Dan @robustus

@CarlosFArellano @KarolCodes Oh lol I just realized I replied to you instead of someone who QT you with a very different point. Apologies.

Anthropic using the fear monger for regulatory capture strategy is such a shame

@CarlosFArellano @KarolCodes 1) I did not say it's AGI 2) That is irrelevant to my post anyway.

@capitao_cripto Oh no, high bar! I will no doubt disappoint in the future. Apologies in advance, but thanks for the kind words.

NewLimit made major progress in the past year toward our mission of radically extending human healthspan. Check out this talk where we run through what we've built, and our vision for where it's going.

Jacob Kimmel @jacobkimmel

a day ago

every year @newlimit, we host friends at the lab & share our progress for 2026, we announced: - 2X increase in discovery rates with AI systems - new program for endothelial cells - 0 -> 1 medicines headed to the clinic - accelerating recovery from alcohol some highlights --

12 24 203 202K 61

Always be suspicious of government regulation. And *especially* of people who think they have a lock on a superior morality calling for government regulation or intervention.

FinancialJuice @financialjuice

5 hours ago

Anthropic’s Amodei calls for tighter AI safety regulations

2 1 20 17K 2

@mert @KyleSamani I feel like you either have the gene to appreciate sov assets and the sov thesis, or you don't. Kyle has demonstrated that he doesn't from the beginning. Fine, but might want to recognize that other peoole *do*. Gold at $30T demands the ack.

(Tweet is (obviously) satire and a subtweet of macro-twitter)

Dude, Where’s My Fable-Driven Productivity? If Fable (Mythos) is the best AI model in history as many predict and markets increasingly price in, it’s odd that 3-hours in, developed world macro productivity data shows no change in previous trends.

Yeah. And while I do think Anthropic's intentions are genuine, have to remain aware that history is full of entities who thought they had a lock on morality ultimately creating some of history's worst horrors in the end.

Matt Van Horn @mvanhorn

21 hours ago

This is wild. "Fable 5 feels like a preview of AI inequality". I felt this at a micro-scale years ago when we were in an email argument with our childcare and I had paid ChatGPT and she had free ChatGPT. It was like I had an expensive lawyer making my arguments and hers was a

50 43 553 59K 190

@ledgerstatus Wow. Having used a tractor to make some minor trails through the woods, I find this pretty impressive.

Our Anthropic overlords deciding which prompts the peasants are allowed to use.

@ItsSnibby Obviously satire. And a subtweet of a prominent macro guy on here.

Not sure you actually read/understood my tweet. But if you did, then you've demonstrated a complete disregard for the last several hundred years of data, over which human flourishing expanded tremendously due to free markets. You can see the data at humanprogress.org/datasets/ And if you didn't read my tweet, then obv you're just here as some sort of data-blind keyboard warrior. Either way, I'm blocking you now cuz the expected value to me of further interaction with you is clearly negative.

No, it's worse than this. There should much less pie *total*, not only that the politicians get way too much of it.

Burnie Senders @BurnieSendersX

a day ago

Equal distribution is one of our greatest achievements, comrades ☭

82 252 2K 441K 75

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner. Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky. When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit. We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted. In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation. H/T to colleagues that shared this with me socket.dev/blog/mini-shai…

@TheStalwart We saw the other tweet

here are some facts about California. some of this is hard to believe. first of all, it's important to understand the concept of "ballot harvesting," which is perfectly legal in CA. this refers to a situation where someone completely unaffiliated with the voter can collect and submit their ballot for them. this flow is completely legal: - a homeless person arrives in LA, where they are eligible for cash assistance, SSI, food stamps, healthcare through medical, and an array of other taxpayer-funded services - they are registered to vote by an NGO (many such NGOs exist and explicitly do this). - they do not have to provide a residential address or any proof of residency to vote. they only have to provide a mailing address, which can be anywhere (church, NGO HQ, homeless shelter). their home address can be "a park" or "an underpass". - their ballot is mailed to the homeless shelter (or whatever address the NGO elects for them) - the only verification done for the mail-in ballot is "signature verification" and uniqueness (only one vote per person is counted theoretically). - the signature can be an X. if they register with an X, they can sign with an X. that is sufficient to pass verification. signature verification is also deliberately loose. the signature does not have to be a perfect match. now consider the hypothetical scenario, which is fraudulent, but virtually impossible to detect: - a homeless person cycles through the LA system. they get registered with their mailing address listed as the NGO HQ or homeless shelter - they "sign" their registration with an X or nondescript, easily replicable signature - they disappear. never seen again. or they exist, but it doesn't matter. they don't get purged from the voter rolls for 4-8 years typically. - the address where they registered receives their ballot for several cycles - operatives are aware that they have X amount of votes to make up. they fill in X many thousand mail-in ballots themselves. the ballots are manually postmarked (permitted). they forge the signature to match whatever signature (could be an X) was submitted upon registration - ballots can be accepted even if they are postmarked at 11.59 pm. polls closed at 8 pm. (you would need an accomplice who is a USPS employee) - the only fraud checks are de-duplication (if the homeless person through some miracle voted in person, only one of their ballots would be counted) and signature verification - because very few of the homeless people in question would have voted in person, this gives NGO operatives tens of thousands of possible mail-in ballots to submit unilaterally. the big problem is that there is NO way to detect this type of fraud. NGOs that register homeless people to vote exist. that isn't a secret. ballot harvesting is fully legal. voting by mail is encouraged. signature verification is as loose as possible. de-duplication doesn't solve anything, since few homeless people vote in person. and no one in power locally is going to spend political capital on rooting out such fraud, since they are all wholeheartedly committed to "voting rights". in a situation where fraud is undetectable, the absence of hard proof of fraud is not evidence that no fraud exists. Raman has gained around 20k votes since election night. She is around 3k votes ahead of Pratt now. there are over 72 thousand homeless people in LA county.

@TheStalwart Those are not good. Possible exception for #9.

@mikealfred And if they *are* wealthy: x.com/i/status/20607…

Dan @robustus

2 weeks ago

Traders on here talk a lot about short term trades, implying that's how they got wealthy. But I think when you push most successful "traders" on where their wealth came from, it's cuz they had 1 or 2 home-run positions that they held for years, on mostly fundamental grounds.

8 6 76 7K 9

Sure, but what they have now is basically what I'd have! (tho tbf I skimmed some sections of the current one, so do that hold me to it). Whereas I never would've produced that earlier one for obvious reasons *that should've been obvious to them too*. The fact that it wasn't obvious to them gives me pause about them in general.