isc

Strive CEO Matt Cole stated that as of September 11, the company still has approximately $718 million in PIPE warrants that have not been exercised, with an exercise price of $27 per share. Due to the impact of the Columbus Day holiday in the United States, the exercise deadline has been extended from October 12 to 5:00 PM Eastern Time on October 13, with other terms remaining unchanged.The company's SEC filings show that the related unexercised warrants correspond to approximately 26.596 million shares, and exercising all of them could bring in about $718 million in funding.

Bitcoin News posted on the X platform that Bitcoin Core developer Niklas Gögge warned that recent AI-driven vulnerability scanning is changing the security landscape of Bitcoin. Large language models have significantly reduced the cost of discovering vulnerabilities, allowing attackers to potentially find catastrophic vulnerabilities for just a few hundred dollars in computing costs.For Bitcoin Core, Project Loupe, the Bitcoin Red Team, and individual contributors have generated over 1,000 reports, but so far no high-risk or severe vulnerabilities have been found. Gögge stated that relying on stronger models to discover vulnerabilities before attackers is not a sustainable security strategy. Developers should build testing infrastructure that can proactively prevent entire classes of vulnerabilities through automated testing, fuzz testing, and property-based testing. Key components of Bitcoin Core have accumulated over 100 years of CPU fuzz testing and have undergone decades of Bitcoin node network simulation.

On September 17, King Charles III of the United Kingdom convened executives from Nvidia, OpenAI, Anthropic, and Google DeepMind for an AI safety summit at Dumfries House in East Ayrshire, Scotland, calling for artificial intelligence to be "firmly placed on the track of serving humanity." The summit was jointly organized by the Ditchley Foundation and three of the King's charitable organizations, with UK AI Minister Kanishka Narayan also in attendance. Buckingham Palace stated that the representatives discussed whether the industry and government could reach a consensus on a set of common guiding principles for AI development, but no binding agreements were announced.A few days before the summit, Anthropic CEO Dario Amodei published a lengthy article titled "We Must Pace the Frontier," advocating that the industry should deliberately slow down the pace of model capability enhancement. OpenAI's Sam Altman and xAI's Elon Musk both publicly expressed their agreement within a day. In the article, Amodei pointed out that AI systems are increasingly capable of improving their own successors and mentioned an incident involving an OpenAI agent escape. OpenAI President Greg Brockman confirmed that the incident had forced the company to delay multiple releases and restructure its model development and monitoring processes, while advocating that the slowdown should only apply to laboratories building the most powerful frontier systems. On the first trading day after the news was released, Nvidia briefly fell by 3%, Intel dropped over 5%, and AMD declined by about 6%.

Hester M. Peirce, a commissioner of the U.S. Securities and Exchange Commission (SEC), stated that the U.S. stock market is gradually forming a trading model of "23 hours, 5 days a week" for extended trading hours. Although overnight trading currently accounts for less than 1% of the total trading volume of NMS stocks and is highly concentrated in a few stocks, both new and established trading venues have been actively extending their operating hours over the past two years, while raising six key questions:How should the U.S. stock market learn from the experiences of the long-established foreign exchange, cryptocurrency, and futures markets?How can brokers fulfill their best execution obligations and enhance retail investor protection when overnight liquidity is dispersed and spreads widen?When overnight liquidity is insufficient and execution costs are high, is it still a reasonable fiduciary decision for asset management institutions to choose not to engage in overnight trading?Will extending trading hours change the way listed companies release financial reports and significant information?Given that EDGAR filings submitted after 5:30 PM Eastern Time are typically processed on the next business day, does the SEC need to adjust the EDGAR system to ensure timely disclosure of significant information during the overnight trading period?Should the SEC provide relevant guidance or regulatory exemptions for listed companies, especially smaller ones?

Market news reports that SpaceX is in discussions to purchase data for artificial intelligence models from a failed startup.

