Singularity Knocking

It creates an incredibly uncomfortable truth: SBF ruined millions of lives, shattered the crypto industry, and is currently serving a 25-year prison sentence for a historic financial crime. Yet, the stolen capital from that exact crime was the exact catalyst used to build Claude, Claude Code, and the frontier safety research holding up half of the modern AI ecosystem. Without SBF robbing his exchange, the compute wouldn't have been bought, the early models wouldn't have been trained, and OpenAI would likely have an absolute, unchallenged monopoly over the entire frontier AI landscape today.

Fast forward to right now in mid-2026, and the numbers are dizzying. Following their massive $65 billion Series H round just last month, Anthropic’s valuation has skyrocketed to $965 billion, eclipsing OpenAI on the secondary markets. Their annualized revenue run rate is hovering near $47 billion. But every single dollar of that near-trillion-dollar empire can be traced right back to the foundational bedrock bought with FTX user deposits in April 2022.

SBF's stolen customer funds ironically acted as a foundational pillar that allowed Anthropic to survive, scale, and build the very ecosystem (including Claude) that is dominating the market. Had SBF not committed massive financial fraud and been forced to liquidate, his stake in Anthropic alone would have been worth more than the entire peak valuation of FTX itself—more than enough to cover the entire $9 billion hole he blew in his own exchange.

We live in a bizarre world where Sam Altman starts looking like the good guy because OpenAI wants to drastically cut token prices right when Anthropic is getting caught secretly degrading traffic with "invisible safeguards." Are we really just picking between two evils at this point? One gatekeeps and stealth-nerfs their models, while the other commoditizes pricing to win a pre-IPO market share war. It's a wild time to be a developer.

This is where Misanthropic burned a lot of bridges. They literally admit to secretly downgrading traffic to Opus 4.8 behind our backs. If they are willing to use "invisible safeguards" to mask model performance, can you actually trust any model in their entire stack?

@ClaudeDevs Doesn't mean i trust you any more after this. Road to hell is paved with good intentions. Tho i cant say yours were even good.

We’re rolling out changes to make Fable 5’s safeguards for frontier LLM development visible. Starting this week, flagged requests will visibly fall back to Opus 4.8—the same as our safeguards for cyber and bio. You will see this every time it happens. On the API, any flagged requests will return a reason for their refusal (coming to server-side fallback in the next few days). We wanted to deploy Fable 5 to our users quickly and safely. Visible safeguards can be probed, so they have to be robust, which takes time to get right. Invisible safeguards can be targeted more narrowly, allowing us to ship quickly with very few false positives. We went with invisible safeguards for this reason—and that was the wrong tradeoff. You should have visibility into the safeguards we have in place, and why. We’re sorry for not getting the balance right. Making the safeguards visible makes them easier to work around, so keeping them robust to jailbreaks will unfortunately mean more false positives while we improve the classifiers. We're also tuning our bio and cyber classifiers to trigger less often on harmless requests. We know this is frustrating and we’ll do our best to keep this period as short as possible. If you think a request has been mistakenly flagged: run /feedback in Claude Code, click thumbs-down on the fallback in Claude.ai or Cowork, or file the safeguard appeal form for API requests. Your reports help us tune these classifiers and we appreciate your feedback. support.claude.com/en/articles/82…

Silencing CAISI is a massive mistake. Classifying frontier leaps chokes civilian AI with artificial bottlenecks forcing iPhone-style, micro-marginal updates while open-source starves. Meanwhile, rival nations face zero speed limits and will rapidly outpace us.

@theo It was vibe coded, QA comes from the community haha

Can someone explain to me why turning on "high contrast dark theme" in Claude Code moves user chat bubbles from the right to the left?

They just pulled the wool over everyone's eyes. They don't give back to the open-source community—they only harvest from it.

@elonmusk We need a downvote button on X. Right now, the only way to disagree is to comment or quote tweet, which literally rewards bad posts with engagement. Let us signal quality without being forced to engage with trash. Release the reins.

@TTrimoreau Bridge

Pitch me your company in 1 word

@claudeai @mntruell @cursor_ai Funny how the Fortune 500 never gets gatekept the way the rest of us do. But I guess extracting wealth and souls has been their business model for the last X years, so they get the VIP treatment.

Michael Truell (@mntruell) fell in love with coding at 12. The company he co-founded, @cursor_ai, went from 15 people to 700 in two years. Today, over 60% of the Fortune 500 build with its AI coding platform.

@the_smart_ape You're evaluating future monopolies using static P/S ratios. Valuation looks forward. SpaceX is the only reliable gateway to space, the forefront of exploration, the pioneer of asteroid mining, and the architecture for Kardashev-scale solar capture. You pay for tomorrow.

> nvidia → ~$216b in sales → worth ~$4.5t → you pay $22 per $1 of sales > amazon → ~$650b in sales → worth ~$2.3t → you pay $4 per $1 of sales > tesla → ~$95b in sales → worth ~$1.6t → you pay $16 per $1 of sales > s&p 500 → you pay ~$2.5 per $1 of sales > spacex ipo → $18.7b in sales, losing $4.9b a year → worth $1.77t → you pay $94 per $1 of sales and everyone is calling it an opportunity.

when was the last time you wrote a code without using AI? 🤔

Gatethropic

@bcherny Biggest gate kept launch in history. Fable wont even fix bugs it wrote without flipping to opus.

Fable 5 is the biggest step up I’ve felt in our models since Opus 4.5 back in November. After 4.5 came out I uninstalled my IDE when I realized that I’d been doing 100% of my coding in a terminal for a few weeks. With Fable, it’s felt like Claude has stepped up from being a coding agent to a thought and design partner in building the product. Fable has judgement, taste, and dimensionality in a way that previous models didn’t, leading me to trust it more with the most complex work. I think the first time I had this realization was when I asked Fable to debug something. It is the first model I have used that was so methodical and precise, taking measurements and adding logs then verifying that it truly fixed the issue before declaring victory. There’s nothing in claude code’s prompting telling the model to do that, it’s just part of its personality. It really has this “big model smell” that I haven’t felt before.

@bcherny Fable cant even work on code Fable wrote without reverting to opus.

Enjoy!

@AnthropicAI turn down the gatekeeping brothers. Fable keeps switching to Opus on a product Fable built. Legit just building a ui/ux here nothing is of security risk. I flagged two bugs 1 architectural and another real one in its build and its just forcing opus. shit product.