 22, aiming to detect failures in advance, increase GPU uptime, and improve the output efficiency of AI factories by integrating infrastructure and AI workload telemetry data. The platform can analyze billions of data points per hour, identify anomalies, predict potential failures, and consolidate related alerts into actionable fault-handling items. Minerva also features a native AI assistant that can query cluster operating conditions through natural language and provide remediation suggestions. Nvidia said Iren is already using the platform, while Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI, and Tesla plan to deploy it. Dell, Lenovo, and Supermicro will also incorporate Minerva into their respective AI factory operations product portfolios.
Bitcoin ETFs attract $470 million in net inflows in Lookonchain update
CoinWorld data: As of October 22, U.S. spot Bitcoin ETFs attracted $470.15 million in net inflows, with single-day inflows of 5,766 BTC and total seven-day inflows reaching 11,503 BTC, valued at approximately $938 million. Meanwhile, Ethereum products saw net outflows of 4,422 ETH, with total seven-day outflows of 120,629 ETH, valued at approximately $290 million.
The Philadelphia Semiconductor Index fell 6.06% intraday.
According to CoinWorld news, the Philadelphia Semiconductor Index fell 6.06% during the day. Among individual stocks, Nvidia fell 3.95%, Broadcom fell 4.25%, Micron fell 4.39%, Intel fell 6.50%, AMD fell 7.64%, Qualcomm fell 2.35%, TSMC fell 4.19%, and ASML fell 5.30%.
2017年3,000枚ETH转入Kraken，9年半升值128倍
Data from Bijie: On-chain analyst Yu Jin reports that 3,000 ETH withdrawn from Gemini in 2017, after lying dormant for nine and a half years, was transferred to Kraken 10 minutes ago. When withdrawn in 2017, ETH was priced at only $19, and these 3,000 ETH were worth about $56,000 at the time. When transferred to Kraken today, they were worth about $7.23 million, a 128-fold appreciation over nine and a half years.
Blackstone President Jon Gray rebuts the AI bubble theory, saying demand is growing parabolically.
According to CoinWorld news, Blackstone President Jon Gray rebutted billionaire Ken Griffin's claim that there is a bubble in artificial intelligence. He said: "When it comes to bubbles, what you think of is building capacity first and then hoping demand appears. But in the AI field, you are building capacity while hoping you can keep up with demand." Gray believes this is "a completely different situation" from phenomena such as the telecommunications bubble in the 1990s. He mentioned ODE, the AI consulting company jointly established by Blackstone and Anthropic, and said: "We simply cannot keep up with client demand; demand growth is almost parabolic."
Perplexity improves AI self-distillation: train the teacher, not just the student
Perplexity proposes a new method in AI self-distillation training that trains not only the student but also the teacher. In this process, the same model plays both the roles of student and teacher: the student solves problems independently, while the teacher can see the reference answer in advance and provide token-by-token guidance. The two roles share parameters, eliminating the need to run an additional larger teacher model, and yield more fine-grained training signals. The SDPG method proposed by teams including UCLA retains the final-outcome reward and allows only those solution processes that perform above the group average to receive distillation guidance, while gradually reducing the distillation weight in the later stages of training to avoid excessive imitation hindering model exploration. Combined with the JOLT method newly proposed by Perplexity and NVIDIA, the teacher uses additional information to solve problems independently and obtain rewards, while KL divergence is used to constrain the output difference between the teacher and the student. JOLT achieves an accuracy of 72.14% on the math test with Qwen3-0.6b, higher than the 61.20% of the previous self-distillation method OPSD.
An unknown wallet transferred 780 BTC, worth approximately $63.85 million.
CoinWorld data: According to Whale Alert monitoring, 780 BTC (approximately $63.85 million) has just been transferred out of an unknown wallet.
NVIDIA launches the Minerva platform, targeting end-to-end AI factories.
CoinWorld news: NVIDIA launches the Minerva platform for end-to-end AI factories.