Odaily News: On September 14, according to SoSoValue data, influenced by the escalating situation in the Middle East and continued developments surrounding AI safety and other factors, the crypto market overall trended downward. The Meme sector led losses with a 24-hour drop of 3.57%, with Cash Cat (CASHCAT) falling 10.36%, Pons (PONS) dropping 14.25%, and Niu Lai plunging 27.82%. Meanwhile, Bitcoin (BTC) declined 0.55% and Ethereum (ETH) fell 1.65%, respectively holding above $76,000 and $2,400.

In addition, the DePIN sector stood out with a 24-hour gain of 3.49%, among which Filecoin (FIL) rose 21.43% and Arweave (AR) climbed 9.41%; the Layer2 sector gained 0.24%, with Lisk (LSK) surging 56.89% within the sector.

In other sectors, the CeFi sector dropped 1.35% over 24 hours, with Gate (GT) falling 2.45%; the Layer1 sector declined 1.86%, with Zcash (ZEC) dropping 4.05%; the DeFi sector fell 1.95%, with Jupiter (JUP) declining 2.61%; the PayFi sector dropped 2.35%, with Monero (XMR) falling 4.11%.

Crypto sector indices reflecting historical sector performance show that the ssiDePIN, ssiAI, and ssiMeme indices rose 5.52%, fell 3.28%, and fell 3.04%, respectively.

Odaily reports: "White-Haired Stock Guru" Serenity posted on X platform, stating that in their view, the probability of Anthropic and OpenAI genuinely slowing down frontier AI development is "essentially zero," and that some regulatory measures pushed by leading AI companies are more likely aimed at raising competitive barriers to restrict the development of competitors.

They stated that OpenAI and Anthropic will not voluntarily cede their lead to competitors like Gemini, xAI, and Meta, and that given all countries are continuously advancing AI technology development, it would be very difficult for the United States to meaningfully slow down frontier AI R&D. Currently, all parties are still competing for scarce resources such as storage and AI chips. They believe that demand for AI-related infrastructure will remain strong, and rebutted the view that AI-related stocks such as Nvidia (NVDA), TSMC (TSM), and Micron (MU) will experience significant declines as a result.

Previously, Anthropic CEO Dario Amodei published a long article stating that the AI industry must slow down the pace of improving model capabilities, but this does not mean stopping model training or technological progress — rather, it means ensuring companies have enough time for model alignment and safety safeguards. OpenAI founder Sam Altman expressed agreement, saying "we need to control the progress of frontier models," and stated that OpenAI will also consider giving independent evaluators access similar to that of employees.

Odaily reports, according to Gate data, South Korea's KOSPI index opened down 3.14%, SK Hynix fell 5%, and Samsung Electronics fell 3.6%.

Odaily News: Arthur Hayes posted on X stating that beyond AI safety risks, if companies such as Anthropic, OpenAI, and SpaceX no longer serve as compute buyers, the U.S. government may need to step in and take on that role; alternatively, if the disappearance of compute demand leads to defaults on related AI debt, the Federal Reserve may need to "print money" to bail out the insurance institutions holding that debt. Hayes said that whichever option is chosen, the situation will ultimately have to be resolved through currency issuance, and he joked about the predicament with the word "Yachtzee."

Odaily Reports - According to MSX.COM data, most of the U.S. AI sector fell in pre-market trading, with Nvidia dropping to $214.34, AMD dropping to $501.81, and Intel dropping to $99.33; Anthropic's pre-IPO over-the-counter contracts saw a 24-hour decline of 3.75% in corresponding market capitalization to $2.0783 trillion, while OpenAI's corresponding market capitalization fell 10.6% to $1.4391 trillion.

Previously, Anthropic and OpenAI rarely synchronized to "hit the brakes." Anthropic CEO Dario Amodei published a lengthy article stating that the AI industry must slow down the pace of improving model capabilities, but this does not mean stopping model training or technological progress—rather, it should ensure that companies have enough time for model alignment and safety protection. OpenAI founder Sam Altman agreed, saying "we need to control the progress of frontier models," and stated that OpenAI will also consider giving independent evaluators access similar to that of employees.

Odaily reports: Meta Platforms (META.O) Chief Artificial Intelligence Officer Alexandr Wang stated that "AI benefiting everyone" means people can trust powerful AI to reliably operate toward their own goals without producing unwanted side effects. Alignment is at the core of fulfilling this promise. We are at a critical moment for the industry, with capabilities growing rapidly, and we must make rapid progress on alignment to keep pace. The positive future we seek, one in which AI truly benefits everyone, requires the entire industry to take prudent and thoughtful action during this critical period. (Jinshi)

Previously, Anthropic CEO Dario Amodei published a long blog post stating that the company would adopt new safety measures, such as introducing third-party evaluation institutions, while also calling on the entire industry to support broader measures to slow the development of AI models. OpenAI CEO Altman quickly said he would adopt Amodei's proposal to "give independent evaluation institutions access similar to that of employees." Elon Musk, who runs xAI, also said: "Dario is right."

Odaily News: Anthropic CEO Dario Amodei published a lengthy blog post on Saturday, stating that the company will adopt new safety measures, such as bringing in third-party evaluators, while calling on the entire industry to support broader efforts to slow down AI model development. OpenAI CEO Altman quickly responded, saying he would adopt Amodei's suggestion to "give independent evaluation institutions access similar to that of employees." Elon Musk, who runs xAI, also said: "Dario is right." Although all three leaders have warned about the risks of artificial intelligence for years, there is almost no precedent for truly coordinated slowdowns in development in this fiercely competitive industry. Companies have long been continuously launching new products to boost user engagement, capture market share, and drive sales growth. It remains unclear to what extent these leading AI companies will actually implement new restrictions or safety checks. (Jinshi)