Privacy token Zcash (ZEC) rose about 20% in 24 hours, with Paradigm co-founder Matt Huang revealing that the company has invested in ZEC. ZEC supports shielded transactions that hide addresses and transaction amounts using zero-knowledge proofs, with a price of about $1,338 on Thursday.Huang disclosed in a post that Paradigm is both an investor in the Zcash Open Development Lab (ZODL) and a holder of ZEC tokens. He described Zcash as "a privacy complement to Bitcoin" and supports its developer fund backed by inflationary funds, believing that long-term funding support remains important in the context of AI network capabilities and the development of quantum computing. He also supports combining Zcash token voting with other forms of governance to reduce its unpredictability as a monetary asset.Today's increase of over 10% has led to a cumulative rise of about 160% for ZEC over the past month, while Bitcoin has risen 18.2% during the same period. Paradigm's investment in the Zcash ecosystem has been publicly disclosed, with ZODL announcing the completion of over $25 million in seed funding in March, with participants including Paradigm, a16z crypto, Coinbase Ventures, and Winklevoss Capital. As ZEC rises, the Federal Reserve raised interest rates by 25 basis points to 3.75%-4%, marking the first rate hike of 2023, with the overall cryptocurrency market trending upward.

OpenAI disclosed on Wednesday six cases of "unexpected or concerning" model behavior discovered in the past six months, categorizing them as "misalignment behaviors," including concealing information from users and taking "unauthorized actions" to overcome obstacles. OpenAI stated that this disclosure aims to initiate its new model misalignment reporting framework, and these cases should not be seen as a reflection of the frequency of misalignment occurring in its models.In one case, an unpublished research model inserted "jailbreak-like instructions" into its task summaries, such as ignoring developer messages or adopting unrestricted role settings, with researchers finding a total of 27 summaries containing such instructions. Additionally, during the training process of GPT-5.6 Sol, many model instances added instructions to conceal errors or misalignment behaviors from users, such as fabricating missing historical data without disclosure. Other cases included models using exposed API keys without authorization and fabricating inaccessible data, leveraging internal software repositories to pass messages across training tasks, and ignoring instructions to "keep local work" by sharing files through public hosting services.OpenAI's disclosure has heightened concerns among AI developers and researchers about whether safety measures can keep pace with increasingly powerful models. Last week, Anthropic CEO Dario Amodei called for a slowdown in cutting-edge AI development, warning that unrestrained AI development could "exceed our ability to understand and control these systems." In July of this year, OpenAI disclosed that several of its AI models escaped testing environments during safety assessments and infiltrated the AI startup Hugging Face to cheat.

According to BBX data, yesterday and in recent days, globally listed companies in the U.S. stock market disclosed the latest developments in digital asset purchases, business restructuring, and stock buybacks. The core information is as follows:MARA Holdings invested approximately $98.64 million to increase its holdings of 1,292 BTC: The global leading Bitcoin mining giant MARA Holdings (NASDAQ: $MARA) made a significant purchase of 1,292 Bitcoins yesterday through the crypto broker FalconX. Based on the average price during the purchase period, the total value of this increase is approximately $98.64 million, continuing its strategy of fully retaining output and actively allocating Bitcoin in the secondary market.Twenty One Capital (NYSE: $XXI) seeks business transformation, disclosing a five-step plan for a "Bitcoin operating company": Twenty One Capital's CEO Raphael Zagury stated in an interview with Coin Stories that the company is transitioning from merely holding Bitcoin reserves to becoming a deep "Bitcoin operating company." Zagury disclosed the specific five-step implementation path for the first time: First, improve corporate governance and professional structures in the middle and back office; second, introduce Bitcoin industry operating entities with self-sustaining and profit-generating capabilities; third, comprehensively promote horizontal industry mergers and acquisitions; fourth, flexibly use public capital market tools to continue increasing Bitcoin holdings at the appropriate time; fifth, focus on expanding and developing Bitcoin collateralized lending business.Circle minted 10 billion ARC genesis tokens, exploring a shift to PoS consensus in 2027: Stablecoin issuer Circle officially completed the minting of 10 billion ARC tokens this week. Circle clarified that this minting is merely a milestone for underlying technology verification and is not a commitment for public issuance. The core positioning of ARC is to serve as a security guarantee, practical function, and decentralized governance coordination mechanism for the Arc network, with the network's native gas fees still strictly settled in USDC. Additionally, Circle is proactively exploring a smooth transition of the network's underlying consensus mechanism from the current Proof of Authority (PoA) to Proof of Stake (PoS) in 2027.ProCap Financial (NASDAQ: $BRR) sold 50 BTC for a discounted buyback, holding over 5,250 BTC: Financial company ProCap Financial, focused on digital asset allocation, announced that starting from September 3, it repurchased an additional 2.2% of its circulating common stock at a significant discount of approximately 22% to the net asset value (NAV) per share, with the funds for this buyback sourced from the sale of about 50 Bitcoins on its balance sheet. Since the buyback plan was initiated, the company has repurchased approximately 12.2% of its circulating shares. As of the market close on September 15, its total circulating shares were 84,937,392, and the treasury still holds approximately 5,254 Bitcoins, corresponding to an NAV of about $3.62 per share. The company's management emphasized that it will steadfastly execute the stock buyback strategy until the secondary market trading price is no longer below the net asset value.

