AI Agents ≠ one LLM They’re multi-layered systems: • General LLMs → reasoning
• Domain LLMs → expertise
• RAG → real-time data
• Tools → execution Real power = orchestration From answers → to actions. Via Giuliano Liguori (
@ingliguori
) #AI #LLM #AIAgents
MACHINE LEARNING
-
AI Agents Are Multi-Layered Systems Powered by Orchestration
By
–
-
GPT-4o and Llama 3.3 Pose Low Harm Risk Study Suggests
By
–
I think the fact that GPT-4o and Llama 3.3-80B did no significant harm is just as important as whether AI helped. If older (less accurate & more sycophantic) chatbots essentially did nothing for people who followed their advice, it means that there is less risk of harm as well.
-
Stanford’s Emma Lundberg on AI Transforming Biology and Open Science
By
–
"AI is changing the way we do biology," said @StanfordHAI Faculty Affiliate Emma Lundberg in @Stanford
's #ResearchMatter series. She's chosen to be in academia to advance open science – making data accessible to accelerate discovery for all. Read more: https://
news.stanford.edu/stories/2026/0
5/research-matters-emma-lundberg
… -
TinyFish Fetch provides clean content to reduce token costs
By
–
If your agent touches the live web, the token cost of raw HTML adds up fast. TinyFish Fetch hands your model clean content instead of markup soup. It’s available via REST, MCP, CLI, Skills, Python, TS SDK, and every major agent framework. Grab your API key for testing
-
Stanford Lecture Offers Deep Dive into LLM Architecture
By
–
Cette conférence de 2 heures à Stanford t’apprendra davantage sur la manière dont les LLM comme ChatGPT et Claude sont construits que ce que la plupart des personnes travaillant dans les meilleures entreprises d’IA apprennent durant toute leur carrière.
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 4 mai 2026
Ajoute-la à tes Signet🔖… pic.twitter.com/JOVAKrBsNSCette conférence de 2 heures à Stanford t’apprendra davantage sur la manière dont les LLM comme ChatGPT et Claude sont construits que ce que la plupart des personnes travaillant dans les meilleures entreprises d’IA apprennent durant toute leur carrière. Ajoute-la à tes Signet
-
Boston Dynamics Spot Uses AI for Whole-Body Tire Manipulation
By
–
Boston Dynamics’ Spot Masters Whole-Body Manipulation — Using AI to Drag, Roll, and Stack 15 kg Tires with Precision
— Ronald van Loon (@Ronald_vanLoon) 4 mai 2026
by @rai_inst#Robotics #Engineering #ArtificialIntelligence #Innovation #Technology pic.twitter.com/nQf4l1xOGSBoston Dynamics’ Spot Masters Whole-Body Manipulation — Using AI to Drag, Roll, and Stack 15 kg Tires with Precision
by @rai_inst #Robotics #Engineering #ArtificialIntelligence #Innovation #Technology -
GenLIP Trains ViT to Directly Predict Caption Tokens
By
–
"Let ViT Speak: Generative Language-Image Pre-training" Instead of contrastive image-text matching or adding a separate text decoder, this paper, GenLIP, trains a ViT to directly predict caption tokens from image patches with a standard next-token loss. A key component they use
-
Free Machine Learning & AI Resources for X Subscribers
By
–
The top 10 programming languages on GitHub over the past 2+ years. Notably, TypeScript has leaped to the top spot. v/
@github -
LimX Tron2 Humanoid Robot Designed for High-Volume Data Capture
By
–
Meet Tron2: LimX’s Next-Gen Humanoid for High-Volume #Data Capture
— Ronald van Loon (@Ronald_vanLoon) 4 mai 2026
via @CyberRobooo#AI #Robotics #MachineLearning #ArtificialIntelligence #ML #Technology pic.twitter.com/LUOABYXT4sMeet Tron2: LimX’s Next-Gen Humanoid for High-Volume #Data Capture
via @CyberRobooo #AI #Robotics #MachineLearning #ArtificialIntelligence #ML #Technology -
Handbook of Data Science and AI for Business: Generate Value from Data
By
–
The Handbook of Data Science and AI for Business — Generate Value from Data with Machine Learning and Data Analytics: http://
amzn.to/3Wc1Iiz