AI Dynamics

Global AI News Aggregator

About

LLMS

  • API Provider Features and Prompt Caching in Deep Agents

    this is a good point around taking advantage of model/api provider features i agree that prompt caching is great! we make sure to use it in deepagents! but that alone doesnt lock in – you can switch pretty easily, just a bit more costly things like encrypted or serverside

    → View original post on X — @hwchase17

  • Thoth: Advanced Agent Harness with State-of-the-Art Memory

    check out thoth – agent harness with sota memory built on langgraph

    → View original post on X — @hwchase17

  • Claude’s Neurosymbolic Elements Revealed in print.ts Source Code
    Claude’s Neurosymbolic Elements Revealed in print.ts Source Code

    i don’t have the full source; print.ts is probably not the only neurosymbolic element, but here’s one account of what is in it via techtrenches.dev/p/the-snake…

    → View original post on X — @garymarcus, 2026-04-11 16:40 UTC

  • Memory Technology in AI Still Emerging and Unfamiliar

    Because memory is still so new, it’s not commonly known

    → View original post on X — @hwchase17

  • Claude Code: Neurosymbolic AI’s Vindication Over Pure Deep Learning

    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.

    → View original post on X — @garymarcus, 2026-04-11 15:27 UTC

  • Graphify: A New Knowledge Graph Tool for Claude Code
    Graphify: A New Knowledge Graph Tool for Claude Code

    Someone built Andrej Karpathy's dream tool 48 hours after he asked for it. Graphify is an open-source tool for Claude Code. Point it at any folder and one command builds a knowledge graph. It reads your code, docs, PDFs, and images. No vector database. No config files.

    → View original post on X — @alphasignalai

  • Your Harness Equals Your Memory in LangChain

    your harness = your memory if you'd rather read in blog form, link here: blog.langchain.com/your-harn… Harrison Chase (@hwchase17) x.com/i/article/204292509094… — https://nitter.net/hwchase17/status/2042978500567609738#m

    → View original post on X — @langchain, 2026-04-11 14:53 UTC

  • Practitioners Express Caution About LLM Hype and Limited Utility

    👇 “I still maintain that LLMs are really dumb and of limited use. it’s my opinion as a practitioner and professional that the hype being peddled by the AI corporations and self promoters (from X grifters right up to academia) are ill-founded at best and and downright dangerous at worst. i am yet to see any significant support for LLMs from practitioners with responsibility. the deeper they go the stronger their caution gets.” fj (@fjzeit) i’ve been using it in-depth for 2.5 years. i discovered the same approach that Karpathy calls “LLM Wiki” over 18 months ago (as have many others). i’ve built a hardware simulator, two operating systems, an embedded systems harness, an ai coding harness, a coding IDE, two games, and a programming language. i still maintain that LLMs are really dumb and of limited use. it’s my opinion as a practitioner and professional that the hype being peddled by the AI corporations and self promoters (from X grifters right up to academia) are ill-founded at best and and downright dangerous at worst. i am yet to see any significant support for LLMs from practitioners with responsibility. the deeper they go the stronger their caution gets. i look forward to the day when the considered opinions of practitioners gets as much airtime as those with a vested interest in the technology. — https://nitter.net/fjzeit/status/2042814019405513141#m

    → View original post on X — @garymarcus, 2026-04-11 14:52 UTC

  • LLM Infrastructure Stuck in 1990s Like Linux

    And just like Linux, LLM infra will never get out of 90s.

    → View original post on X — @tunguz