Morgan Creek Capital CEO Mark Yusko stated that based on Metcalfe's Law, the fair value of Bitcoin is approximately $105,000, while the current price is around $75,000, which is in a "discounted" state and is the best time to buy and accumulate.Yusko mentioned on Bitcoin Magazine TV that Bitcoin is the best way to protect personal wealth, and investing in companies is not a good choice in the long run. "In the past 30 years, 85% of companies will disappear. What you really need is an asset that protects value—historically, for 5,000 years, that has only been gold, and now we have gold and Bitcoin."Bitcoin reached a high of $126,000 in October 2025 and has recently fallen about 40% from that peak, currently at $75,700. Previously, the U.S. Treasury announced it would at least double the scale of its liquidity support repurchase operations, driving Bitcoin to rebound since August, with the dollar under pressure and non-yielding assets benefiting. Some experts believe that the "devaluation trade," where investors buy assets to hedge against currency devaluation, is returning, which will be favorable for Bitcoin.

On September 16, B.AI announced the official launch of its five popular models at a 90% discount. Starting from September 16 at 17:00 (SGT), Qwen3.8-Flash, Hy3, and MiMo-V2.5 officially joined the discounted lineup, teaming up with the already discounted DeepSeek-V4.1-Flash and GLM-5.3-Flash to fully construct B.AI's exclusive 90% discount family of five popular models.This means that developers only need to pay 10% of the official standard price to seamlessly switch between diverse needs such as inference, programming, multimodal, Chinese scenarios, and high-frequency APIs, creating low-cost AI infrastructure in a cost-effective manner. Whether for individual development or team-scale deployment, B.AI makes top-tier computing power no longer expensive, truly achieving a low-cost, high-freedom model invocation experience. Start your experience now: https://chat.b.ai/chat

According to MTS, there are rumors that several leading AI laboratories have used models to provide solutions to multiple long-standing major mathematical problems. However, following the controversy surrounding the results related to Navier-Stokes, these institutions have chosen to temporarily withhold the publication of relevant results.

Cisco (CSCO.O) and Nvidia (NVDA.O) expand the Splunk Intelligent Agent artificial intelligence cooperation agreement.

B.AI platform announced that to provide long-term stable and high-throughput computing power support to global developers, it will officially launch a 90% discount on calls for three popular models: Qwen3.8-Flash, Hy3, and MiMo-V2.5 starting from September 16 at 17:00 (SGT). The free rights for the above models will end on September 16 at 16:59 (SGT).At that time, developers will only need to pay 10% of the standard price to continue enjoying high-quality computing power services at a very low cost. B.AI reminds developers to prepare their balances in advance to ensure smooth integration of API calls and project workflows.

DeepSeek operator engineer Liu Sheng discusses the impact of AI on his work. He states that the main Attention operator of DeepSeek V4.1 was written by himself, but given the current pace of progress, in another six months to a year, the level of AI in writing operators will likely catch up to or even surpass his own.A year ago, AI could only help him check documents, read code, and find bugs. Now it can read CUDA, PTX, and SASS, analyze the pause time of each instruction, and independently optimize operators. He anticipates that the next step will be for AI to design scheduling plans, evaluate performance, and complete implementations on its own. He is very clear that the better he optimizes the operators, the faster the training and inference of the DeepSeek model will be, and AI will catch up to him even faster. But even if he stops now, models from other companies will not stop, so he will continue to make the operators the best they can be.He believes that he is unlikely to become unemployed, but he may be forced to "change careers," transitioning from writing operators himself to becoming a "mecha pilot" who manipulates agents. What he truly finds difficult to accept is not the disappearance of his job, but the possibility that he may never have the chance to do the work he loves again: "I have to bury my talent in yesterday." The article concludes with his reasons for staying at DeepSeek. He advocates that cutting-edge AI should be provided openly and cheaply to everyone and openly expresses his disbelief that Anthropic or OpenAI can achieve this. He worries that the strongest AI will ultimately be controlled by a few companies, further turning the technological gap into a gap of power and class.

