Today's AI Pulse
OpenAI is developing its own AI‑powered smartphone
OpenAI is reportedly building a smartphone that puts autonomous AI agents at its core, replacing the traditional home screen with an “Agent Task Stream” that can handle bookings, market analysis and more. The device is expected to run on custom MediaTek or Qualcomm silicon and be assembled by Luxshare, the same contract manufacturer behind iPhones.
Also developing:
By the numbers: Acko targets $2‑$2.5B valuation in upcoming IPO
Pt.2 Beyond Phishing: Cyber Threats in the Age of AI with Four Flynn
The post serves as a detailed guide to Professor Hannah Fry’s 2025 podcast series, spotlighting Episode 10 Part 2 which examines the human side of cybersecurity, the motivations of attackers, and AI‑driven defenses such as passkeys and risk‑based authentication. It also recaps the surrounding episodes, from the historical Operation Aurora case in Part 1 to discussions on AI‑generated worlds, drug discovery, robotics, health, philosophy, and the future of AI assistants. By weaving together these diverse topics, the author highlights how AI is reshaping both threats and solutions across multiple fields, while offering listeners a roadmap to the season’s interdisciplinary conversations.
Unlock Existing ChatGPT Auth, Not New Sign‑In Layer
Huge if true. Wrote a bunch about the Sign in with ChatGPT opportunity back in April. “Today, there’s no secure way to link a ChatGPT account to another tool. No OAuth. No scoped tokens. No way to persist context responsibly. What’s needed...
Real-Time World Models Promise Holodeck‑Level Robotics
OK @drfeifei was already near the top of my list of favorite people alive, but this pushed her up to the top. Infinite real time realities in a Holodeck. This work has sped up my expectations about robotics. @adcock_brett told @Benioff...
How Anthropic’s ‘Skills’ Make Claude Faster, Cheaper, and More Consistent for Business Workflows
Anthropic launched 'Skills,' a capability for its Claude AI that lets organizations package instructions, code, and reference materials into reusable, composable folders the assistant loads on demand, using "progressive disclosure" to avoid context-window limits. Skills work across Claude products and...
New Chapter Releases Symbolic Verifier for LLM Evaluation
Chapter 3 on building a symbolic verifier for LLMs from scratch is now live: https://mng.bz/lZ5B . And with this, the first 176 pages of Build A Reasoning Model (From Scratch) are now available. This verifier is a useful method for...
Spotify Says It’s Working with Labels on ‘Responsible’ AI Music Tools
Spotify announced a formal agreement with major labels — Sony, Universal, Warner, Merlin and Believe — to jointly develop “responsible” generative AI music products and to stand up a dedicated research lab and product team. The pact emphasizes upfront deals,...
Is Agentic Commerce a Thing?
Forrester analysts outline “agentic commerce” as nascent AI systems—composed of foundation models and task-specific agents—that can take actions to assist or conduct parts of the shopping journey, from discovery to, in rare cases, checkout. The firm distinguishes owned experiences (brand-controlled...
ACE Prevents Context Collapse with ‘Evolving Playbooks’ for Self-Improving AI Agents
Stanford and SambaNova introduced Agentic Context Engineering (ACE), a framework that prevents “context collapse” by treating an LLM’s context as an evolving, itemized playbook updated incrementally by Generator, Reflector and Curator modules. In evaluations ACE outperformed strong baselines—improving agent-task performance...
Inside the Web Infrastructure Revolt over Google’s AI Overviews
Cloudflare has automatically updated robots.txt files on roughly 3.8 million domains and rolled out a new Content Signals Policy—covering about 20% of the web—to let site operators opt out of AI uses (ai-input and ai-train) while distinguishing traditional search from...
Analytics: The Essential Tool for Content Marketers
🔴 If you're a content marketer, this is why you have Analytics! Some content gets real results. Some content …not so much. This is obvious to every marketer who's pushed the publish button. But it’s less obvious that content performs...
AI Partnership Accelerates Fusion Toward Unlimited Clean Energy
Super excited to be collaborating with Commonwealth Fusion Systems @CFS_energy to use AI to accelerate fusion development - and move closer to a sustainable future with limitless clean energy
Is Haiku 4.5 Really THIS Good? OpenAI's Erotic Mode & Are MCP Apps the Right Approach? EP99.21
The post introduces the "This Day in AI" podcast, hosted by Michael and Chris Sharkey, who present themselves as average tech enthusiasts sharing candid, humor‑filled discussions about AI experiments, tools, and mishaps. It emphasizes the show’s informal, non‑expert vibe, offering...
Funding General Intuition to Advance AI Spatial-Temporal Reasoning
Super excited about funding @PimDeWitte on this exciting journey towards AI understanding the physical world !
Real-Time 3D World Model Runs on One H100
Very excited to share @theworldlabs ‘s latest research work RTFM!! It’s a real-time, persistent, and 3D consistent generative World Model running on *a single* H100 GPU! Blog and live demo are available below! 🤩
Under the Hood of AI Agents: A Technical Guide to the Next Frontier of Gen AI
Agentic AI—LLM-based systems that autonomously run tools in a thought-action-observation loop—are rapidly moving from chat prototypes to production, enabling tasks from booking travel to coding. Key infrastructure components include agent development frameworks, cloud model hosting, tool-call protocols (notably the year‑old...
