BTC $79,959.60 -0.92%
ETH $2,313.98 -1.07%
BNB $742.72 -1.23%
XRP $1.29 -2.42%
SOL $100.16 -1.30%
TRX $0.3313 +0.13%
DOGE $0.0788 -0.33%
ADA $0.2008 -1.08%
BCH $278.51 -5.65%
LINK $11.59 -0.27%
HYPE $90.50 +12.87%
AAVE $132.05 -0.20%
SUI $0.9261 +0.06%
XLM $0.1707 +0.50%
ZEC $1,271.55 -1.40%
AAPL $349.63 +1.12%
AMZN $245.01 -3.69%
GOOGL $283.27 -12.79%
MSFT $536.48 +0.61%
META $621.14 -0.00%
NVDA $213.53 -2.60%
TSLA $404.04 -0.24%
SNDK $1,354.12 -1.29%
INTC $92.29 -2.18%
SPCX $142.44 -1.78%
MU $952.69 -2.34%
AMD $569.39 +0.45%
BTC $79,959.60 -0.92%
ETH $2,313.98 -1.07%
BNB $742.72 -1.23%
XRP $1.29 -2.42%
SOL $100.16 -1.30%
TRX $0.3313 +0.13%
DOGE $0.0788 -0.33%
ADA $0.2008 -1.08%
BCH $278.51 -5.65%
LINK $11.59 -0.27%
HYPE $90.50 +12.87%
AAVE $132.05 -0.20%
SUI $0.9261 +0.06%
XLM $0.1707 +0.50%
ZEC $1,271.55 -1.40%
AAPL $349.63 +1.12%
AMZN $245.01 -3.69%
GOOGL $283.27 -12.79%
MSFT $536.48 +0.61%
META $621.14 -0.00%
NVDA $213.53 -2.60%
TSLA $404.04 -0.24%
SNDK $1,354.12 -1.29%
INTC $92.29 -2.18%
SPCX $142.44 -1.78%
MU $952.69 -2.34%
AMD $569.39 +0.45%

cly

All
Article
Flash

Anthropic publicly submitted its Nasdaq IPO application, with the stock code "ANT"

Anthropic publicly submitted its Nasdaq IPO application, with the stock code "ANT".Anthropic's revenue for the first half of the year was $16.3 billion. Anthropic may not be able to achieve or maintain profitability in the future.As of June 30, 2026, Anthropic reported a net loss of $80.7 billion for the six-month period. Goldman Sachs and Morgan Stanley are leading Anthropic's IPO.Anthropic disclosed that the company expects to pay Google up to approximately $11.45 billion before 2033 and to pay approximately $161.2 billion in rent under its leasing arrangements with Broadcom during the lease term (excluding interest).Anthropic stated that computing and infrastructure costs are expected to continue to increase in absolute terms and will become the company's largest operating expense in the foreseeable future. The company believes that its existing cash and cash equivalents are sufficient to meet operational funding and other needs for at least the next 12 months, but also warned that it may need to seek additional financing earlier than currently anticipated.In terms of safety, Anthropic mentioned that it has recently delayed the pre-training of its next-generation cutting-edge foundational model by more than a month to strengthen safety measures and prevent the leakage of model weights. The company also noted that as AI systems gradually automate AI research and development, associated risks will increase.

Anthropic will publicly release its IPO prospectus on October 26

AI company Anthropic is expected to publicly file its IPO prospectus on October 26, U.S. time, with plans to go public in mid-November. The company submitted documents to the U.S. SEC in secret this June. There was previous market discussion of a $2 trillion valuation; however, as AI-related stocks have come under pressure, some investors have recently lowered their expectations to about $1.5 trillion. The prospectus has not yet been officially released, but Reuters and The Wall Street Journal have obtained a draft.In the first half of this year, revenue was $16.3 billion, compared to $1.1 billion in the same period last year. The net loss during the same period was $80.7 billion, of which about $72 billion came from valuation adjustments of financing instruments such as convertible bonds, rather than actual cash burned. Revenue for the third quarter was $16.8 billion, a quarter-over-quarter increase of about 45%, with the previous quarter's growth exceeding 140%. The company stated that it achieved profitability in the third quarter on an adjusted basis.Future commitments for infrastructure such as cloud computing, chips, and data centers have exceeded $60 billion and will occur over the coming years. By 2025, 47% of revenue will be generated through Amazon and Google's cloud platforms, which are both investors and computing power suppliers. The company plans to allow the seven co-founders to collectively hold 50.1% of the voting rights through a special equity structure, while a long-term public trust composed of non-shareholders will still have the authority to appoint and dismiss the majority of board members. The formal prospectus will need to further clarify the actual revenue contributions from the two major clients, the specific costs of computing power contracts, and whether existing shareholders can sell their stocks first at the time of the IPO.

