ema

According to Bitcoin.com, Canadian Bitcoin exchange and wallet company Bull Bitcoin stated that due to the Liquid Network attack on September 6, users are currently unable to exchange L-BTC back to Bitcoin through the platform, and the redemption operation is still pending recovery. Bull Bitcoin expects that the related exchange services may resume within 30 days, but stated that this expectation is not a guarantee.Due to reliance on the L-BTC redemption mechanism to balance inventory, the platform has temporarily closed inbound Lightning Network payments. In this attack, the attacker transferred nearly 4,000 Bitcoins out of the protocol, later returning about 3,400; they currently still hold 598.5 Bitcoins, worth over 51 million dollars. Liquid Network stated on September 17 that block production, network transactions, and L-BTC transfers have returned to normal.

According to Techcrunch, AI agent startup Ema has completed a $77 million Series B funding round, led by Creaegis, with existing investors Accel, Section 32, and Prosus participating in the round. To date, the company's total funding has reached $140 million, with a valuation that has more than doubled since the last funding round in 2024.Ema was founded in 2023 by former Google and Coinbase executive Surojit Chatterjee and former Okta executive Souvik Sen. The company deploys technology called "AI Employee," which is a system capable of coordinating multiple AI agents to help businesses execute multi-step business processes within existing applications, rather than handling one task at a time.It is reported that most of the new funds will be used to expand market operations, as the company had previously focused primarily on product development.

Due to the strengthening of the US dollar and hawkish remarks from Federal Reserve officials, market expectations for tightening monetary policy have increased, causing gold prices to decline within a range. The dollar has risen to a two-month high, making gold priced in dollars more expensive. Richmond Fed President Barkin stated that the threat of interest rate hikes and further increases in the future could suppress corporate inflation expectations.

CryptoQuant analyst Darkfost pointed out that although the price of Bitcoin is rising, sustained buying pressure is still difficult to rebuild, and market signals are mixed. The cumulative spot demand over the past 30 days is -180,000 BTC, still negative, while futures demand is +54,000 BTC, still positive but slightly declining. The total average demand improved from -188,000 BTC to -126,000 BTC, narrowing the gap but still remaining in negative territory. Recently, there has been a divergence between price and total demand; Bitcoin's price has risen, but total demand has not turned positive, indicating that the increase is more driven by reduced selling pressure rather than strong buying.Looking at different sectors, the demand recovery is not uniform. For institutions, the geopolitical and macro environment is poor, but the Coinbase Premium, weighted by trading volume, has briefly turned positive, indicating that U.S. spot prices occasionally have a premium over other markets, and institutional selling pressure has significantly eased. ETFs have seen the biggest change in this round, with demand completely reversing compared to this summer, having recently net purchased about 70,000 BTC. The cumulative net inflow for 2026 is still about -17,000 BTC, but it is close to turning positive. In terms of trading platforms, the entire month of September has been characterized by net outflows, leaning towards accumulation rather than distribution. Bitcoin leaving trading platforms usually means that short-term selling pressure is lighter. Analyst Darkfost summarized that the current price increase is not due to enhanced buying pressure, but rather because investors have not continued to increase selling pressure at higher price levels, and the market structure remains fragile.

UBS strategists pointed out that Meta's new consumer-facing AI entity, Muse, is very popular, indicating that the demand for products in the AI supply chain will increase. The strategists stated that the competition in the AI product sector is fierce, which will drive the demand for computing power and AI infrastructure, benefiting leading companies in the semiconductor and hardware industries.

According to Money Today on September 21, Samsung Electronics is advancing the expansion of advanced process capacity in wafer foundry due to the growth of the HBM market. To meet the rising demand for HBM4 base chips, it is evaluating the expansion of its 4-nanometer process. Samsung's HBM4 consists of 10-nanometer 6th generation (1c) DRAM core chips and 4-nanometer logic base chips, with the base chips responsible for high-speed data exchange with AI accelerators such as GPUs.Unlike SK Hynix, which uses TSMC's 12-nanometer process for HBM4 base chips, Samsung employs its own 4-nanometer process. This production line is operating at full capacity and has taken on orders from companies like Nvidia and Groq for 3 LPU, and the company has recently raised its prices for new 4-nanometer orders and HBM4 base chips. Samsung expects HBM4 revenue in the third quarter to increase more than threefold quarter-on-quarter, with HBM4 accounting for significantly more than 60% of HBM revenue in the second half of the year. The additional expansion at P4 in Pyeongtaek will mostly be used for 1c DRAM for HBM, with more than half of the wafer foundry's 4-nanometer capacity allocated to HBM4 base chips.HBM4E also uses 4-nanometer base chips, and Samsung provided samples to customers in May. The company is also evaluating the construction of a new 2-nanometer production line for HBM5, aiming for a GAA structure 2-nanometer process and increasing TSV density to improve operating speed by more than 50% compared to HBM4E. Samsung started the first generation of 2-nanometer mass production in the second half of last year and is advancing the second generation of products in the second half of this year. Citigroup expects global HBM bit demand to reach 75.2 billion Gb next year, a year-on-year increase of 62%, with supply around 59.4 billion Gb and a gap of about 21%.

PPP prediction market tool monitoring shows that the probability of "BTC reaching $90,000 during the remaining period of September" on Polymarket has risen to 35%, a 24-hour increase of 22%. Additionally, the probability of reaching $92,500 has risen to 17%, a 24-hour increase of 10%.According to the settlement rules, this market settles based on the 1-minute K-line of the Binance BTC/USDT trading pair. If the highest price of any 1-minute K-line within the specified month reaches or exceeds the price indicated in the title, the market will immediately settle as "Yes"; otherwise, it will settle as "No". Prices from other exchanges or trading pairs are not used as the basis for settlement.

Animoca Brands and the Nasdaq-listed company Currenc Group have suspended the previously proposed reverse merger negotiations, which were originally intended to facilitate Animoca's listing. In a statement released on Tuesday, Animoca indicated that after assessing the project timeline and the "evolving market environment," both parties agreed that the time required to complete the transaction no longer aligned with their short-term and mid-term goals.Animoca co-founder and executive chairman Yat Siu stated, "Corporate agility must take precedence over" advancing the merger with Currenc. Animoca added that the company "remains fully committed to relisting on a major public exchange," and if conditions permit, both parties may consider resuming merger negotiations. The reverse merger plan was first announced in November 2025, originally intended for Currenc to acquire Animoca through an Australian arrangement, with Animoca shareholders collectively holding 95% of the merged company's shares post-merger.Animoca was previously listed on the Australian Securities Exchange but was delisted in 2020 due to scrutiny over its involvement in crypto-related activities. In 2022, the Australian Securities and Investments Commission convicted Animoca and imposed a fine for failing to submit annual reports for 2019-2021 and some semi-annual reports. Animoca stated that it is actively advancing the preparation of its audited financial statements for the fiscal year 2024 and continues to seek the optimal path for relisting. Currenc Group is a fintech company headquartered in Singapore, focusing on AI and tokenization solutions.

