Steph S. ♻️

If we write down 3 new things we are grateful for 21 days in a row – 3 new things each day – our brain starts to retain a pattern of scanning the world not for the negative, but for the positive first. ted.com/talks/shawn_ac… #Happiness #Wellbeing #Kindness #Success #Gratitude

Phases to master Agentic AI amzn.to/4nolZ2l

This is a step-by-step roadmap to building your first AI agent. Save this for later.

How AI/ML Models are Built and Deployed in 12 Steps? by @techyoutbe #ArtificialIntelligence #MachineLearning #ML #Technology cc: @ylecun @ogrisel @meisshaily

Today we are launching my favorite feature of ChatGPT so far, called Pulse. It is initially available to Pro subscribers. Pulse works for you overnight, and keeps thinking about your interests, your connected data, your recent chats, and more. Every morning, you get a custom-generated set of stuff you might be interested in. It performs super well if you tell ChatGPT more about what's important to you. In regular chat, you could mention “I’d like to go visit Bora Bora someday” or “My kid is 6 months old and I’m interested in developmental milestones” and in the future you might get useful updates. Think of treating ChatGPT like a super-competent personal assistant: sometimes you ask for things you need in the moment, but if you share general preferences, it will do a good job for you proactively. This also points to what I believe is the future of ChatGPT: a shift from being all reactive to being significantly proactive, and extremely personalized. This is an early look, and right now only available to Pro subscribers. We will work hard to improve the quality over time and to find a way to bring it to Plus subscribers too. Huge congrats to @ChristinaHartW, @_samirism, and the team for building this.

Quantum experiment shows entangled light can outpace classical systems by millions of years. bit.ly/46rHxVK

Vibe coding won't save your career. Those, who want to control AI must understand engineering. And it’s not rocket science. The best free resources to get started: - System Design 101 by ByteByteGo (GitHub): github.com/ByteByteGoHq/s… - Engineering Visual Guides: bytebytego.com/guides/ - Become a Supabase Pro in 1.5 Hours (1:26:39): youtu.be/dU7GwCOgvNY - Overview of HTTP by Mozilla: developer.mozilla.org/en-US/docs/Web… - HTTP Request Methods: developer.mozilla.org/en-US/docs/Web… - The most important concepts of Node.js and Express.js (17:09): youtu.be/2YIgGdUtbXM?si… - Every React Concept Explained in 12 Minutes (11:52): youtu.be/wIyHSOugGGw?si… - TypeScript - The Basics (12:00): youtu.be/ahCwqrYpIuM?si… - CSP Guide by Mozilla: developer.mozilla.org/en-US/docs/Web… - OWASP Top 10: owasp.org/www-project-to… - Getting Started with Lovable: docs.lovable.dev/introduction/w… - Replit AI Agent, Full Course: youtu.be/DaXQ5L7r7Lg?si… - n8n Masterclass: youtu.be/AURnISajubk?si… - How to Integrate Supabase with Clerk: clerk.com/docs/integrati… ------------ Guides for those building AI-powered products: - How To Build, Deploy, And Scale Your AI Product Strategy: productcompass.pm/p/openai-how-t… - 14 Prompting Techniques: productcompass.pm/p/prompting-te… - AI Evals: productcompass.pm/p/ai-evals - Error Analysis: productcompass.pm/p/evaluating-a… - A Guide to Context Engineering: productcompass.pm/p/context-engi… - AI Agent Architectures: productcompass.pm/p/ai-agent-arc… - 17 Penetration & Performance Testing Prompts: productcompass.pm/p/penetration-… - 3-Layers Distribution Framework To Win Market Share: Subscribe to get notified: productcompass.pm/subscribe ------------ Finally, learn by doing: (no coding, understanding tech) - LLM Chatbot: productcompass.pm/p/llm-api-fram… - RAG Chatbot: productcompass.pm/p/how-to-build… - Voice Agent: productcompass.pm/p/n8n-mcp-serv… - Multi-Agent Research System: productcompass.pm/p/multi-agent-… - Separate DEV/TEST/PROD: productcompass.pm/p/lovable-bran… - Full-Stack App with Lovable (2:00:00): productcompass.pm/p/full-stack-a… ------------ [Bonus] My favorite tools: - Database: Supabase - Vector store: Pinecone - Coding agents: Lovable (easiest), Replit, Cursor (most advanced) - Hosting: Netlify, Vercel - Authentication: Clerk - Hardening: Cloudflare - Transactional emails (avoid SPAM): Postmark - SEO for SPA apps: Prerender - Cache: Redis + Netlify - Storage: S3 - Analytics: Clarity, PostHog - Logging: Logtail - Metrics: Grafana - Feature flagging: GrowthBook - System status: Uptimerobot - Payments: Stripe - Agentic workflows: n8n (easiest), LangChain (most advanced) - Evals off-the-shelf: LangSmith ------------ What did I forget?

