与「Closed」相关的搜索结果

If AI large-model talent in China is looking for domestic career opportunities, WeChat is actually worth considering. Not many people know that the WeChat team also has its own large model, WeLM, which is closed-source and has already iterated to version V4. At this stage, joining established, scaled large-model teams is the more conventional choice. But there are also people who choose to join WeChat’s relatively small-scale large-model effort, which is somewhat like trying to bottom-fish in the stock market—of course, it still depends on individual judgment and ability to evaluate the situation. Revisit this post in three months.

TOMORROW SPACEX GOES PUBLIC AT A $1.77T VALUATION. THE LAST TIME ELON MUSK TOOK A COMPANY PUBLIC, A REPORTER LISTED EVERY REASON THE DEAL WAS A "LEAP OF FAITH." The company was not making a profit. No mass production. No distribution. Just a prototype. Here was Musk's response, on Tesla IPO day, June 29, 2010: "The smartest money in the world is betting on Tesla. They must have a reason for doing so because obviously, to go public and increase the range and increase the price in this market is not easy, and these are very well-informed people." Then the reporter pressed him on the losses. Musk fired back: "If Tesla were just making the sports car and doing our powertrain supply business, we would be profitable as a company. We're expanding by 3,000% in the space of 2 to 3 years. It's just not possible for the company as a whole to be profitable with that rate of expansion. It's not some fundamental flaw in the Tesla business." Tesla $TSLA priced at $17, opened at $19, and closed at $23.89. Up 41% on day one. The deal raised $226M. At the time, Musk himself had just $650,000 in liquid assets per public filings and was taking loans from friends to cover personal expenses. Almost everything he had was tied up in Tesla. A $10,000 investment in Tesla at the IPO would be worth ~$3.41 million today. BlackRock alone just ordered at least $5B worth of SpaceX shares.

We’ve reached an agreement to acquire @ona_hq. Its secure cloud execution technology will help Codex take on longer-running work, even when laptops are closed, and help more organizations deploy agents securely in production. After closing, Ona will join OpenAI’s Codex team.

A few years from now, nobody is going to remember how many likes you got on a random post. They’ll remember how you made them feel. I learned that the hard way. Some of the posts I’m most proud of barely got noticed. Meanwhile, some of the ones that performed best were forgotten within days. The numbers looked great, but they didn’t leave a mark on anyone, including me. That’s when I realized attention and impact are not the same thing. Attention is easy to measure. Impact isn’t. One disappears with the next scroll. The other stays with someone long after they’ve closed the app. That’s why I’ve started caring less about reach and more about whether something I share is actually worth reading. Not because it’s the best strategy. Because life is too short to spend it saying things you don’t believe. Part of why @RallyOnChain stands out to me is that it pushes people to contribute ideas, not just collect impressions. So here’s something I’ve been thinking about: If every post you’ve ever made disappeared tomorrow, which one would you still be proud of?

Most investors are asking: which AI company wins? That's the right question for the last cycle. The GPU constraint formed. The investors who identified it early — before the crowd arrived, before Wall Street coverage was broad — participated in some of the most significant returns in technology investing history. That window is closed. The GPU trade is consensus now.

In 2008 Elon Musk had $0 left. He had poured his entire $180,000,000 PayPal fortune into SpaceX and Tesla. SpaceX had just failed its third rocket launch in a row. Tesla was on the verge of bankruptcy. His marriage fell apart. He lost his house. He was borrowing money from friends just to pay rent. In an interview that year he broke down in tears and said: “I never thought I was someone who could ever be broken. But 2008 broke me. I just didn’t know how I was going to make it.” Three months later SpaceX launched successfully for the first time. Six months later Tesla closed a financing round on Christmas Eve, literally hours before bankruptcy. Today SpaceX is worth $350,000,000,000. The people who told him to save his money and go sit on the beach are watching.

California Is Blocking a Federal Audit of Its Voter Rolls California allows first-time voters to register using forms of ID that most Americans would find surprising, including: -Gym membership card -Employer ID card -Credit or debit card -Prescription drug label -Insurance card (California provides free health coverage to undocumented immigrants) Full list: This is permitted when a voter fails to provide a Social Security number or driver’s license at registration. Our office believes this policy deserves a closer look. We also have serious concerns about how California maintains its voter rolls. There are open questions about whether the state is promptly removing deceased voters, people who have moved, and individuals convicted of disqualifying felonies. On top of that, California allows third parties to collect and turn in ballots on voters’ behalf (a practice known as ballot harvesting) with few restrictions. This makes it difficult to track who actually received, completed, and submitted each ballot. For over a year, the Department of Justice has been trying to audit California’s voter rolls. Federal law gives the Attorney General the authority to review state voter files and confirm that only eligible U.S. citizens are voting in federal elections. @AAGDhillon sent California a letter explaining our legal authority. California refused to comply, claiming state privacy laws block the review, an argument that does not hold up because those laws don’t apply to the federal government in this context. We’ve sued California in federal court, and the case is before the Ninth Circuit Court of Appeals. If California genuinely wants voters to trust its elections, it should open its records, not fight to keep them closed. What are they afraid of?

Last week in crypto was not just about prices going down. It was more about seeing how fragile the market still is when liquidity starts to thin out. BTC traded below 60K during the week. According to CoinShares, digital asset investment products saw 1.67B in outflows. 1.438B of that came from Bitcoin products. 257M came from Ethereum products. At the same time, ETF demand clearly started to weaken. Then the stronger US jobs data reminded the market of something simple: the liquidity story is not over yet. The US added 172K jobs in May. That made rate cut expectations weaker, and pressure came back across risk assets. Not just crypto. Nasdaq fell 4.2% on Friday. S&P 500 closed the week down 2.6%. In crypto, that pressure was amplified by leverage. Around 1.6B in positions were liquidated. The lesson for me is pretty simple. When the market is going up, everyone looks at the story. When the market is going down, you have to look at the mechanics. → who was buying → are they still buying → how much leverage was sitting in the system → does the macro environment still support the move This week was a good reminder of something people often forget. A good narrative does not mean there is always a real buyer at every price level. When ETF inflows are strong, dips get absorbed more easily. When ETF flows turn negative, the same dip feels much heavier. That was the main takeaway for me. Not that crypto is broken. More that even strong narratives become fragile when liquidity is not there to support them.

Horizon V 0.2 is LIVE for closed beta! >Strategies now run live >Stability fixes across the board >2x faster generation on every strategy Spots are limited: