SIBuilder (@SIBuilder) profile picture

Love exploring X, diving into AI, experimenting with new tech, and sharing interesting discoveries. 🚀

Joined November 2023
Photo by SIBuilder (@SIBuilder)
Man, I’m getting absolutely wrecked in this market. I get the whole deleveraging argument. AI stocks ran too far, valuations got out of hand, and a lot of these names probably needed to come back down to earth. But Jesus, how much is enough? Just because a stock was overpriced doesn’t mean the underlying business is suddenly worthless. Some of these companies are putting up incredible numbers. Revenue is growing, earnings are strong, demand is still there. And the stocks just keep getting hammered. At this point, you could announce that you’ve figured out nuclear fusion and the market would probably sell off anyway. Remember when Jensen Huang could do no wrong? NVIDIA was printing money, memory stocks were flying, and everyone wanted a piece of the AI trade. Now it feels like nobody wants to touch any of it. I was still making jokes about Claude yesterday. Today I’m looking at these hardware stocks and I don’t even have the energy to get mad anymore. And yeah, I know. Some of these stocks were stupid expensive. Maybe they still are. But seriously, what are we even using to value these things anymore? If you owned a $10 million property that was consistently bringing in $5 million a year, would you really dump it just because everyone suddenly decided real estate was a terrible investment? You’d at least look at the income before making that call. That’s all I’m saying. I’m not asking the market to give every AI company some ridiculous multiple forever. I just don’t understand how we went from paying anything for growth to acting like growth doesn’t matter at all. Maybe this is just what a real shakeout looks like. Maybe there’s still a lot more pain coming. I don’t know. But watching profitable companies get sold off like they’re going out of business is getting pretty damn exhausting. At what point does the selling stop being about valuation and start being just plain panic?
1
1
229
Why I’m Not Buying Anthropic Stock Anthropic just went public. Shares were priced at $83, jumped after the opening bell, and have since pulled back to around $90. I didn’t buy a single share. In fact, I think there’s a real chance the stock eventually falls below its IPO price. Here’s why. 1. Anthropic’s revenue growth may not be as sustainable as it looks. I’m a heavy Claude user, and if you’ve spent any serious time with Claude Code, you know how quickly the costs can add up. Run the /cost command, and it’s not unusual to see $100–$200 in API usage for a single session. For power users like me, that can add up to thousands of dollars a day. Right now, companies are racing to adopt AI and encouraging employees to use these tools as much as possible. That’s fueling incredible revenue growth for Anthropic. But here’s the question: What happens when CFOs start looking closely at those bills? Eventually, companies will have to figure out whether the productivity gains actually justify the spending. And once AI budgets come under serious scrutiny, I’m not convinced Anthropic can maintain its current growth trajectory. 2. Great technology doesn’t necessarily mean great shareholder returns. The AI model business reminds me of the auto industry. Massive capital requirements. Brutal competition. Constant technological disruption. These companies are spending billions training models and building infrastructure, only to risk seeing their competitive advantage disappear when someone else releases a better, cheaper alternative. That’s a difficult business model for investors. You can build incredible technology and still struggle to generate attractive long-term returns on capital. 3. Claude’s competitive moat isn’t nearly as deep as people think. I’ve been running the same tasks across Claude, Google’s Gemini, and OpenAI’s ChatGPT. My takeaway? All three are remarkably capable. Claude is a fantastic product, but it’s nowhere near irreplaceable. And competition is only going to intensify. Google and OpenAI have enormous resources. Open-source models are getting better every day. As the performance gap narrows, pricing pressure will only get worse. That’s not exactly an environment where I’d expect profit margins to expand. Here’s the bigger picture. I’m incredibly bullish on AI. But being bullish on AI doesn’t mean I have to be bullish on every company building AI models. The technology will keep improving. Models will become more capable, more accessible, and increasingly affordable. Over time, I wouldn’t be surprised if AI models become commoditized, with much of the value shifting toward the infrastructure that powers them. That’s where I see the real scarcity: GPUs, electricity, data centers, and the physical resources that can’t simply be scaled overnight. AI could absolutely change the world while still delivering disappointing returns to investors in some of its biggest companies. Both things can be true. And that’s why I’m staying on the sidelines for this IPO.