LIMI: Less is More for Agency arxiv.org/abs/2509.17567 ✍️ Keyu Li et al. via @arxiv H/T @connordavis_ai 👉 "This paper's findings establish the Agency Efficiency Principle: machine autonomy emerges not from data abundance but from strategic curation of high-quality agentic demonstrations" 💡"The age of "thinking AI" is over. The age of "working AI" demands strategic intelligence, not brute force scaling" @Corix_JC @jeanyvesgonin @sonu_monika @JagersbergKnut @ahier @sim010101 @maponi @EstelaMandela @Shi4Tech @CEO_Aisoma @SpirosMargaris @IngridVasiliu @dinisguarda @mvollmer1 @RamonaEid @ALouiseLavalett @Ym78200 @ChuckDBrooks @FernandaKellner @PVynckier @JoannMoretti @NeiraOsci @tlloydjones @theomitsa @SusanHayes_ @TarakRindani @FrRonconi @sulefati7 @Nicochan33 @mikeflache @Khulood_Almani @TysonLester @CurieuxExplorer @Eli_Krumova @amalmerzouk @smaksked @pchamard @MaryRich78 @Analytics_699 @MHcommunicate @TheAIObserverX @NathaliaLeHen @sminaev2015 @jeancayeux @WillyRayNick @DanielleLargier @RLDI_Lamy @richardkimphd

Pew have updated their ChatGPT polling from a year ago. Usage has roughly doubled since summer 2023. The share of employed adults who use ChatGPT for work has roughly tripled over the same period to 28%. Two thirds of US adults have still not yet used ChatGPT.

🇺🇸 Today is a day we have been working towards for six months. We are announcing America’s AI action plan putting us on the road to continued AI dominance. The three core themes: - Accelerate AI innovation - Build American AI infrastructure - Lead in international AI diplomacy and security. It’s a long document (!) and I highly encourage you read it. There is a lot of exciting actions in here but one I’m very partial to is the focus on open source and open weights and making sure the U.S. leads in this critical area. This represents the work of many many people but I want to particularly highlight @DavidSacks @mkratsios47 and the person who has spent the most time and sweat and effort on this : @deanwball Personally, this is an effort I’ve been incredibly proud to have been a part of and why I signed up for this job. I’m confident this and the other announcements today put us on the road to AI dominance as called for by President Trump.

New roles for #AI by @Khulood_Almani #DataScientist #BigData #ArtificialIntelligence #ML #MI cc: @meisshaily @ylecun @JimMarous

Today we’re releasing GitHub Spark — a new tool in Copilot that turns your ideas into full-stack apps, entirely in natural language.

If #AI can write a song that moves you and no one tells you it was synthetic, did it matter who made it, or did it matter what you felt? #EmergingTech #Music #Tech #ArtificialIntelligence thenextweb.com/news/a-viral-b…

Browsers are becoming the frontline for #AI action and the frontline for risk. Granting AI logged‑in access makes your browser both agent and attack surface, unseen vulnerabilities could be the price of convenience. theverge.com/command-line-n… #EmergingTech #InfoSec #Privacy

Sam Altman is recasting #AI as a grassroots economic engine. His Fed-stage pitch reframes billions of prompts as proof that democratized AI can lift everyday productivity, turning users into co‑drivers of #innovation.  theverge.com/news/710867/op… #EmergingTech #Tech

So I gave Claude Code a Mac Mini. And it’s called Claudeputer. It runs 24/7 and it’s allowed to do whatever it wants - it’s in complete control of its computer. Watch for a 2min demo.

Robot swapping batteries on its own! 🔋 @UBTECHRobotics just gave humanoid robots the power to work non-stop — literally. The Chinese robotics firm has unveiled a world-first: a fully autonomous, hot-swappable battery system for humanoid robots. Their latest model, Walker S2, can now replace its own battery — without human help — in under three minutes, using dual-arm precision, advanced positioning, and adaptive algorithms. That means robots can now work 24/7 with zero downtime. Beyond just swapping, the system lets robots choose between charging or swapping depending on task urgency — enabling real-time, energy-aware decision-making. Most robots today shut down to recharge. Walker S2 just rewrote that rule.

BREAKING: OpenAI just launched ChatGPT Agent It allows ChatGPT to think, plan, and execute complex tasks on its own virtual computer while you do other things I had early access, and ChatGPT Agent built me a complete early retirement plan in 20 minutes: > Found local tax laws (Vancouver) > Analyzed average monthly spend rates > Calculated savings needed to retire at 30 > Researched optimal investment allocations > Found tax optimization strategies I'd never heard of > Built multiple FIRE scenarios > Created a downloadable presentation with results This would've cost me $5,000+ from a financial advisor and taken weeks I think with ChatGPT Agent now, and especially as it gains access to more tools, we're finally going to see the rise of a new AI skill category in *Agent Management* Agents are finally becoming capable of doing real work autonomously, so anyone who learns how to effectively orchestrate agents will have a huge advantage

The car dashboard is becoming a canvas for AI intent modeling. Unlike phones or desktops, it knows motion, urgency, destination, even mood… an underused signal set for building truly anticipatory software. pcworld.com/article/285068… #EmergingTech #AI

#Alexa+ is starting to feel less like a voice interface and more like a sidekick. It gets context, adjusts plans, and handles fuzzy requests, nudging smart assistants closer to real agency. pcworld.com/article/284928… #EmergingTech #AI

#AI can now spot hidden immune cell subtypes by reading shape, behavior, and location all at once. Tools like CellLENS might help scientists design #cancer treatments that actually match what cells are doing in real tissue. news.mit.edu/2025/ai-system… #EmergingTech #AI #MedTech