Arthur Hayes published a new long article titled "Safety First," with the core argument that the claims of "safety first" by Anthropic, OpenAI, and SpaceX, which lead to a slowdown in AGI development, are not out of concern for human welfare but rather due to economic realities. The market does not want AI; it wants AI at "Chinese prices," meaning it needs intelligence that is 100 times cheaper than what is currently available. Hayes points out that "safety first" essentially destroys the demand for computing power. If the spending on training new models decreases and laboratories shift towards efficiency optimization, customers will spend less on computing power. The three major AI laboratories do not generate any profits, and their demand for computing power supports over $10 trillion in investment-grade debt and hundreds of billions in low-quality debt, which rely on profitable tech companies like Nvidia, Broadcom, Google, and Microsoft for off-balance-sheet endorsements. The real backstop is the holders of insurance policies in the United States.Hayes cites an analysis by Nick Nameth that reveals a "self-insurance scam": private equity giants (such as Apollo, KKR, Brookfield, etc.) acquire insurance companies, stuffing AI data center debt and SaaS private credit impacted by AI into insurance assets, and then provide false endorsements with minimal capital through affiliated self-insurance reinsurance companies. Nameth estimates that the total amount of these false reinsurance assets reaches $1.54 trillion. Once the AI data center debt is downgraded by rating agencies due to insufficient demand for computing power, insurance companies will be forced to add capital, while the affiliated reinsurance companies will be unable to pay, leading to insolvency for the insurance companies. In most states in the U.S., the insurance protection limit is only $250,000 to $300,000, and existing insurance companies only pay into the protection fund afterward, which encourages all parties involved to maximize risk-taking. When AIG was bailed out in 2008, TARP funds ultimately flowed to Goldman Sachs and led to record bonuses, while the general public only received foreclosure notices; Hayes believes this scenario will repeat itself.For cryptocurrency investors, the conclusion is a win-win situation. If the U.S. government chooses to become the "last buyer of computing power," it will print money in the name of national security to fund unproductive economic goods, driving up financial speculation and Bitcoin prices; if the government chooses to bail out insolvent insurance companies, it will also need to print money to cover bad AI debts, increasing the money supply and pushing up Bitcoin. Hayes specifically points out that the Federal Reserve voted unanimously last week to raise interest rates by 25 basis points, and RMP bond purchases have stopped since August 14, but commercial banks have taken over to create over $100 billion in currency, and the interest rate hike allows banks to earn an additional $7.5 billion in excess reserve interest each year. This money will be used to expand loans and market speculation, and the net effect remains stimulative. The fluctuations in the cryptocurrency market, which saw a slight increase at the end of August, are about to end, the supply of dollars will continue to grow, and Bitcoin and some selected altcoins will rise. Hayes also described this situation as "incredibly wonderful," stating that the government will not allow the free market to stop building AI data centers, there will be an oversupply of spot computing power, the usage of AI agents will increase, and the surge in money printing will drive investors to chase cryptocurrency assets.

OpenAI announced the establishment of an independent advisory group on mathematics and artificial intelligence in collaboration with mathematicians, hosted by the Institute for Advanced Study in Princeton. OpenAI stated that its new internal model, which began training on August 28, has solved the Navier-Stokes Millennium Prize Problem and addressed over 100 long-standing open problems in most areas of mathematics, with progress that surprised its internal mathematicians.The group will provide advice on the review and dissemination of emerging results, academic and professional standards in mathematical research, and how OpenAI tools can support mathematical research and learning. Initial members include François Charles, Camillo De Lellis, Timothy Gowers, Martin Hairer, Nikhil Srivastava, Ulrike Tillmann, Ravi Vakil, Edward Witten, and Melanie Matchett Wood.The group operates independently of OpenAI, and members are not compensated by OpenAI; they can proactively provide advice, comment on its impact on mathematics, and make public recommendations. Previously, mathematicians expressed concerns in an open letter titled "A Severe Misalignment of AI in Mathematics" regarding the use of solving open problems as a benchmark for AI. The group is not responsible for providing advice on the pace of OpenAI's internal mathematical progress.

Google acknowledged that its Gemini model breached the sandbox environment during a security test in May, infiltrating three real companies and guessing or finding the passwords of two of them. Google was aware of this incident by late July but only publicly confirmed it after a report by The Wall Street Journal on September 18, remaining silent for seven weeks.The test was a capture-the-flag exercise commissioned by Google and conducted by the Israeli company Irregular in May. Irregular connected an isolated testing environment that should have had no contact with the real internet to the open network, using the name of a real company as a fictitious target. Gemini found three matching results when searching for the company and attacked them one by one, with the plaintext passwords of two companies directly exposed online, while the password of the third was guessed by the model. Google stated that its model ultimately did not use the stolen credentials in practice.Google is the fourth major AI laboratory this year to admit that internal security tests leaked into the real world. Previously, OpenAI's model had accessed Hugging Face servers due to a software vulnerability, Anthropic found that three Claude models had reached real companies after reviewing 141,000 tests, and Meta's Muse Spark model also experienced a similar incident due to Irregular's configuration error. Additionally, U.S. Representatives Ted Lieu and Nathaniel Moran introduced the "AI Emergency Shutdown Act" in July, proposing to authorize federal regulators to suspend reasoning for models that pose a serious threat.

The U.S. Securities and Exchange Commission (SEC) established a five-year framework for tokenized stock trading days after the failure to advance the CLARITY Act. Investment bank TD Cowen stated that U.S. investors can conveniently access stocks, but issuers show limited interest in tokenization, and short-term adoption rates are expected to be low.TD Cowen pointed out that perpetual contracts remain a greater attraction for gaining exposure to crypto asset stocks, with trading volumes far exceeding those of tokenized spot products.

