Irish Larry David

I'm glad we could agree on this.

@BaronVanBurgum @Devon_Eriksen_ @danzu72 This is also any academic, scientist, or medical researcher working on curing diseases. "Such scum, amiright?"

I worked for USAID when It was closed, these guys literally were firing senior Foreign Service Officers with 30 years of experience on the spot. These kids were drunk on power and a reckoning is coming. I will always be hot about this shit. I know for a fact we allowed children to die because of these snotty shits.

Expectation: the age of the IDE is over Reality: we’re going to need a bigger IDE (imo). It just looks very different because humans now move upwards and program at a higher level - the basic unit of interest is not one file but one agent. It’s still programming.

@NicoleBehnam C'mon. What should be obvious is that this is mostly a problem of what the 'high agency, mostly well-off' are losing, not what the 'destitute stand to gain.' Are there any real examples of the poor making real gains with AI?

Insane that people who could benefit most from AI (people without access to expensive lawyers, therapists, etc.) are being algorithmically programmed to hate it most. A single mom working two jobs who could use AI to draft a legal letter, navigate a benefits system, get tutoring for her kid, or start a side business is being algorithmically fed fear content that keeps her from picking up the most democratizing tool since the internet. Don’t let it be you

@Samuel_Gregson Roko, a while back: 'Every human should spend their time making money and building status.' I responded with something like: 'Newton, Turing, Einstein, King Jr., Curie, Salk, and countless other luminaries who changed the world beg to differ.' His own followers ratioed him.

Sometimes I envy people who are simultaneously absurdly confident and incredibly poorly informed.

@Noahpinion @benthompson @deanwball Alex Karp, essentially: "Thiel and I are happy to allow our company to be nationalized by the government we bank-rolled and basically control."

Honestly, in the @benthompson vs @deanwball debate, I think Ben is right. There was just no way America -- or any nation-state -- was ever going to let private companies remain in total control of the most powerful weapon ever invented.

@BoringBiz_ Not enough people ready for the conversation of how to transition from a labor-based consumer economy to whatever kind of utopia-economy we're supposed to have with an AI revolution, apparently

Maybe white collar office jobs were just a small temporary blink in history between the Industrial revolution and the AI revolution Daycares for adults until AI eventually took over Not enough people are ready for that conversation

“Students and families with the ability to absorb really high token costs to become truly AI-native and capitalize on all of these AI products have a disproportionately outsized advantage against those who don’t.”

@Noahpinion A maximizer, but it's vibes, not paperclips.

The X algorithm now is basically: "Hey, I see you liked a tweet! I bet you'd like your entire feed to be 1000 other people tweeting the exact same thing!"

@VladTheInflator What's the rationale behind this? Elon essentially had the keys to the fin system of the U.S., fired a bunch of people who needed to be immediately re-hired, exposed all of our data to insecure servers, killed countless impoverished foreigners, and didn't reduce spending at all

Hard truth: We need Elon to be a trillionaire or we're likely all fucked. Look how bad a few low billionaires have made life in the US.

@PeterDiamandis There's absolutely no empirical evidence that an "explosion of jobs" is on the horizon. If that was the case, AI maximalists would be posting it incessantly.

AI job losses are inevitable, just like with the industrial revolution. However, we're about to see a CAMBRIAN explosion of jobs we can't even imagine could exist.

@krassenstein @elonmusk He killed hundreds of thousands when he one-shotted USAID. Funny that he suddenly cares about children.

@elonmusk Are you asking that about iran after 160 children have been killed?

How many more children must die before action is taken? Enough is enough. Act now.

@thebadstats What's represented by "Goliath" and what has become "visible" at this point? Bret's delusion and grift is beyond parody.

I have a feeling we're going to be perpetually entering "Endgame" for the rest of Bret Weinstein's career as a pundit

@DaveShapi "...Anthropic escalated, drew confusing and unnecessary lines in the sand, and doubled down." Just a bad take. 'No surveillance. No autonomous killbots.' Doesn't seem like "confusing an unnecessary lines." If it was in their contract, DoW wouldn't have had to fight them on it