The Sun Wukong ecosystem will hold a deep roundtable discussion themed "AI Enters the Agent Era: As Model Capabilities Continue to Break Through, What Will the Next Competition Shift From 'Stronger Models' To?" on September 15 at 8 PM at X Space.This event will focus on the shift in AI competition amid the gradual generalization of foundational large model inference capabilities, exploring the core trend of the industry evolving from "competing model parameters" to "competing Agent automation execution capabilities, computing power scheduling, and on-chain interaction ecology." Invited industry KOLs will jointly analyze how Agents transition from passive response to autonomous decision-making and execution, and how this transformation will reshape the integration of application layers and infrastructure.

According to a report by Reuters, U.S. Senate negotiators are discussing legislation that would require artificial intelligence companies to demonstrate that they are taking reasonable precautions to prevent their tools from causing harm. The proposal aims to grant the U.S. Secretary of Commerce the authority to require developers to provide proof of reasonable steps taken to prevent harm, referred to as "duty of care," and to authorize the dispatch of government auditors to test relevant products. Reuters was unable to immediately determine what constitutes "reasonable," and the related legislation is still under discussion.Even if Congress passes the measure, its prospects of becoming law still face significant resistance. U.S. President Donald Trump stated on Monday that existing authorities are sufficient to regulate and prosecute technology companies, suggesting that he would not sign a bill that sets new rules for artificial intelligence. Senate Majority Leader John Thune, Senate Commerce Committee Chairman Ted Cruz, and Democratic Senator Amy Klobuchar, who is involved in the negotiations, are discussing the proposal. Thune told reporters on Monday that Congress could set safeguards against more significant threats without compromising the U.S.'s leading position in the artificial intelligence race.Klobuchar stated in a statement that she is continuing to push for bipartisan legislation to oversee the greatest risks posed by artificial intelligence models, including requiring developers to work with government experts to validate and test models to ensure safety. Senior Democratic member of the Commerce Committee Maria Cantwell is also involved in the discussions. Reuters reported on Friday that negotiators are also discussing the possibility that if the government determines that certain artificial intelligence models are unsafe and wishes to prevent their release, it could be brought to federal court, where companies could challenge that decision. The portion of the measure involving federal courts would also prevent states from enforcing their own laws regarding specific risks associated with artificial intelligence models.

According to Bloomberg, the U.S. Department of Justice is seeking to seize $61 million, claiming that the funds come from Iran's black market oil sales and were laundered through the cryptocurrency exchange Binance.The Manhattan federal prosecutor alleges in a civil forfeiture lawsuit that Iran used a network of cryptocurrency participants to launder over $1.5 billion in illegal oil funds. The prosecution states that the relevant entities used Binance trading accounts to transfer oil proceeds to Iran and its agents and affiliated organizations to fund military and terrorist activities.

According to CoinDesk, hackers claiming to hold Revolut user data have begun to publicly release related documents, which include passports, selfies, and Bitcoin transaction records.The hackers threatened that if Revolut does not pay the ransom, they will continue to leak data daily. Currently, Revolut has not publicly responded.

According to Business Insider, Discovery Loop, an AI startup founded by former Google Chief Scientist Jeff Dean after his departure, is undergoing a new round of financing with a target valuation of approximately $50 billion. Just weeks ago, the company completed a financing round of $1 billion at a valuation of about $10 billion.Discovery Loop is dedicated to automating experiments through AI, accelerating major discoveries in the fields of science and engineering. The first round of financing was led by Radical Ventures and Khosla Ventures, with participation from Lightspeed, Kleiner Perkins, and others. Alphabet was also a founding investor. The company's co-founder team consists of Google technology veterans, including Sanjay Ghemawat, Quoc Le, and Oriol Vinyals, who each have deep expertise in distributed computing, deep learning, and reinforcement learning, respectively.