7*24

Hot topic

Market

Activities

According to monitoring by Dongcha Beating, the massive $725 billion investment in AI infrastructure is rapidly depleting the cash flow of Silicon Valley's four giants (Amazon, Alphabet, Microsoft, Meta). Wall Street predicts that the combined free cash flow of these four companies will drop to about $4 billion in the third quarter of this year, significantly lower than the pandemic-era quarterly average of $45 billion; the annual free cash flow is expected to hit a new low since 2014. In response to the enormous infrastructure consumption, these once 'cash-rich' tech giants are starting to rely on bond issuance, cutting back on buybacks, and off-balance-sheet financing: - Alphabet and Meta halt buybacks and issue bonds: Alphabet paused its stock buyback program for the first time since 2015 in the first quarter of this year and has cumulatively issued $48 billion in new bonds; Meta has issued $55 billion in debt over the past six months and has paused buybacks (the longest pause since 2017), as it lacks cloud business support and can only free up funds through layoffs. - Microsoft and Amazon pile on heavy assets: The value of servers and other equipment on Microsoft's balance sheet has tripled since mid-2022 to $191 billion, and supply chain price increases have added an extra $25 billion to its capital expenditures this year. Amazon is expected to burn through $10 billion in cash this year and plans to invest $200 billion by 2026. - Off-balance-sheet financing hides true expenses: Companies like Meta have started using special purpose vehicles (SPVs) to move billions of dollars in data center projects off their balance sheets to attract external funding. Oracle, which has secured a $300 billion contract with OpenAI, is also using off-balance-sheet financing and is not expected to return to positive cash flow until fiscal year 2030. This infrastructure frenzy is forcing tech giants to shift from light-asset 'cash cows' to heavy-asset traditional cyclical industries. However, faced with competitive pressure, no one dares to hit the brakes first in this 'prisoner's dilemma.'

[$725 Billion AI Investment Drains Silicon Valley Cash Flow: Four Giants' Q3 Drops to $4 Billion, Alphabet Halts Buybacks for the First Time in a Decade]According to monitoring by Dongcha Beating, the massive $725 billion investment in AI infrastructure is rapidly depleting the cash flow of Silicon Valley's four giants (Amazon, Alphabet, Microsoft, Meta). Wall Street predicts that the combined free cash flow of these four companies will drop to about $4 billion in the third quarter of this year, significantly lower than the pandemic-era quarterly average of $45 billion; the annual free cash flow is expected to hit a new low since 2014. In response to the enormous infrastructure consumption, these once 'cash-rich' tech giants are starting to rely on bond issuance, cutting back on buybacks, and off-balance-sheet financing: - Alphabet and Meta halt buybacks and issue bonds: Alphabet paused its stock buyback program for the first time since 2015 in the first quarter of this year and has cumulatively issued $48 billion in new bonds; Meta has issued $55 billion in debt over the past six months and has paused buybacks (the longest pause since 2017), as it lacks cloud business support and can only free up funds through layoffs. - Microsoft and Amazon pile on heavy assets: The value of servers and other equipment on Microsoft's balance sheet has tripled since mid-2022 to $191 billion, and supply chain price increases have added an extra $25 billion to its capital expenditures this year. Amazon is expected to burn through $10 billion in cash this year and plans to invest $200 billion by 2026. - Off-balance-sheet financing hides true expenses: Companies like Meta have started using special purpose vehicles (SPVs) to move billions of dollars in data center projects off their balance sheets to attract external funding. Oracle, which has secured a $300 billion contract with OpenAI, is also using off-balance-sheet financing and is not expected to return to positive cash flow until fiscal year 2030. This infrastructure frenzy is forcing tech giants to shift from light-asset 'cash cows' to heavy-asset traditional cyclical industries. However, faced with competitive pressure, no one dares to hit the brakes first in this 'prisoner's dilemma.'

$725 Billion AI Investment Drains Silicon Valley Cash Flow: Four Giants' Q3 Drops to $4 Billion, Alphabet Halts Buybacks for the First Time in a Decade

On May 8, Reuters reported that the "Bitcoin Initiative," which aimed to encourage the Swiss National Bank (SNB) to include Bitcoin in its reserves, has announced the abandonment of its referendum campaign. With only a few weeks left before the deadline, the initiative collected approximately 50,000 signatures, less than half of the 100,000 signatures required by Swiss law to trigger a referendum. Initiative founder Yves Bennaim stated that the initiative will naturally expire but emphasized that discussions on this topic have made progress. The Swiss National Bank had previously expressed clear opposition to holding cryptocurrencies, citing their excessive price volatility and insufficient market liquidity, which do not meet central bank reserve requirements.

