According to BlockBeats, on September 14th, TradingBeats monitoring revealed that Smart Money TestingThingsOut recently placed HYPE and ZEC on one end of its trading range, while BTC and ETH were on the other. The former two combined held approximately $20.894 million in long positions, and the latter two combined held approximately $20.906 million in short positions. The nominal size of the long and short positions was almost equal, totaling over $40 million.

The account position shows a logic biased towards relative strength trading, using short positions in BTC and ETH to hedge against market volatility, and betting that HYPE and ZEC, which have performed well recently, will continue to outperform.

This round of position building began on the evening of September 11th. TestingThingsOut bought 170,200 HYPE tokens, opening a position of approximately $14.005 million; ZEC saw a net long position of 6,500 tokens. Simultaneously, 140 BTC were newly established as short positions, and the ETH short position was expanded to 4,000 tokens.

As of press time, the four positions are as follows:

- HYPE long position: valued at approximately $13.631 million, with a floating loss of approximately $374,000;

- ZEC long position: valued at approximately $7.263 million, with a floating loss of approximately $415,000;

- BTC short position: valued at approximately $10.857 million, with a floating profit of approximately $109,000;

- ETH short position: valued at approximately $10.049 million, with a floating profit of approximately $142,000.

The four positions totaled approximately $41.8 million, with a net unrealized loss of approximately $538,000. This bet has not yet been cashed out. Starting from around the evening of September 11th, BTC appreciated by approximately 0.57% relative to HYPE; during the same period, BTC appreciated by approximately 0.41% relative to ZEC.

If we calculate from the end of the position building in the early morning of September 12, the HYPE/BTC price ratio fell by about 0.80%, and the ZEC/BTC price ratio fell by about 3.44%, with ZEC showing a more obvious relative weakness.

In addition, its Polymarket-linked address still holds approximately 9,500 "HYPE fell to $20 this year" No., with an implied probability of approximately 95.5% for this No., and no new transactions have been added in the past three days.

According to BlockBeats, on September 14th, Serenity published an article stating that it is highly unlikely that leading AI labs like OpenAI and Anthropic would intentionally slow down their development, and that most of the "regulation" they attempt to implement is regulatory capture aimed at slowing down competition. Public pronouncements about "slowing down development" are largely superficial.

Serenity also criticized the claims that AI stocks on the X platform (such as NVDA and TSM) are "about to collapse," emphasizing that the memory and chip supply chains will remain tight until 2031, indicating that the competition for resources in AI development is still accelerating.

Beating AI News Flash: Following Trump's refusal to slow down AI development citing competition from China, OpenAI CEO Sam Altman subsequently published a lengthy article reiterating the need to "apply the brakes." He stated that slowing down does not mean halting research and development, but rather allowing models to progress more slowly, giving time for security and surveillance. "No matter how much competitive pressure the United States faces, it cannot be a justification for reckless development."

OpenAI also plans to move security reviews further upstream. Previously, the security framework primarily focused on model releases; now, companies will write clear "security statements" before any cutting-edge reinforcement learning training, which could significantly enhance capabilities, begins, identifying risks and protective measures. Altman also supports nationally unified standards for cutting-edge AI security and independent reviews, but believes that companies should start implementing these measures now, without waiting for congressional legislation.

Altman also pointed out two worst-case scenarios. One is that humanity completely loses control over AI, and the other is that a super AI is monopolized by one person, one company, one laboratory, or even one country. He believes that both scenarios must be avoided, and AI must ultimately always obey humanity.

BlockBeats reported on September 14th that the GoPlus Chinese community issued a security alert stating that on September 12th, a newly activated EOA address launched a proposal attack against Ampleforth, demanding the transfer of 2.5 million USDC (almost all of the vault's liquid funds) to the proposer under the pretext of "applying for completed work funding for the SPOT ecosystem analysis tool." As of press time, the proposal remains pending, with no votes cast and the funds not yet transferred.

GoPlus points out that the threshold for a proposal is 75,000 FORTH, and the threshold for approval is 600,000 FORTH. Based on the price of $0.27 at the time of the alert, approximately $160,000 would be needed to reach the required number of votes. The proposer/receiving address currently holds approximately 87,200 FORTH voting rights; the proxy address also holds approximately 87,200 FORTH and delegated its voting rights to the proposer 19 minutes before the proposal was submitted.