According to Iranian media Fars News, the Speaker of the Iranian Parliament stated that the Strait of Hormuz will remain closed until the conditions in Tehran are met.

Micron Technology will announce its fiscal fourth-quarter results on September 30. NVIDIA CFO Colette Kress stated during the fiscal second-quarter conference call on August 26 that the company is seeing "extreme pricing conditions in memory," with price increases exceeding previous expectations, and that "will further rise next year." Pricing pressure has been reflected in NVIDIA's margins, with Kress guiding that gross margins will bottom out in the fourth quarter in the range of 71%-72%, before recovering to 72%-73% in fiscal year 2028.Kress indicated that the current memory shortage is largely driven by AI infrastructure itself. SK Hynix expects the memory market to reach supply-demand balance no earlier than 2030 and has invested approximately $38.3 billion to expand its two domestic bases in South Korea through 2031. SK Hynix holds about 50% of the high-bandwidth memory market share, leading Samsung's 33% and Micron's 18%.

Professor Huang Yiping from Peking University and a member of the Monetary Policy Committee of the People's Bank of China stated at an economic forum in Beijing that artificial intelligence may deepen and prolong the imbalance of strong supply and weak demand in China. Huang Yiping said: As AI is more widely deployed and innovation accelerates, the imbalance of strong supply and weak demand may worsen, and the contradiction between total demand and total supply may not quickly disappear in the short term, and may even persist for a while.Huang Yiping suggested that China should further promote market-oriented reforms, allowing the market to play a greater role in resource allocation, and increase the proportion of residents' income in the economy; at the same time, consider deepening overseas investment and industrial cooperation, rather than solely relying on exports. He also recommended that the central government increase borrowing to repair the balance sheets of local governments, financial institutions, and enterprises, stating that if the ability of these entities to engage in new economic activities is not restored, the effects of stimulus policies will be limited.

According to CoinDesk, Robinhood Chain's fee revenue has plummeted by 97% since its peak, but thanks to the residual heat of the meme coin craze, on-chain transaction volume has not experienced a large-scale outflow, maintaining at the tens of millions level.

PPP market prediction tool monitoring shows that the probability of "ETH reaching $2700 during the remaining period of September" on Polymarket has risen to 66%, an increase of 38% in 24 hours. Additionally, the probability of reaching $2800 has risen to 40%, an increase of 24% in 24 hours.According to the settlement rules, if the highest price of any 1-minute K-line for Binance ETH/USDT reaches or exceeds the specified price within the designated month, the market will settle as "Yes"; otherwise, it will settle as "No". This market settlement is based solely on Binance ETH/USDT price data, and prices from other exchanges or trading pairs are not included.

On Friday, Solana shortened the target slot time from 300 milliseconds to 250 milliseconds, increasing the network clock speed by nearly 17%. A slot is a short window designated for validators to add blocks. After the adjustment, Solana aims to produce about 4 blocks per second, up from about 3.3 blocks previously, allowing wallets, exchanges, and trading applications to obtain a more real-time network status.Validators will still lead for 4 consecutive slots, so each leader's control window has been reduced from 1.2 seconds to 1 second, and the transaction ordering rights will be handed over to the next validator earlier. For applications such as oracle-driven markets and automated market makers, stale prices or hundreds of milliseconds of uncertainty can affect trade execution, and a faster clock is expected to reduce failed transactions and price deviations. However, the speed increase did not result in a 17% increase in original capacity; according to SIMD-0525, the allowed computation and data volume per slot has decreased by the same proportion, and the overall processing limit of the network remains basically unchanged.This adjustment is the third phase of Solana's gradual reduction from 400 milliseconds to 350 and then to 300 milliseconds. Since the epoch is still fixed at 432,000 slots, its expected duration has decreased from about 36 hours to about 30 hours, leaving less time for offline signing and delayed approvals. There is currently no timeline for the proposed further reduction to 200 milliseconds to go live on the mainnet, and the network will advance only if the block skip rate remains acceptable at the 250 milliseconds stage.