I just wrote down my thoughts on context engineering every sota model has its quirks, but they're all pretty capable the difference between ai slop and magical experiences is the context you give to the model

Designing for autonomy isn’t a tweak—it’s a transformation. In “Designing for Autonomy: How Agentic AI Will Reshape Enterprise Architecture,” we explored how businesses must reimagine their tech stacks to support agents that can observe, decide, and act without human prompting.

IBM sees enterprise customers are using 'everything' when it comes to AI, the challenge is matching the LLM to the right ... venturebeat.com/ai/ibm-sees-en…

Open, closed, or hybrid? When it comes to LLMs, there’s no one-size-fits-all. In “Choosing the Right Model Strategy for Enterprise AI,” three tech leaders got real about the trade-offs in building multi-model AI stacks: - Xuedong D. Huang, Chief Technology Officer at Zoom -

Why at 𝗜𝗕𝗠 𝗮𝗻𝗱 𝗥𝗲𝗱 𝗛𝗮𝘁 we are very bullish on 𝘃𝗟𝗟𝗠 and why you should be too If you’re serious about AI infrastructure, vLLM isn’t just a nice-to-have; it’s rapidly becoming the new default. Here's why: 𝟭. 𝗜𝘁 𝘂𝗻𝗹𝗼𝗰𝗸𝘀 𝗿𝗲𝗮𝗹 𝗚𝗣𝗨 𝗥𝗢𝗜 LLM

It’s been one year since 𝗜 𝗯𝗲𝗰𝗮𝗺𝗲 𝗮 𝗨.𝗦. 𝗰𝗶𝘁𝗶𝘇𝗲𝗻. I’ve been meaning to write about it, not the high-level announcement version, but the real extended story. The personal one. The journey I went through and the factors I considered before making the decision.

This is the technology behind the Wimbledon Digital Experience - Hybrid Cloud Architecture with IBM Cloud and AWS working together - Red Hat OpenShift as the universal deployment and orchestration layer - IBM watsonx AI stack for data, governance, orchestration, and NLP -

Imagine this: an AI with Grok 4-level performance but fully open source. Right now, I only see one player bold enough (and resourced enough) to make it happen...

Grok CLI >>> Cursor and Claude Code We wanted an IDE for xAI's Grok so we did something meta: we prompted the newly released Grok 4 to create... Grok CLI itself! Grok CLI can: 1. Modify local files and use the shell 2. Go through huge codebases and fix them 3. Persist for

This was a historic week for AI policy. President Trump gave his first major speech on AI, the White House released its AI Action Plan, and President Trump signed three important executive orders. If you missed it, here is a recap of the major themes and actions. 🧵

GPT-5 is here. Rolling out to everyone starting today. openai.com/gpt-5/

new waymo next to old waymo. spotted in SF today

it is the first time in a more than 15 years i consider replacing my iphone very impressed with pixel 10 and gemini integrations

AI agents are only as good as the facts they’re grounded in. I spent some time checking my friend's cool project, 𝐚1𝐟𝐚𝐜𝐭𝐬, a framework that lets you build agents on top of a knowledge graph, define your domain in plain YAML, ingest structured facts, and query them in

having a founder still deeply involved and pushing hard is a moat #ORCL

Meta just unveiled AI glasses with a built-in display, controlled by a band that reads muscle signals. I sat down with Mark Zuckerberg to cover how these glasses could replace your phone, superintelligence, the metaverse, and more. 0:00 Intro 1:07 Meta's new glasses revealed

Welcome back Picasso😭

IBM is back! They just joined Hugging Face Enterprise & released Granite 4.0 in open-source with a new hybrid Mamba/transformer architecture that reduces memory requirements without reducing accuracy much. This set of models is great for agentic workflows like tool calling,

Today is my last day at IBM. Grateful for the people, the learnings, and the chance to help shape the future of enterprise AI. Thanks to everyone who made the ride unforgettable 💙

Angel Island might be one of Bay Area’s best kept secret. A 10-minute ferry from SF or Tiburon and you get nature, history, and unreal views. A reminder why California is unbeatable.