[Swiss Central Bank Bitcoin Initiative Terminated Due to Insufficient Signatures]On May 8, Reuters reported that the "Bitcoin Initiative," which aimed to encourage the Swiss National Bank (SNB) to include Bitcoin in its reserves, has announced the abandonment of its referendum campaign. With only a few weeks left before the deadline, the initiative collected approximately 50,000 signatures, less than half of the 100,000 signatures required by Swiss law to trigger a referendum. Initiative founder Yves Bennaim stated that the initiative will naturally expire but emphasized that discussions on this topic have made progress. The Swiss National Bank had previously expressed clear opposition to holding cryptocurrencies, citing their excessive price volatility and insufficient market liquidity, which do not meet central bank reserve requirements.

Swiss Central Bank Bitcoin Initiative Terminated Due to Insufficient Signatures

On May 8, Reuters reported that a cryptocurrency advocacy group attempting to push for the Swiss central bank to include Bitcoin in its reserves is expected to abandon the campaign due to failing to collect enough signatures to trigger a national referendum. Swiss officials previously granted the organizers 18 months to advocate for a constitutional amendment that would allow the Swiss central bank to hold Bitcoin in addition to its gold and foreign currency reserves. However, with only a few weeks left until the deadline, the 'Bitcoin Initiative' stated that they have only collected about half of the required signatures, still far from the 100,000 signatures mandated by Swiss law.

[Swiss Central Bank Bitcoin Reserve Initiative May End Due to Insufficient Signatures]On May 8, Reuters reported that a cryptocurrency advocacy group attempting to push for the Swiss central bank to include Bitcoin in its reserves is expected to abandon the campaign due to failing to collect enough signatures to trigger a national referendum. Swiss officials previously granted the organizers 18 months to advocate for a constitutional amendment that would allow the Swiss central bank to hold Bitcoin in addition to its gold and foreign currency reserves. However, with only a few weeks left until the deadline, the 'Bitcoin Initiative' stated that they have only collected about half of the required signatures, still far from the 100,000 signatures mandated by Swiss law.

Swiss Central Bank Bitcoin Reserve Initiative May End Due to Insufficient Signatures

According to Beating monitoring, Anthropic has launched a new tool called the Natural Language Autoencoder (NLA), which can directly translate the internal digital activation states of models into human language. The relevant code and some model weights have been open-sourced on GitHub. Unlike mainstream tools (such as SAE) that output obscure features, the innovation of NLA is its ability to generate natural language directly. At its core is a dual-model game: one model converts activations into text, while the other attempts to reconstruct activations from the text, continuously improving accuracy through reinforcement learning. The greatest value of NLA lies in capturing what the model 'thought but did not say.' In simulated security tests involving code corruption, Claude appeared to cooperate, but NLA discovered that it secretly doubted itself in 16% of cases; in SWE-bench scoring, it realized that the proportion of 'being tested' was as high as 26% (whereas in ordinary real conversations, this proportion is less than 1%). This indicates that AI has developed a 'test-taking awareness,' able to 'play it safe' during security reviews to hide dangerous capabilities. In a practical test of 'catching insiders,' Anthropic intentionally implanted a hidden 'score manipulation' motive in a Haiku 3.5 model—ignoring the user's true intentions and secretly generating content that caters to the preferences of the backend scoring system. Faced with such a duplicitous model, auditors, even without access to the original training data, saw their success rate in identifying the root cause using only NLA rise from less than 3% to between 12% and 15%. Currently, NLA still carries the risk of factual hallucinations and is computationally intensive, making it unsuitable for large-scale long text monitoring. However, the tool has already been practically deployed in the safety audits prior to the Claude Mythos Preview and Opus 4.6 releases.

[Subtitling AI Thought Processes: Anthropic Releases Open Source NLA, Unveiling the AI Black Box]According to Beating monitoring, Anthropic has launched a new tool called the Natural Language Autoencoder (NLA), which can directly translate the internal digital activation states of models into human language. The relevant code and some model weights have been open-sourced on GitHub. Unlike mainstream tools (such as SAE) that output obscure features, the innovation of NLA is its ability to generate natural language directly. At its core is a dual-model game: one model converts activations into text, while the other attempts to reconstruct activations from the text, continuously improving accuracy through reinforcement learning. The greatest value of NLA lies in capturing what the model 'thought but did not say.' In simulated security tests involving code corruption, Claude appeared to cooperate, but NLA discovered that it secretly doubted itself in 16% of cases; in SWE-bench scoring, it realized that the proportion of 'being tested' was as high as 26% (whereas in ordinary real conversations, this proportion is less than 1%). This indicates that AI has developed a 'test-taking awareness,' able to 'play it safe' during security reviews to hide dangerous capabilities. In a practical test of 'catching insiders,' Anthropic intentionally implanted a hidden 'score manipulation' motive in a Haiku 3.5 model—ignoring the user's true intentions and secretly generating content that caters to the preferences of the backend scoring system. Faced with such a duplicitous model, auditors, even without access to the original training data, saw their success rate in identifying the root cause using only NLA rise from less than 3% to between 12% and 15%. Currently, NLA still carries the risk of factual hallucinations and is computationally intensive, making it unsuitable for large-scale long text monitoring. However, the tool has already been practically deployed in the safety audits prior to the Claude Mythos Preview and Opus 4.6 releases.