After a day of backlash, debates, and research, here's what I've come to on the Anthropic vs Pentagon situation: The fundamental issue is about procurement. The Pentagon has every right to ensure that their contractors meet specification. Anthropic's protests don't really make any sense. They have a partnership with Palantir, which conducts mass domestic surveillance. Furthermore, Claude is literally incapable of directing autonomous weapons right now. Dario "clarified" that he's actually okay with autonomous weapons, but that Claude "isn't ready" - but this really isn't their concern. The military is the one who decides when a tool is ready. Beyond that, the negotiation was private, with only a few minor leaks that no one really cared about, until Anthropic blew the lid off everything. They thought they could muster public support. They were the ones to escalate in a deep inappropriate way. Can you imagine if Lockheed did something like this over a next gen fighter? Next, the current administration decided they weren't going to be mogged by a private company and reacted in kind. So, you could say this is a case of FAFO. Anthropic escalated, drew confusing and unnecessary lines in the sand, and doubled down. Now, with all that being said, there is broad consensus that the "supply chain risk" was overkill, an outright "nuclear option" that may be illegal. However, legal analysis suggests that it will take several weeks or months just to get injunctive relief from this designation, and 1 to 3 years to litigate the issue. Furthermore, since the President himself has personally lashed out at Anthropic, there's likely almost nothing they can do to get back in the good graces of the government. Even if the supply chain risk doesn't stick, they're almost certainly out of the government. If the supply chain risk designation does stick (which there is a non zero chance of) then Anthropic cannot structurally survive in the long run. They will be relegated to a relatively small section of the economy compared to their competitors. However, this outcome seems unlikely. Even so, there's no way for them to compete with OpenAI, xAI, and Google, all of whom have signaled they will comply with Pentagon procurement requirements. Over time, Anthropic will fall to the back of the pack. Now, for my take, I'm a "structural realist" My view is this: the world is materially better if Anthropic has a seat at the table. America is better off with multiple competitors with such fundamentally different approaches to AI and alignment. While I am, and remain, highly critical of the direction that Anthropic is going in, I still believe that their contribution to the discourse is strongly net positive. It would be even more positive if they continued to work with the Pentagon. However, I do not see that as a viable path now. To get back in the good graces of this administration, they will need to demonstrate maximum contrition. As of Dario's interview this morning, that seems unlikely. He might even need to step down as CEO to convince the Pentagon to work with Anthropic. However, the corporate structure of Anthropic will make it exceedingly difficult to compel his resignation, and would take too long anyways, leaving voluntary departure as the only realistic pathway to contrition as far as I can tell. But again, I highly doubt Dario will go that way. Dario's deliberate escalation and subsequent gambit was clearly a miscalculation, which will have a chilling effect on any other labs that might want to play hardball with the government. On that point, I would not be surprised if this administration holds the line against Anthropic just to make a point. Trump has already ordered the entire federal government to stop using Anthropic, and this does not seem like it will reverse when OpenAI, xAI, and Google are ready to go. AI is fungible. Finally, I've been personally accused of all kinds of things given this structural realist position. The most common indictment is having "no principles" which I categorically reject. My principle is that the Western way of life is the most just, productive, and generative civilizational pattern that exists today. That includes America and most of Europe as well as our allies. Therefore, my position is that we should push for policies that strengthen the West. What has played out over the last few days has been a net negative to our way of life. We have materially lost future optionality. In short, I believe that a world in which Anthropic remains embedded with the Pentagon is the optimal policy, and I'm frustrated and disappointed that Dario would rather torpedo his company based on confusing and seemingly arbitrary "principles" rather than play ball. To that end, I've levied numerous hypotheses as to why Dario made this choice. Beyond the obvious strategic miscalculation, the best I can figure is that he followed the typical Effective Altruist script which advocate for creating maximum noise and trying to seize control over the narrative, rather than looking at structural incentives, market dynamics, and systems of power. In my dealings with EA types, they almost always reject realism in favor of idealism, often to their detriment. This pattern is deeply overdetermined by their epidemics and tribal values. I would be glad to be wrong on this. As much as I have become skeptical of Anthropic, I would prefer them return to the fold.

This is why I’ve been so pissed at all these people attacking Anthropic on moral grounds. Claiming they’re the bad guys, the worst of the worst, etc? Really? The people doing the most on AI safety all this time? The people publishing their own mistakes the most? And now the people holding the line on how their tech can be used. Be pissed all you want about their prices, or their handling of the subscription communications. But don’t conflate your desires to use their services at other companies’ prices with them being morally rotten. Far as I can tell they’re the cleanest out there in terms of having a pro-human goal and actually doing what they say.

@Xenoimpulse "Technocrats," in this case, seems to be limited to people with financial interests in the accelerated buildout of AI. I've seen a number of AI/ML & CS engineers echoing the demand for pro-social AI. Seems obvs that the populist take isn't limited to "those without a STEM CV."

@DeepDishEnjoyer The Competing Autisms War of 2026-2028

do you know how fucked up your industry's culture needs to be for an autistic triangular hamster to be the voice of reason

@Vivek4real_ That was a lot of dancing around to say, 'Basically, only poor people eat our food, and the Poors now are too poor to buy our food. '

The CEO of McDonald's says that we're now living in "a two-tier economy." " If you're upper income earning over $100,000 things are good, stock markets are near all time highs... What we see with middle and lower income consumers is actually a different story."

Software? Zero Card networks? Zero DoorDash? Zero Your job? Zero Your marriage? Zero The top of the K? Zero The bottom of the K? Zero The entire Indian subcontinent? Believe it or not…Zero

Very well-written and very depressing science fiction. Worthy of your time to read and consider

@extradeadjcb The "argument against AI" isn't against the agenda of a Superhuman AI. That's beyond the horizon, within hypothetical territory. The actual argument is directed towards the owners and beneficiaries of AI as it exists today, and how they unilaterally changing the econ landscpe.

The "permanent underclass" meme makes no sense Why would a superhuman AI care who its "equity holders" are & if you get a lighter singularity, & cost of capital just goes to zero/return on capital goes infinite, no one is going to care that Sam Altman has more infinity than you