EP 683 : Thinking Machines Lab Inkling 2, Xiaomi Humanoids, MIT Memory Fix & Waymo Robotaxi Update

Imagine a security camera that can take off, zip around your hallway, and investigate a break‑in on its own. Amazon’s Ring just made that a reality with the $270 Ring Drone Cam, a new indoor drone that patrols your home when alarms trigger. While Ring pushes the boundaries of home surveillance, Meta is stepping into the physical world with its first humanoid robot, Max, built in just three months with Chinese partner Fourier. Max can pick up objects and will soon get upgraded hands, sensors, and stair‑climbing skills, giving Meta a tangible platform for training AI on household tasks.Across the Atlantic, Italy slapped Meta’s Facebook with an €805 million fine for accessing Android users’ gallery photos without clear consent—a stark reminder that data‑privacy penalties are soaring and that blanket permission models are no longer safe. The ruling forces Meta to disable the feature by default and seek explicit user approval, highlighting the growing regulatory pressure on big tech.In the stock market, Agility, the maker of the Digit humanoid, went public on Nasdaq, offering investors a pure‑play exposure to humanoid robotics despite modest revenues. The listing underscores the hype around robots that can lift 50 lb, recharge in nine minutes, and run 20‑hour shifts.To keep you ahead, we’ve rounded up the hottest AI productivity tools: Templify AI for effortless content editing, Claude Meetings that adds Claude to your calls, Trainn MCP for auto‑generated product videos, Gemini 4 Hydrogen for agentic coding, qorqbot for proactive business tasks, and Aven App Agent that guides users through apps.Listen now, share with fellow tech enthusiasts, and subscribe for weekly AI & robotics deep dives. Tools mentioned: Templify AI, Claude Meetings, Trainn MCP, Gemini 4 Hydrogen, qorqbot, Aven App Agent.

Breaking news shakes the AI world: Nvidia’s legendary founder Jensen Huang has stepped down as CEO after 33 years, sparking a power shift at the company that fuels almost every modern AI model. In the same episode, Google rolls out its first Gemini‑powered AI smart glasses, promising privacy‑first features and a new way to interact with the world through voice and vision. We break down why Huang’s exit matters for chip strategy, China ties, and upcoming earnings, and how CFO Colette Kress will steer Nvidia in the interim. Then we dive into Google’s $395 AI glasses, built with Samsung and framed by Warby Parker, that can "Hey Gemini" your surroundings, translate 70+ languages, and hide a recording light for bystander privacy. Plus, we cover Anthropic’s Claude Meetings joining video calls, DeepSeek’s alleged use of Nvidia Blackwell chips, Meta’s new Muse phone‑calling, OpenAI’s $2 M child‑safety pledge, and the latest AI productivity tools. Want to stay ahead of the AI curve? Listen now, share with fellow tech fans, and subscribe for daily deep dives. Tools mentioned: Marqly, Foxpit, Toki, Lemon, Gemini 4 Hydrogen, Claude Meetings, Genie.

Breaking news rattles the AI world: Nvidia’s board is reportedly moving to oust co‑founder Jensen Huang after a second alleged affair, a drama that could upend the company that powers almost every modern AI model. Meanwhile, Google DeepMind unveiled AlphaGenome 2, an AI that maps tiny DNA variations to specific cell types, promising faster disease‑target discovery. In the same whirlwind, OpenAI announced its first $10 billion‑quarter, Anthropic’s IPO is oversubscribed, and SpaceX’s Grok Bot 1.0 graduated from beta to a full‑time teammate across Slack, Teams and video calls. We also dive into Fiddler’s new Enterprise Playbook for governing Claude Code agents, a sneak peek at the Blue AI phone, and a health‑tracking ring from Natura. United Airlines is tightening AI agent access, and a Maya‑site discovery using NASA laser maps highlights AI’s expanding reach beyond pure tech.What does this mean for developers, investors, and anyone watching the AI boom? We break down the potential fallout for Nvidia’s chip ecosystem, why AlphaGenome 2 could accelerate drug pipelines, and how Grok Bot’s proactive features might reshape remote teamwork. Plus, we highlight today’s hottest AI tools that can boost your productivity right now.Listen now to get the full analysis, share with your network, and subscribe for daily AI deep dives. Tools mentioned: Scribe V3, Gemini Decisions API, Grok Bot 1.0, ScroogeFrog AI, Nod, Kippy AI, IT Code.