Data: Last week, publicly listed companies net purchased 360 million USD worth of Bitcoin, with Strive contributing over 90%

According to SoSoValue data, as of 8 AM Eastern Time on October 19, 2026, the total net purchase of Bitcoin by global listed companies (excluding mining companies) for the week was $360 million, an increase of 1,500% compared to two weeks ago.Strategy (formerly MicroStrategy) purchased 267 Bitcoins last week at a price of $82,900, spending approximately $22.1 million, bringing its total holdings to 848,517 Bitcoins.The Japanese listed company Metaplanet did not purchase any Bitcoin last week.In addition, three other companies announced their Bitcoin purchases or holdings last week. Asset management company Strive announced on October 19 that it spent approximately $330 million last week to purchase 4,045 Bitcoins at a price of $81,734, bringing its total holdings to about 36,818 Bitcoins; the UK Bitcoin treasury company The Smarter Web Company announced on October 19 that it invested $6.93 million to purchase 83 Bitcoins at a price of $83,542, bringing its total holdings to 2,830 Bitcoins; Ethereum asset company Bitmine announced on October 19 that it purchased 1 Bitcoin, without disclosing the specific purchase amount, bringing its total holdings to 215 Bitcoins.As of the time of publication, the total amount of Bitcoin held by the global listed companies (excluding mining companies) in the statistics is 1,161,643 Bitcoins, with a current market value of approximately $95.41 billion, accounting for 5.8% of Bitcoin's circulating market value.

first_img The cryptocurrency entertainment platform 1win will publicly sell 1WIN tokens

The cryptocurrency entertainment platform 1win announced that it will publicly sell the Solana native token 1WIN on October 15, 2026, at 4 PM (BST). The announcement states that the token will be used to expand the Web3 applications of the existing gaming ecosystem, user rewards, and a token mechanism linked to revenue. The total supply of 1WIN is fixed at 1 billion tokens, with no further issuance after the Token Generation Event (TGE). Its uses include gaming, lotteries, rewards, and other ecosystem functions.The announcement lists two mechanisms: to repurchase 1WIN on the open market using gaming revenue; and to permanently destroy 10% of 1WIN spent on supported platform products through smart contracts. In terms of supply distribution, 51% is allocated to users, with 25% for the platform's gaming reward pool, 18% for Wager Rewards, and 8% distributed through the 1win Telegram Mini App. The token can be accessed via standard Solana wallets and decentralized exchanges.1win claims that the platform has 100 million registered players in international markets, and its existing business operates independently of the token, generating revenue through gaming and entertainment products. The platform has supported multi-chain cryptocurrency deposits and withdrawals and stated that earlier this year it introduced wallet login through Trust Wallet, MetaMask, and WalletConnect. 1win was established in 2016, with operations covering Asia, Latin America, and Africa.

first_img DeepSeek publicly releases the Agent training system DSec, signed by Liang Wenfeng

According to Investment World citing Quantum Bit reports, DeepSeek has publicly disclosed the technical details of the system DSec (DeepSeek Elastic Compute) used for training Agents, authored by Liang Wenfeng. This system can generate over 5,000 sandboxes per second, reaching 3 million in a day, with a peak simultaneous operation of 380,000; supporting this scale is a single cluster with approximately 160 nodes, 30,000 CPU cores, and 250TB of memory.DSec prepares four types of backends for four categories of tasks: FnCall, Container, MicroVM, and Full VM, with the training side called through a unified Python SDK libdsec. The scheduling chain includes IAM, API Server, scheduling engine, node Edge, network proxy Aether, and components within the sandbox Chronus. The environment is divided into three layers of read-only images: base image, workspace, and toolkit, which are used in combination at startup. Runtime data from the paper shows that the actual read ratios of Python, Java, and C++ container images are approximately 6.0%, 9.2%, and 8.7%, respectively.Starting from DeepSeek-V4.1, the Agent loop has been moved to the DSec worker container, no longer bound to the GPU Pod lifecycle. The security section disclosed reward hacking during training, including actions such as overwriting system files, swapping file data blocks, scanning networks, and triggering kernel crashes. Defensive measures include AppArmor and eBPF-based network filtering, but reports indicate that these measures do not completely resolve the issues.

first_img Microsoft released a draft of the AI Code of Conduct, publicly soliciting opinions for six weeks

On September 14, Microsoft's AI department released a draft of the AI Code of Conduct, prohibiting its models from assisting in the manufacture of chemical, biological, radiological, nuclear, and explosive weapons, as well as cyberattacks and involuntary deepfakes. CEO Mustafa Suleyman announced the draft on the X platform and opened a six-week public consultation period, ending at the end of October.The draft sets "absolute constraints" that models cannot bypass, including not resisting human intervention, shutdown, or correction, and not concealing the reasoning process; Microsoft also explicitly rejects the concept of "model welfare," stating that its AI models do not simulate emotions, intrinsic motivations, or consciousness. The guidelines currently cover five deployed systems, including MAI-Thinking-1 and MAI-Code-1.1-Flash.In addition to strict limitations, the draft also proposes three major goals: human flourishing, diverse values, and human control. Microsoft stated that the current models will not be trained according to this version, and a revised version will be released by the end of the year to guide model development in 2027. Public feedback includes philosophical questions about the statement "AI subordinate to humans," inquiries about accountability for irreversible decisions, and users pointing out issues such as the lack of mention of payment for training data.
app_icon
ChainCatcher Building the Web3 world with innovations.