Miguel Fierro

#AI, #Language, and the Future of Understanding with Pilar Manchón at Google It was great to explore the evolution of language models, the challenges of black-box AI, and what’s next for human-machine interaction in this amazong podcast! https://lnkd.in/dtpMKYKj

55

3 comentarios

A few days ago, Gary Marcus published a thought-provoking post arguing that many of today’s most advanced AI systems already qualify as neurosymbolic AI -- not because of what’s inside the model, but because of how they interact with symbolic, often deterministic tools. We tend to associate neurosymbolic AI with architectures that embed symbolic reasoning within the model itself. But Marcus makes the case that tool-using LLMs (systems that call out to code interpreters, search engines, and calculators) are just as much in that tradition. The symbolic logic may live outside the model, but it’s doing real work in shaping the system’s behavior. Set aside the baggage of the Marcus vs. LLM-world debate for a moment -- whatever side you take, he’s hitting on an important point. The reliability of LLM-powered systems is being driven not only by big improvements in model performance, but also by architectures that connect those models to external tools (e.g., web search, code interpreters), many of which add symbolic reasoning, verification, determinism. Looking at Grok 4 and Grok 4 Heavy, these made a splash this week with SOTA results on key benchmarks. But when you look closely, you see that performance gets a big boost when the models are allowed access to tools, especially those with deterministic logic like a Python interpreter. That’s a neurosymbolic system, whether or not the model internals were designed that way. This has me thinking about architectural paths forward for improving LLM reliability and security in concrete enterprise contexts. The big question I’m thinking about is: How far can generalist neurosymbolic architectures take us on reliability and security, versus approaches that anchor LLMs in domain-specific workflows and logic? Generalist systems are exciting if they generalize well. But in high-stakes and high-volume domains, we may still need tight coupling with deterministic layers and trusted domain-specific workflows to get the reliability and trustworthiness we need at scale. My question for Gary and others who've been in this space for some time: Are there examples where generalist neurosymbolic systems give us strong, niche reliability guarantees? Or is that still an open research question? Link to Marcus's post: https://lnkd.in/d6U_MSjK #artificialintelligence #ai #trustworthyai #aisafety #aisecurity

107

34 comentarios

Heard about grok- xAI's recent announcement in adding 12 Billions $ (targeting 1 M GPUs) to build the next versions of LLM ? in this second post about LLM, let me share you the building blocks of LLM and how this helps in making state of the art models. This building blocks can help you in making key decisions - right LLM choosing for your business usecase, fine tuning the model (LoRA) and mainly how this huge influx of money can help getting better model. In today’s LLM pipelines, context windows, multi-head self-attention, and transformer layers define how models generate coherent responses. But these models — like Grok, GPT-4, Claude, or LLaMA — aren’t magic; they’re built from embeddings, forward/backward propagation, query-key-value attention, and positional encoding. Understanding how data flows through stacked layers, how gradients shape the model in training, and how context length impacts performance is key to making informed architectural, fine-tuning, and product decisions. Whether you’re managing compute cost, designing RLHF loops, or debugging drift, this technical understandings is no longer optional. That’s where the #Stanford CME 295 LLM Cheatsheet (attached) becomes a game-changer. It distills the core building blocks — from tokenization and embeddings to cross-attention and transformer math — into digestible visuals and definitions. it’s a perfect starting point for engineers, product leaders, and researchers looking to move beyond the hype and into technical clarity. let me know how you are using #GenAI especially LLMs in your usecases. #LLM #Elon #Data #GenAI #xAI

23

1 comentario