Imagine your brand showing up not on a search engine page but right inside the chat where millions are already asking questions. OpenAI just turned that vision into reality with the new 'Businesses' beta in ChatGPT – a free‑to‑claim verified profile, real‑time mention analytics, and a customizable AI agent that can answer shopper queries or serve as a paid‑ad landing spot. Early access is limited to U.S. online retailers and digital services, but the dashboard already promises insight into discovery topics and click‑through rates that were previously a black box.While OpenAI is rewriting the rules of discovery, Anthropic is turning Claude into an office‑wide coworker. The beta rollout of Claude Sheets and Claude Whiteboards lets teams generate spreadsheets with formulas or collaborative boards by simply prompting Claude, and Claude Data now produces narrated video dashboards from raw metrics. Claude Meetings even sits in on video calls, and new connectors let Claude edit Adobe, Atlassian and other apps without leaving the chat.The AI‑future clock is also ticking faster. The AI Futures Project moved its AI 2027 timeline up, citing METR’s analysis that Anthropic’s models are improving twice as quickly as expected – a shift that could bring super‑human research capabilities within months.Other headlines: Adobe previewed MCP connectors for Photoshop and Premiere and the Firefly Graph workflow builder; Google unveiled Gemini 4 Carbon for Ultra subscribers; OpenAI acquired the travel‑planning startup Odessia; and France’s Paris Call on AI safety gathered signatories including OAI, Anthropic and Mistral.Ready to see how your brand can claim a spot in the next‑gen assistant? Listen now, share with your marketing team, and hit subscribe for weekly AI breakthroughs. Tools mentioned: Businesses in ChatGPT, Claude Sheets, Claude Whiteboards, Claude Data, Gemini 4 Carbon, MiniMax H3‑Image, Adobe Firefly, Alsona, Maxma, TK by Medium, Marketing Standup.

What if the AI frontier is moving faster than anyone expected? This week Anthropic’s fresh Claude Fable 5.2 shatters expectations—20 % cheaper, up to 30 % quicker, and now #1 on Artificial Analysis’s Intelligence Index, edging out Google’s unreleased Gemini 4 Carbon by five points. Meanwhile, Meta’s new Muse Ember proves that speed doesn’t have to cost a fortune, matching Anthropic’s Claude Haiku 5.5 performance while delivering tasks for roughly a third of the price. Google counters with two specialized TPU chips, the 8tx and 8i Max, designed to keep data encrypted and double processing power for massive‑scale agents. The new TPUs aim to challenge Nvidia’s dominance by offering on‑prem encryption and a unified memory architecture that can handle whole codebases or long video streams in a single pass. We also cover a flurry of side stories: Nvidia’s legal drama in Singapore, Anthropic’s looming IPO, Adobe’s Disney‑powered custom models, Instagram’s free Effect Builder, and a recent hack at Fireworks that exposed billing data. These moves signal a shift toward tighter hardware‑software integration and more aggressive pricing strategies across the board. Each development reshapes the competitive map and hints at where the next breakthroughs will land. If you’re building with AI or just love staying ahead of the curve, hit play, share the episode with your team, and subscribe for weekly deep dives. And don’t miss our quick roundup of the hottest tools you can start using today. Tools mentioned: Claude Fable 5.2, Seedream 5.5, Muse Ember, Inkling 2 Small, Higgsfield Miracle, Floot, Spira, The Code.

An explosive week in AI: Nvidia’s founder‑CEO Jensen Huang publicly admits an affair amid a board‑ordered spy‑investigation, sending shockwaves through the sector, while Anthropic reveals how its Claude model “dreams” to clean and prioritize its memory—boosting performance in games like Pokémon. Apple is rumored to be fast‑tracking a Siri‑as‑agent overhaul to take on Meta’s Muse, OpenAI’s new funding round, and a breakthrough from a Physical Superintelligence lab using Anthropic tech to solve a decade‑old quantum conjecture. We also break down Higgsfield’s layer‑by‑layer video editor, the latest AI‑for‑physics milestones, and why AI stocks are jittery after the midterms. Stay ahead with our curated AI‑tool roundup, from Levanto Sage2’s live‑conversation data extraction to Memorias’ automatic biography builder, outages.live’s real‑time incident feed, Cybersecurity Code’s weekly briefs, Meta’s Muse avatars, Thinking Machines Lab’s Inkling 2 Small model, and OpenAI’s ChatGPT Images 2.6 that now creates GIFs. Listen now for deep analysis, share with fellow tech enthusiasts, and hit subscribe to never miss a pulse on the AI world. Tools mentioned: Levanto Sage2, Memorias, outages.live, Cybersecurity Code, Muse, Inkling 2 Small, ChatGPT Images 2.6.

Ever wondered how AI could buy your groceries with a single text, or rewrite a speaker’s lips in real time? In this episode we unpack Amazon’s new Alexa‑for‑Shopping pilot that lets invited US users shop, compare deals and checkout entirely via SMS, and a bold Microsoft patent that could splice words into live video while synchronizing mouth movements. We discuss why these moves could reshape trust and control in digital commerce. On the science front, NASA’s Voyager 1 is set to cross the one‑light‑day threshold, meaning a round‑trip radio ping will take 48 hours—a stark reminder of the vastness of space after a recent communications scare. We also explore Perseverance’s discovery of ultra‑fine Martian dust that may threaten astronaut health, CERN’s breakthrough observation of a beauty‑quark particle turning into two muons, and a stunning report from Curtin University of mineralized blood cells preserved in 115‑million‑year‑old dinosaur bones. Our AI & Science roundup also covers the tiniest Martian dust particles that could infiltrate lungs, the ancient Teghaza 001 meteorite unlocking billions‑year‑old water signatures, and the implications of the CERN beauty‑quark finding for future physics experiments. Stay tuned for quick takes on the latest AI tools that can boost your workflow, and don’t forget to listen, share, and subscribe for more tech‑savvy insights. Tools mentioned: beacon.sh, Wanaka Studio, Scalekit, Tobira Site Agent, Teamily AI, UnleashX.

