Brian Hacker

@haralabob Bingo!

The draft should be abolished, but since that won't happen... NBA Finals teams pick 30th and 29th Conference Finalists pick 28th and 27th Bottom 4 teams pick 26th-23rd (reverse order) Flat lotto odds for everyone else.

@GaryMarcus @BalvinderKalon What is that from?

@Hacker_Poker @BalvinderKalon but was drawing on this

Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a 3,167 line kernel called print.ts. print.ts is a pattern matching. And pattern matching is supposed to be the *strength* of LLMs. But Anthropic figured out that if you really need to get your patterns right, you can’t trust a pure LLM. They are too probabilistic. And too erratic. Instead, the way Anthropic built that kernel is straight out of classical symbolic AI. For example, it is in large part a big IF-THEN conditional, with 486 branch points and 12 levels of nesting — all inside a deterministic, symbolic loop that the real godfathers of AI, people like John McCarthy and Marvin Minsky and Herb Simon, would have instantly recognized.* Putting things differently, Anthropic, when push came to shove, went exactly where I long said the field needed to go (and where @geoffreyhinton said we didn’t need to go): to Neurosymbolic AI. That’s right, the biggest advance since the LLM was neurosymbolic. AlphaFold, AlphaEvolve, AlphaProof, and AlphaGeometry are all neurosymbolic, too; so is Code Interpreter; when you are calling code, you are asking symbolic AI do an important part of the work. Claude Code isn’t better because of scaling. It’s better because Anthropic accepted the importance of using classical AI techniques alongside neural networks — precisely marriage I have long advocated. It’s *massive* vindication for me (go see my 2019 debate with Bengio for context, or to my 2001 book, The Algebraic Mind), but it still ain’t perfect, or even close. What we really need to do to get trustworthy AI rather than the current unpredictable “jagged” mess, is to go in the knowledge-, reasoning-, and world-model driven direction I laid out in 2020, in an article called the Next Decade in AI, in which neurosymbolic AI is just the *starting point* in a longer journey.* Read that article if you want to know what else we need to do next. The first part has already come to pass. In time, other three will, too. Meanwhile, the implications for the allocation of capital are pretty massive: smartly adding in bits of symbolic AI can do a lot more than scaling alone, and even Anthropic as now discovered (though they won’t say) scaling is no longer the essence of innovation. The paradigm has changed. — *Claude Code is plainly neurosymbolic but the code part is a mess; as Ernie Davis and I argued in Rebooting AI in 2019, we also need major advances in software engineering. But that’s a story for another day.

@BalvinderKalon @GaryMarcus Response @GaryMarcus ?

print.ts is the terminal output renderer, not the reasoning engine. of course display logic has a ton of if-then branches for syntax highlighting, markdown, diffs, and ANSI codes. that's like finding a switch statement in Chrome's CSS renderer and concluding Google doesn't trust neural nets.

@phtevenstrong @mezzanine_fi Have you seen @OG_RogerTennis ? Guy seems like a jerk but his analysis is great!!

Lads, I've been posting defi yields since 2022. I'm an old man in the yield posting space. I "made it" as a farming KOL and am now happy to be transitioning into builder/founder and with @mezzanine_fi. Who are the new yield content creators in this space who are worth following? I would love to follow and boost.

@adamjohnsonCHI @MECAForPeace That's rad Adam. Well not the genocide, but the book.

Reminder: I have a book coming out in 12 days. It’s a data-driven and, I hope, thorough account of how US mainstream press sold the public on mass death in Gaza. Pre-ordering helps a lot so do that if you can. 100% of royalties go to @MECAForPeace amazon.com/How-Sell-Genoc…

@GergelyOrosz All AI companies already argue that they can steal work and can’t be stolen from. There is no cognitive dissonance. They don’t care if it’s consistent.

You can imagine Anthropic being in a pickle: 1. Do they just leave this, and look the other way, ignoring that it's not exactly fair to transform their code and leave it up there 2. Do they claim copyright applies... but this could be bad for their own business in much bigger ways: eg imagine regulation coming into play that bans this. Claude Code and other tools would have to refuse this kind of generation. Lawsuits against AI labs could spike etc So my bet is #1 happens. Not the interest of an AI lab to expand copyright protections to derived work cretated by an LLM...

This is either brilliant or scary: Anthropic accidentally leaked the TS source code of Claude Code (which is closed source). Repos sharing the source are taken down with DMCA. BUT this repo rewrote the code using Python, and so it violates no copyright & cannot be taken down!