Dagger’s Container Use Simplifies AI Agent Parallelism
Docker is overkill for running AI agents in parallel. @Dagger_io’s new Container Use is a much simpler, agent-first alternative: → No Dockerfile needed → Git-native isolation → Full logs & real‑time visibility + easy one-liner commands to do it all! 🙌 Let’s compare 🧵↓ https://t.co/etVd1uHheZ
OpenAI Slashes Microsoft Revenue Share, Gains $50B
Major shift at OpenAI: revenue share with Microsoft & partners to fall from 20% → 8% by end of decade, potentially netting @OpenAI an extra $50B. Big move on margins, and implying pressure on terms and structure. 🔗 https://t.co/kjyCLj1C6B #AI...
OpenAI Needs $1 Trillion!?, Google Kills Ads, BlackRock's $40B Data Center
Jaeden Schafer’s post surveys a whirlwind of AI‑industry headlines, from Amazon’s lawsuit against Perplexity’s “Agent Shopping” feature and OpenAI’s $38 billion shift to AWS, to Nvidia’s historic $5 trillion market cap and OpenAI’s purchase of the Sky desktop‑control AI. The author intersperses...
Even LLMs Can't Answer Simple Emoji Queries
Some questions shouldn't be asked of LLMs. :-)
Top AI Minds Ranked: Who Defines Our Future
All the best in AI are on this list: https://t.co/2dDZtIT7Gf
Google Cloud C4 Brings a 70% TCO Improvement on GPT OSS with Intel and Hugging Face
Intel and Hugging Face benchmarked OpenAI’s GPT OSS on Google Cloud’s new C4 VMs (Intel Xeon 6/Granite Rapids) and report a 1.7x improvement in total cost of ownership versus prior-generation C3 instances. The C4 machines delivered 1.4x–1.7x better throughput per...
Add AI to Any Site with a Tiny Widget
[tiny project] ai-share-widget the easiest way to add* ai capability to a site with a front-end widget where you can customize the prompt and allow the user to pass that into their AI tool of choice github (102 lines of code) in...
Video AI Faces Its Own Turing Test
The Turing Test for video … 😅
A Holiday Season Gift-Wrapped In Agentic Commerce
OpenAI has published a Product Feed Specification to let merchants list and sell products directly through ChatGPT’s Instant Checkout, complete with product-level controls (enable_search, enable_checkout), support for rich media/3D assets, and 15-minute feed updates. Major partners including Walmart and Salesforce...
Walmart Explores ChatGPT-Powered Shopping Experience
Walmart to allow customers to shop using ChatGPT? https://t.co/GcOFDzhy0f #shopping #onlineshopping #wallmart #ChatGPT @PawlowskiMario @chidambara09 @Ym78200 @CurieuxExplorer @efipm @bigfundu @sayedflah @Ronald_vanLoon @cyngn @belindabeibi @odisseiaalfa @DigitalColmer @MyCompanionsAI @KirkDBorne @patricegorissen @jeffkagan @EdwardKens50830 @enilev @insom_ai333 @ChrisCCrowley @sallyeaves @gurmeet_judge @VairagyaSadhana @andresvilarino @amomsimpression @amit6060 @0xAmol @OfficialDabier
Mac Mini Excels at Inference; DGX Spark Needed for Fine‑tuning
Saw that DGX Spark vs Mac Mini M4 Pro benchmark plot making the rounds (looks like it came from @lmsysorg). Thought I’d share a few notes as someone who actually uses a Mac Mini M4 Pro and has been tempted...
How to Filter Out AI on Pinterest
Pinterest has added a desktop setting that lets users toggle off GenAI interests to reduce AI-generated content in their recommended Pins, accessible under Settings → Refine Your Recommendations → GenAI interests. The change is rolled out via the desktop app...
AI Art Hits Sameness, Lacks Emotional Storytelling
AI is amazing but it isn't good enough yet. @Kantrowitz says, below, AI has a sameness problem. The problem is if you hang out here on X all day long we see the sameness. Look at my AI Artist's feed. If...
From Moonshot to Millions: Veo’s Video Revolution
In 2018, video generation was a moonshot project by the Google Brain team. Today, hundreds of millions of videos have been generated with Veo. 🤯 In this episode of Release Notes, @doomie and @OfficialLoganK reflect on Veo's journey from research to...
Anthropic’s Latest AI Model, Claude Haiku 4.5, Doubles Down on Speed and Safety
Anthropic launched Claude Haiku 4.5, a lightweight AI model that it says matches Claude Sonnet 4’s coding performance while running twice as fast and at one-third the cost. The model is available across Anthropic’s apps, Claude Code, Amazon Bedrock and...
AI Contests Accelerate Product Fixes and Empower Creators
Since I just judged another, similar, contest, these are smart for companies to do. First, they are a forcing function to get AI companies to ship and fix things, but they get feedback from real users and get lots of...