1
8
631
Wow. Dario Amodei, your IPO actually fell below its offering price. Seriously? After all that big talk about AI accelerating science and squeezing 100 years of medical breakthroughs into just 10, this is where we are? Anthropic stock is already trading below its IPO price on day two. And now the whole AI sector could be in for a reality check. If demand was so strong, why jack up the fundraising target by 60% just a day or two before the IPO? Come on. That looks a lot more like greed than confidence. And here's the real kicker: all seven founders reportedly sold shares on day one. You piss off your customers with your product decisions, then turn around and piss off your investors with your IPO. Keep this up, and pretty soon you'll have burned every bridge. Honestly, Anthropic's post-IPO flop could end up being a major turning point for the AI hype trade on Wall Street. You can talk all day about changing the world, curing diseases, and ushering in a new era of science. But none of that gives you a free pass to put greed ahead of investors. AI might change the world, Dario. But the market just reminded you that it still gets to decide what your company is worth.
2
8
775
AI is supposed to create value, not leave retail investors holding the bag. Anthropic, Wall Street, and Ray Dalio all deserve scrutiny. Who gets preferential access, who bears the risk, and who walks away with the profits? The AI boom needs fair markets, real transparency, and accountability. Innovation means little without trust.
translation: Anthropic fucked retail investors
2
1
7
1,018
Quick correction: I meant Dario Amodei (Anthropic CEO), not Ray Dalio. Mixed up the names. The point remains the same: AI should create real value, not leave retail holding the bag. Transparency and accountability matter.
1
5
250
Dario Amodei believes he is saving humanity. He locked 50.1% of the vote into a single share, built a $900 billion loop with the cloud giants, and left Wall Street the rest. His danger is not that he lies. It is that he feels no guilt.
11
1,844
Anthropic’s IPO is being treated like the second coming of Christ. Wall Street, give me a fucking break. Silicon Valley is selling us a future where your job is disposable, your electric bill keeps climbing, and the companies replacing workers are suddenly worth trillions. And we’re supposed to celebrate? They call it “productivity” when workers get laid off. They call it “innovation” when billionaires get richer. They call it “progress” when ordinary Americans are told to learn to code, adapt, or get left behind. Funny how the future is supposed to be cheaper and more efficient, but the first thing these companies want to cut is your paycheck. And the people telling you not to worry about AI taking your job? They’re betting billions on companies needing fewer people. Think about that. If AI works, shareholders get the upside. If workers get replaced, families eat the losses. If data centers drive up electricity costs, consumers get the bill. And if Wall Street’s trillion-dollar gamble blows up? Don’t be surprised when somebody comes looking for a bailout. I’m not against AI. I’m against pretending that making billionaires richer automatically makes America better off. A trillion-dollar valuation doesn’t mean a trillion dollars of value has been created for society. It means investors are willing to bet that much on the future. Those are two very different things. So spare me the speeches about changing humanity. Who gets rich? Who loses their job? Who pays for all of it? Because if the answer is that Silicon Valley gets the profits while working Americans get the bill, that’s not some glorious technological revolution. That’s a wealth transfer with better marketing. And the most insulting part? They expect us to applaud.
2
2
12
882
Anthropic’s IPO: Wall Street Gets Rich. Everyone Else Gets the Bill. Anthropic hit the Nasdaq on November 19, raising over $81 billion and climbing 17% on its first day. A trillion-dollar valuation. Just like that. Wall Street had plenty to celebrate. Early investors stood to make a fortune. Meanwhile, protesters were out on the streets of New York and San Francisco, raising concerns about where AI is taking us. Some were arrested. Same AI boom. Two very different realities. And let’s be real, there are some uncomfortable questions we can’t keep ignoring. Jobs: AI is getting better at doing work people get paid for. Companies see bigger profits. But what happens when those paychecks stop coming? Wealth: Tech billionaires are piling up fortunes while ordinary Americans are struggling with rent, groceries, and bills. How much of this supposed prosperity is actually reaching them? Wall Street: Early investors have opportunities to cash out. Retail investors are buying in at eye-watering valuations. And if the AI bubble bursts? A lot of ordinary people could be left holding the bag. Safety: The same companies warning us about AI’s potential dangers are racing to build even more powerful systems. Who’s making sure things don’t get out of hand? The Real Cost: Data centers are consuming enormous amounts of electricity and water. Surveillance, privacy, and military applications are raising serious concerns. The profits are concentrated. The costs don’t always stay that way. Control: We’re developing AI systems that can help build the next generation of AI. At some point, we have to ask who’s actually calling the shots. Look, AI could bring real breakthroughs in medicine, education, science, and productivity. That’s exciting, and it matters. But here’s what bothers me. A small group of companies is gaining extraordinary wealth and influence over technology that could reshape millions of lives. The rest of us have surprisingly little say in how it unfolds. And we’re expected to just trust that it’ll all work out. That’s a pretty sweet arrangement if you’re the one making the decisions. For everyone else? Not so much. No wonder trust is wearing thin. Anthropic’s IPO feels like the beginning of a much bigger conversation about who benefits, who takes the risks, and who gets to decide. AI could make the world incredibly wealthy. But if most people never see the benefits, what exactly are we celebrating?