@TheZvi He is denying that training data was stolen which seems patently obvious (using BT so as not to pay for vast content libraries) in some cases. Whether the tech will be found more broadly liable for plagiarism/theft seems a very open legal question. I smelled 6 feet of straw.

@Hacker_Poker As in, those in denial here (not all of anyone, of course, as always) are usually denying what AI can already do, whereas in other cases often people deny what AI will be able to do shortly (as Dean and I have a longstanding disagreement on).

To each their own denial, at whatever level of AI capabilities past, present or future, based on what their social, emotional or actual salary relies on them not believing. The version Dean is describing is unique largely because it is denying the past rather than the future.

@AriDavidPaul That’s crazy. Sorry to hear it brother.

Went to Bellagio, wired money ahead of time. Went to cage to get it. They asked me to sign a contract with MGM that allows them to do background and credit check stuff (even in my case, just wanting to wire in advance and pick it up.) The wouldn’t budge on the contract so having them reverse the wire. Is this the same at every Vegas casino?

🐐 shit

@Seanfrank Some use evaporative cooling. The water does not go back to the river or the golf course lol. It effectively vanishes. It does this in places where there is already drought conditions.

The “ai data centers are using all the water” thing was very radicalizing. I saw smart people, respected people, scientists- echo this back. You can not like data centers near you. You can complain they make electricity prices rise… But the water point is a total hoax. Every data center on earth uses less water than American golf courses. And the water isn’t polluted, it isn’t destroyed. It’s a little warm. IT COULD STILL BE USED ON THE GOLF COURSES IF YOU WANT. Your local McDonald’s is using more water than a data center. It’s just shocking how unreal and fake that narrative is. Radicalizing.

@GaryMarcus Have you seen this paper? arxiv.org/pdf/2509.15541

trust me bro, this may be missing a few zeros, but it’s a good meme

@tylermacro10 @whhatisachoice Think I should delete or?

@Hacker_Poker @whhatisachoice Dont show intelligence Its dangerous in this climate

Someone genuinely has to explain to me how truflation works

@whhatisachoice @tylermacro10 They haven't? What is the OP's graph measuring?

@tylermacro10 yes and the prices right now haven't adjusted, lets say your grocery store has a 6mo contract with sysco, sysco observes higher transportation costs, now what? now it waits until that contract expires to pass through those costs

@AviFelman These people were respected (probably incorrectly in a lot of cases) for their crypto takes. They history and political takes would be considered rudimentary and stupid if they were in 10th grade.

@divine_economy Most people in crypto have an extremely rudimentary understanding of politics and history. Ignorance runs rampant. Add racism and adoration for our deal leader and this is what you get.

nearly the entire og guard of anti-state crypto—udi, nic, bankless, ameen, toly—is now whiling away their days clamoring for war i say this with no great animus. it's just a testament to how profoundly out of touch this industry is with the world's ideals—and its own.

@sillytuna Sorry to hear it mate. WTF.

@dwpoker Same. Phone is completely unusable.

The spam calls are too much. Even with the Apple iOS filtering I get constant voicemails. Thinking of disabling voicemail. Any reasons why I shouldn’t do this? Can’t think of the last time I got a voicemail that mattered.

@phtevenstrong @WhiteHouse Good analysis. How do you feel about this? huffpost.com/entry/troops-b…

When was the last time Iran attacked American non-combatants on American soil? When was the last time Iran attacked American non-combatants directly anywhere? AFAIK, they haven't. And even if you use Hezbollah as a proxy, it would be 40 years since the last attack on U.S. non-combatants (Beirut). I don't support the Ayatollah (whose regime we are largely responsible for), but it is not clear to me that Iran was a threat to U.S. non-combatants.

President Trump isn't interested in endless wars, he's interested in ENDING decades-long chaos in the Middle East.

@BasicMountain @MartinJuza This isn't society writ large, there isn't a reason to give a second chance. If you get caught cheating twice, ban for life. And that might be one time too many.

@MartinJuza Anyone cheating that obviously on camera multiple times is an addict that can’t be “rehabilitated” by a suspension.

We need to punish people who do stuff like these vampiric tutor shuffle cheats harder. Ban for life. You had your chance and you decided to cheat. Gtfo of our game Surely we dont want to enter another era when the best way to improve your winrate is by shuffling your opps deck