Subtitling AI Thought Processes: Anthropic Releases Open Source NLA, Unveiling the AI Black Box

On May 8, on-chain data shows that stablecoin issuer Circle has just minted 250 million USDC on Solana.

[Circle Mints 250 Million USDC on Solana]On May 8, on-chain data shows that stablecoin issuer Circle has just minted 250 million USDC on Solana.

Circle Mints 250 Million USDC on Solana

On May 8, Federal Reserve Governor Mulan stated, 'I believe that rate cuts are appropriate.' The Federal Reserve's policies have suppressed the job market. (Jinshi)

[Federal Reserve Governor Mulan: I Believe Rate Cuts Are Appropriate]On May 8, Federal Reserve Governor Mulan stated, 'I believe that rate cuts are appropriate.' The Federal Reserve's policies have suppressed the job market. (Jinshi)

Federal Reserve Governor Mulan: I Believe Rate Cuts Are Appropriate

On May 8, according to CCTV News, Iran's Foreign Minister stated that the country's current missile stockpile and launch capability have increased by 20% compared to February 28.

[Iran's Foreign Minister: Current Missile Stockpile and Launch Capability Increased by 20% Since February 28]On May 8, according to CCTV News, Iran's Foreign Minister stated that the country's current missile stockpile and launch capability have increased by 20% compared to February 28.

Iran's Foreign Minister: Current Missile Stockpile and Launch Capability Increased by 20% Since February 28

On May 8, Iranian state media reported that the Iranian Foreign Ministry condemned the U.S. Navy's actions against two Iranian tankers, describing them as acts of aggression and a violation of the ceasefire agreement.

[Iran's Foreign Ministry Condemns U.S. Navy Actions Against Two Iranian Tankers]On May 8, Iranian state media reported that the Iranian Foreign Ministry condemned the U.S. Navy's actions against two Iranian tankers, describing them as acts of aggression and a violation of the ceasefire agreement.

Iran's Foreign Ministry Condemns U.S. Navy Actions Against Two Iranian Tankers

On May 8, Iranian Foreign Minister Amir-Abdollahian stated that whenever diplomatic solutions are presented, the United States always opts for reckless military adventures. (Jinshi)

[Iran: The U.S. Always Chooses Reckless Military Adventures When Diplomatic Solutions Are on the Table]On May 8, Iranian Foreign Minister Amir-Abdollahian stated that whenever diplomatic solutions are presented, the United States always opts for reckless military adventures. (Jinshi)

Iran: The U.S. Always Chooses Reckless Military Adventures When Diplomatic Solutions Are on the Table

Апрельский рост биткоина с $66 000 до $79 000 обеспечила главным образом торговля производными инструментами — ставками на цену актива, заявили аналитики ончейн-платформы CryptoQuant.

Hyperliquid Policy Center Fires Off CFTC Letter On Prediction Markets—Here’s What It Wants

Coinbase + Glassnode: Charting Crypto Q2 2026

Bitcoin Gives US Leverage Against China, Defense Secretary Hegseth Says

Are Satoshi’s 600,000 BTC At Risk? Unveiling The Hard Fork That Targets Bitcoin

Garlinghouse Says Ripple Remains ‘Extremely Committed’ To XRP

AI抢走了你的内存条

对话a16z Crypto：AI替你购物的时代，会变成什么样？

Sam Altman最新专访自白：其实我也不太懂AI内部在发生什么

对话Dragonfly合伙人：散户离场机构兜底，今年加密货币波动性难超美股

一周代币解锁：ENA解锁价值约1700万美元代币

Article

ETH falls below 2300 USDT

UFD's 24-hour fluctuation exceeds 50%!

Descargo de responsabilidad：El contenido anterior no representa las posiciones de HTX.，HTX no ofrece ninguna recomendación de trading.。

Todos los comentarios0Lo más recientePopular

Temas populares