1
3
442
Claude is fucking impressive. Claude Code, in particular, is on another level. Throw a messy codebase at it, and it can track down bugs, refactor code, build features, run tests, and work through problems without needing its hand held every step of the way. What really gets me is how well it picks up on intent. You can give it a rough idea of what you want, and it’ll often figure out the details, catch things you missed, and get shit done. It’s also damn good at digging through massive documents, connecting the dots, and making sense of complicated problems. Does it screw up? Absolutely. You still have to check its work. And the competition isn’t standing still. But here’s why Claude stands out to me: Give it a tough problem, some context, and room to work. There’s a damn good chance it’ll come back with something genuinely useful. That’s the kind of AI worth paying attention to.
4
481
AI might be moving a lot faster than most people realize. Google DeepMind researcher Tim Genewein thinks the world’s capacity to run AI agents could grow over 1,000x by 2030, and potentially over a billion times within a decade. He sees four paths from AGI to superintelligence: Bigger, smarter models. Breakthroughs in AI architecture. Massive networks of AI agents working together. AI building better AI. Here’s the kicker: these paths reinforce each other. If one hits a wall, the others can keep moving forward. But that last one is the big deal. Once AI starts building better versions of itself, the pace of progress could get pretty wild.
2
574
On Anthropic’s first day of trading, the way they handled things was truly disgraceful. The initial offering price was originally set at $70–$78, but it was raised straight to $83. Not only did they raise the price, but they also increased the number of shares offered from the planned 820 million to 980 million—a jump of nearly 20%. At least with instant noodles, they usually offer more for the same price—but Anthropic went and did both: increasing the quantity and the price. Even more outrageous is that on the very first day of trading, management cashed out $2 billion. Are they really that desperate? They could at least put on a show—don’t rush to cash out right after going public. How are retail investors supposed not to be scared by that? And that’s not all. Reportedly, 10% of employee stock options were unlocked on the first day of trading—amounting to approximately 300 million shares, equivalent to nearly one-third of the total offering. What’s truly hair-raising is the unlocking schedule that follows. A batch will be unlocked in mid-December, followed by successive unlockings in early, mid, and late January 2027. Throughout the first quarter of 2027, the total volume of shares expected to be unlocked is projected to reach 3 billion. What does that mean? The total offering size is only 980 million shares, yet the number of shares set to be unlocked in just one quarter is more than three times that amount. This isn’t a gradual release of shares—it’s a barrage that leaves no room to catch your breath. That’s why I said yesterday: I’m staying away from Anthropic stock. Now that I’ve seen these numbers, I’m even more convinced. There will be other opportunities to make money; there’s no need to risk my own money on a situation like this. I’d like to live a few more years.
372
689
165
5,598
1,615,038
OpenAI is quietly turning ChatGPT Health into more than just a place to ask health questions. It’s rolling out to desktop, Android, Dots, and Sites. But what’s interesting isn’t the expansion itself. It’s what ChatGPT can actually do with your data. On desktop, it can look at your sleep and workout history to help you figure out how recovery affects performance. Android gets Health Connect integration for sleep, activity, and workouts. Dots can spot a scheduling conflict between a meeting and your workout, then suggest moving it. And Sites can turn your running data into a personalized training plan website. That’s where things get interesting. ChatGPT isn’t just answering questions anymore. It’s starting to connect the dots between your data, your schedule, and what you might want to do next. The rollout is still U.S.-focused and gradual. Desktop and Android are available across Free, Go, Plus, and Pro. Sites requires Plus or Pro, and Dots requires Pro. It’s not a replacement for medical advice, obviously. But I think this points to something bigger. The next AI race might be less about which model scores highest on benchmarks and more about which assistant actually understands your day-to-day life and helps you get things done. With your permission, of course. help.openai.com/en/articles/…
1
1
359
The jump from AGI to superintelligence could happen faster than most people expect. Google DeepMind researcher Tim Genewein recently suggested that the number of AI agents running worldwide could grow 1,000x by 2030, and potentially over a billion-fold within a decade. He sees four main drivers: Bigger models, more data, more compute. New AI architectures. Millions of AI agents working together. AI helping develop better AI. The interesting part is how these all feed into each other. If one approach hits a wall, competition and investment could accelerate the others. But the biggest shift might be AI helping improve the next generation of AI. Once that feedback loop gets going, progress could compound in ways we’re not used to. Maybe the path from AGI to ASI isn’t one massive breakthrough, but millions of smaller improvements stacking up faster than anyone expects.
