also “best case” really mean “shortest case”. whether it is *good* is a complex question!
@garymarcus
-
Gary Marcus Vindicated: AGI Requires Symbols and Neural Networks
By
–
. “
@garymarcus has won, others too” And the smart people already know it. AGI requires both symbols and neural networks. Period. -
Claude Code Validates Breakthrough in Neurosymbolic AI Hybrid Approach
By
–
Gary Marcus strikes again! He directly reveals the core truth after Claude Code source code leak: ✅ Claude Code is the biggest advance since the LLM era
✅ But it's not pure LLM, and it's not pure deep learning
✅ Core file print.ts has 3,167 lines, packed with if-then branches + deterministic symbolic logic Anthropicrely on classical symbolic AI at the critical moment to make Agent truly reliable. This move directly validates the Neurosymbolic AI (neural-symbolic hybrid) approach that Marcus has been advocating for over 20 years! Scaling is no longer the only answer; hybrid approach is the future The full long-form article is worth reading carefully 👇 — Gary Marcus (@GaryMarcus) 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 Geoffrey Hinton 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 to do an important part of the work. Claude Code isn't better because of scaling. [Translated from EN to English]→ View original post on X — @garymarcus, 2026-04-11 22:22 UTC
-
Search box improvement through stable economic implementation
By
–
if it works stably and economically they can put it in the search box. what’s there is utter garbage.
-
Search functionality deterioration raises concerns about tech products
By
–
Search never used to be THIS bad.
-
Critique of X Search Reliability and Its Failures
By
–
X search is completely unreliable. An undergrad could make a better system for classifying trends. And yet they are supposed to be part of a leading rocket-and-AI company? [Translated from EN to English]
→ View original post on X — @garymarcus, 2026-04-11 21:32 UTC
-
IP dispute over Uber demand prediction algorithms clarified
By
–
This is brazenly false, @vkhosla
. 1. To clarify my understanding, I followed up with someone who stayed at Uber longer than I was. The core intellectual property that we developed at Geometric (I was a co-developer thereof) was embedded in supply and demand prediction of pricing -
Claude Code Vindication: Neurosymbolic AI Emerges as Next Paradigm
By
–
Couldn’t agree more! We’re moving towards Nuerosymbolic AI – there’s masses just haven’t realized it yet, in the sense of understanding what the literature proposed and why the latest developments are starkly aligned Gary Marcus (@GaryMarcus) 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. — https://nitter.net/GaryMarcus/status/2042987819333738929#m
→ View original post on X — @garymarcus, 2026-04-11 19:48 UTC
-
Gary Marcus vindicated: Claude Code proves neurosymbolic AI superiority
By
–
Yep … @GaryMarcus has been right for 25 years; some AI Godfathers, not so much at all! #SaturdayAISurvey @AnthropicAI Gary Marcus (@GaryMarcus) 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. — https://nitter.net/GaryMarcus/status/2042987819333738929#m
→ View original post on X — @garymarcus, 2026-04-11 19:46 UTC