AI Autonomously Resolves IT Outages, Easing Human Overload
AI is stepping in as an autonomous operator for overloaded humans. ➡️ AI agent detected VPN outage ➡️ Authenticated into New Relic ➡️ Re-established expired integration ➡️ Streamed telemetry data ➡️ Identified SW update as root cause ➡️ Suggested a rollback IT lead reviewed...
Google’s AI Videos Get a Big Upgrade with Veo 3.1
Google unveiled Veo 3.1, an upgraded text-to-video model that improves prompt adherence, audio realism and now supports both landscape and portrait (16:9) outputs, plus a lower-cost “Fast” variant. The model is rolling out across Google’s ecosystem—Gemini app, Flow filmmaking tool,...
Nano Banana Expands to Slides, NotebookLM, Gemini Tools
Several of our launches this week have been absolutely 🍌… Here are the new places you can find Nano Banana: — Edit images using custom prompts directly in Slides @googleworkspace — @NotebookLM’s video overviews got a huge upgrade with 6 new...
Dreamina AI Surpasses Nano‑Banana, Offers Multi‑Media Creation
I'm playing with @dreamina_ai (used it on my post earlier) and it's quite special for creating a wide variety of things, videos, photos, posters. It's from ByteDance and is beating the others, lets you upload six images and mix them...
Major Federation of Unions Calls for ‘Worker-Centered AI’ Future
The AFL-CIO launched a “workers first initiative on AI,” urging employers and policymakers to embed labor protections into AI deployment through stronger collective bargaining, regulation, and worker involvement in development. The federation — representing 63 unions and nearly 15 million...
India Launches EdGini: Personalized, Fun Learning Platform
Excited to share my new venture– EdGini, a Made-in-India learning platform built to empower students! • 🎯 Smart, engaging & personalized learning experience • 📊 Real-time performance tracking & skill growth • 🧩 Curated challenges to make learning fun • 💡...
Veo 3.1 Adds Audio, Object Tools; Flow Goes Mainstream
Veo 3.1 just dropped - Google listened and added audio everywhere it was missing. You can also insert objects into video and (soon) remove them too. Flow is quickly becoming less of a labs demo, and more of a creation tool....
The Right Mental Model For Agentic AI
A forthcoming report preview argues that companies need a clear mental model for 'agentic' AI—autonomous, goal-driven systems—to effectively reshape work and capture value. It outlines practical frameworks and design principles for integrating agents across enterprise functions, plus metrics for measuring...
AI Success Requires Ops Redesign, Not Just UI Upgrade
Overheard at World Forum today ⬇️ #ServiceNowPartner “We're reimagining critical processes such as hire to retire and source to pay and record to report, so that everyone and everything, people, systems, suppliers, agents, can work together in one place.” If your team's...
Codex CLI Exit Saves Sessions, Huge QOL Boost
whoever added this to the codex cli when you exit is a lifesaver cannot tell you how many times I've mindlessly ctrl+c trying to copy paste only to break a session then i gotta dig up the latest chat. such a simple...
OpenAI Targets $1 Trillion From $13 Billion in Five Years
"OpenAI has five years to turn $13 billion into $1 trillion." Do you think they can do it? https://t.co/py7EgsSvnP
NanoChat Code Sparks Fresh Look at LLM Theory
Karpathy’s nanochat codebase motivated me to revise llm theory
Anthropic Releases Comprehensive Claude Code Blueprint for Developers
Anthropic just published a killer blueprint for Claude Code users! 🔥 → GitHub CLI workflows → .md prompt tuning → Custom slash commands → Headless + multi-agent agents Dense, yes... but pure gold if you're building with LLMs. Download link in 🧵↓ https://t.co/kCbF9amMFW
Launch of Readsail: Curated AI Learning Platform
If your company wants to help employees stay up to date with AI, I've teamed up with @natolambert and others to launch @readsail. It's essentially a platform that makes it easy to manage ongoing AI learning and brings you a...
AI Will Redefine Managers and Jobs, Says ServiceNow SVP
I spoke with Josh Kahn, SVP at @ServiceNow, about future AI interfaces, the role of mid-managers, and AI-driven job transformation. Here’s everything he said 🧵 #ServiceNowPartner
FactoryAI CLI Setup Guide Arriving Soon
getting setup with @FactoryAI's CLI i hear you - more how-to's/guides are coming https://t.co/gRvIl7hmzY
Securing AGI Future with 40k GPUs and 2GW Texas Campus
onwards to agi 🫡 looks like we’ve got stuff to chat about at the @stateofaireport nyc launch tmrw @eisokant :) https://t.co/lUthvM1nhD
Humanlike Chatbots Trigger Mental Health Crises
Earlier this year, after receiving emails from people in the throes of AI psychosis, I began to keep in touch with one man as he journeyed to recovery. I'm grateful to James that he was willing to go on record. He...
Found a Goldmine for Scalable Search and RAG
I've been looking for how search and RAG can be done on large scale and actual data, and there's just toy examples everywhere I look. Not just some pdfs or a website with everything in context, but actual search, retrieval,...