1
226
Tutorial for Beginners: How to Use Codex to Automate Tasks Even If You Don't Know How to Code
🤖 Made with AI
1
2
176
AI makes knowledge abundant, but independent thought more valuable than ever. What will set people apart is the depth of their insight, the clarity of their judgment, and their understanding of human nature. As machines push the limits of intelligence, humanity must rise to a higher level of wisdom.
4
134
🚀 Google launches Genie 4, integrating the Gemini API and Google AI Studio! What is a world model? Simply put, while ordinary AI primarily generates text, images, and videos, a world model can create a virtual world that constantly evolves, allows for free exploration, and enables real-time interaction. For example: You tell the AI, “Create an ancient Chinese city with streets, shops, pedestrians, and rivers.” Conventional AI might generate an image or a video of the city. A world model goes a step further: you can step into this virtual city, walk around freely, turn corners, browse the shops, and explore different scenes—and the surroundings will change in real time as you interact with them. The significance of this technology lies not only in generating more realistic visuals but also in enabling AI to learn and understand space, motion, the relationships between objects, and the potential outcomes of different actions. Through the Gemini API and Google AI Studio, developers can integrate Genie 4’s capabilities into their own applications. The future application scenarios are vast: 🎮 Game Development: Quickly create freely explorable game worlds using text. 🎬 Film and TV Production: Adjust scenes, camera angles, and character movements in real time. 🏬 Commercial Displays: Allow customers to browse stores online and experience virtual shopping environments. 🤖 Robot training: Simulate various tasks and operations in virtual environments. 📚 Education and training: Create historical and scientific scenarios that users can explore and interact with firsthand. From generating images and videos to creating virtual worlds that users can freely explore, AI is entering a whole new phase. 🌍✨
Genie 4 lets you step inside interactive environments and reshape them mid-session: shifting the weather, summoning characters, and generating adaptive native audio on the fly try it today via the Gemini API and in AI Studio: aistudio.google.com/live?mod…
There's a new version of this post
1
1
128
The OpenAI Responses API now includes a background task completion window, allowing users to trade time for cost savings on long tasks: • 2 hours: 30% reduction in token fees • 8 hours: 40% reduction in token fees Set `background: true`, then select 2h or 8h via `background_options.completion_window`. Currently in Public Beta, this feature supports select GPT-6 series models and is limited to the Standard tier; it does not support Flex. Note: This is an API feature, not a new button in ChatGPT. Official documentation: developers.openai.com/api/do…
1
3
229
Anthropic is expected to emerge from its IPO with $232B in cash and short-term investments, against a staggering $901B in future compute and infrastructure commitments. That’s the scale of the AI arms race. Billions are pouring in, compute capacity is being locked up, and Big Tech is making trillion-dollar bets on the future. The bigger the commitments, the higher the bar for returns. Capital can buy compute. It can’t guarantee profits. AI is entering its defining phase. Innovation will drive growth, but capital efficiency will determine who survives. The real test is just getting started.
1
308
AI isn’t just creating a gap in how well people use it. More importantly, it’s creating a gap in how much people can experiment and what they think is possible. When you’re low on tokens, you think about how to save them. When you have plenty, you start thinking about what else you can do with AI.
1
132
Your coworker used AI to finish a proposal in 10 minutes. The boss picked his. Not because he was faster. Because he was willing to put his name on it. AI is making answers cheap. Judgment is still expensive. Everyone can get an 80% solution now. The real value is in the last 20%: What to cut. What risks to take. What can’t go wrong. AI won’t get fired, lose money, or lie awake at 3 AM regretting a decision. You will. The future doesn’t belong to people who know the most. It belongs to those who make the right calls, own their mistakes, and stay in the game. AI can give you answers. It can’t take responsibility.
2
68