Robots are stealing the show again—some clocked an 8‑hour factory shift, others learned to tidy supermarket aisles, and a few ended up on a stretcher. In this episode we dive into the biggest breakthroughs and mishaps from the past week.First up, Thinking Machines Lab unveiled Inkling 2 Small, a 17‑billion‑parameter active mixture‑of‑experts model that rivals larger systems on vision, coding and enterprise tasks. The team promises an Apache 2.0 open‑weight release pending safety reviews, a move that could hand true model control back to developers.Next, Xiaomi’s humanoid fleet proved they can run a full day on the assembly line of its EV plant, autonomously sorting parts, self‑correcting errors and keeping pace with human workers. Meanwhile, MIT researchers gave robots a self‑repairing memory system that flags uncertain object descriptions and rewrites faulty visual records, fixing 92 % of ambiguities in tests.On the road, Waymo announced it will end its exclusive ride‑hailing partnership with Uber, allowing riders in Austin and Atlanta to book directly through the Waymo app starting January 2028, aiming for a smoother, brand‑consistent experience.We also round up the hottest AI productivity tools: Tobira Site Agent turns your website into a selling AI, Turnstone offers a free, self‑hosted Grok‑style bot, Rankfender v3.0 boosts SEO visibility for humans and AI agents, crystl provides a multitasking terminal for AI coding agents on macOS, TraceX acts as an AI engineer for production issues, and Datastripes Data Stories converts spreadsheets into instant data narratives.Listen now to get the full details, share the episode with fellow tech fans, and subscribe so you never miss a breakthrough. All of these stories illustrate how quickly AI is moving from labs to real‑world impact. Tools mentioned: Tobira Site Agent, Turnstone, Rankfender v3.0, crystl, TraceX, Datastripes Data Stories.

Ever wonder how quickly the frontier of AI is actually moving behind closed doors? In this episode we unpack METR’s exclusive 11‑week deep dive into Anthropic’s internal data, which shows the company’s AI development may be accelerating at 1.3‑3× its previous pace—largely powered by its own models. We discuss what the redacted “Model 4” and the looming danger threshold in Anthropic’s IPO filing could mean for safety and competition, and why even the lab’s insiders were “seriously wrong” about the speed of progress.Next up, OpenAI rolls out Images 2.6, the first image model that can generate animated GIFs directly in ChatGPT and via API. We break down the two flavors—Sunburst for premium quality and Flare for lightweight jobs—plus the 50 % speed boost that puts the new model ahead of its predecessor on Design Arena.We also cover the rest of today’s AI headlines: OpenAI’s GPT‑6 Cyber hunting software bugs, Perplexity’s new “Teammates” agents, Anthropic’s personal‑finance hub inside Claude, a smart‑ring from startup Aina, Google’s cost‑effective Gemini Omni 1.2 Flash video model, and more.Ready to stay ahead of the curve? Hit play, share with your network, and subscribe for weekly deep‑dives into the AI breakthroughs that matter. At the end of the show we list the top tools mentioned so you can start experimenting right away.Tools mentioned: Images 2.6 (OpenAI), Gemini Omni 1.2 Flash (Google), Atmee v2 (Atmanity), Wanaka (3D AI apps), Glasshand (device‑level AI APIs), Pancake (go‑to‑market agents).

Imagine never waiting on hold again—your AI assistant can pick up the phone for you. In this episode we break down SpaceXSI’s newest breakthrough: Grok Bot can now place calls on U.S. numbers, book appointments, negotiate bills and even order pizza, all while showing you a live transcript. Users remain the “caller of record” and must obey phone‑law disclosures, a detail Elon Musk flagged as “unsustainably high” in cost. We explore why this matters for the future of AI agents and how Grok Voice already handles 20,000 Starlink calls a day.Next, we turn to the U.S. midterm results that sent a clear message: voters want tighter limits on data centers and are skeptical of AI, with 57 % saying the technology will do more harm than good. The outcome could reshape federal oversight of AI labs, especially as Democrats prepare hearings on regulation.We also cover the week’s hottest AI headlines: Pangram’s Image 1.0 detection model and Chrome extension; Ricursive’s Ripple‑65B preview, an open‑source coding champion; OpenAI’s acquisition of Pierre and the rollout of ChatGPT Images 2.6 that now creates GIF‑style motion; Conductor’s merger with Meta to manage multiple agents; Atmanity’s Atmee v2 avatar model that turns a single photo and 10‑second voice clip into a lifelike video companion; and corporate policy shifts as Apple trims AI usage limits for engineers while Goldman Sachs, Meta and Microsoft restrict AI tools for security reasons.Ready to stay ahead of the AI curve? Listen now, share this episode with your network, and hit subscribe for a weekly dose of AI strategy and product news.Tools mentioned: Grok Bot, Hy4, Kimi K3.1 Pro, Oboe, Midjourney, Gemini Notebook, Lovable.