Good morning, AI enthusiasts. AI keeps tanking in public opinion polls, but Anthropic’s new data shows the answer is a lot more nuanced than a simple "thumbs down."

Using Claude itself as the interviewer, Anthropic collected 80K+ conversations across 70 languages in one week. The finding that stood out: most people aren't choosing sides between hope and fear… They're carrying both.

In today’s AI rundown:

• Anthropic surveys 81k people on AI hopes, fears

• Cursor’s coding model cuts costs near the frontier

• Use Replit’s Tasks feature to improve your site

• Microsoft AI’s image model climbs leaderboards

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

AI RESEARCH

📊 Anthropic surveys 81k people on AI hopes, fears

Image source: Anthropic

The Rundown: Anthropic just released what it says is the biggest qualitative AI attitudes study ever, using Claude to interview 81k of its users across 159 countries about where they think the tech is headed and what scares them about getting there.

The details:

• Anthropic introduced Claude Interviewer in December, building a special version of Claude that ran open-ended conversations in 70 languages.

• Professional excellence was the top-reported hope, with freeing up time, financial independence, and broader life management frequently mentioned.

• Fear of AI getting things wrong outranked every other concern, with job anxiety, losing personal agency, and over-reliance close behind.

• AI sentiment varied by region: India and South America skewed above average, while the U.S., Europe, Japan, and South Korea ran neutral or below.

Why it matters: AI's favorability numbers have cratered in mainstream polls, but Anthropic's study adds nuance that those surveys miss. Almost as notable is Claude running 80K in-depth interviews across 70 languages in a single week, a wildly strong proof of concept for the tech as a research tool that simply didn’t exist a year ago.

TOGETHER WITH NORTON NEO

🌐 The browser built for the AI agent era

The Rundown: Norton Neo is the first AI-native browser designed to be powerful, productive, and private by default — combining intelligent assistance with built-in protection so you don't have to trade privacy for productivity.

With Norton Neo, you can:

• Search across tabs, history, and the web from one universal bar

• Research and write side-by-side with built-in Split View multitasking

• Security is embedded at the core, with a built-in VPN powered by Norton

Try Norton Neo and experience the future of browsing.

CURSOR

⚙️ Cursor’s coding model cuts costs near the frontier

Image source: Cursor

The Rundown: Anysphere, the company behind AI code editor Cursor, just shipped Composer 2, a third-generation in-house model that is competitive with frontier coding models from OpenAI and Anthropic at a fraction of the cost per task.

The details:

• Composer 2 topped Opus 4.6 on the independent Terminal-Bench 2.0 (61.7% vs 58%) and sits within 5 points of GPT-5.4 on Cursor's own CursorBench.

• At $7.50/M output tokens on its fast tier, Composer 2 costs roughly 1/10th of GPT-5.4 and 1/20th of Opus 4.6 at comparable speeds.

• Composer’s scores on the company’s internal CursorBench have climbed from 38% to 61.3% across three model generations shipped since October.

Why it matters: Cursor quickly went from harnessing other top AI models to building one of its own at this price point. Nearing the frontier as an application-layer company is an impressive feat, and the speed, cost, and performance of Composer 2 could change the math for developers paying full price for coding with GPT-5.4 or Opus 4.6.

AI TRAINING

⚙️ Use Replit’s Tasks feature to improve your site

The Rundown: In this guide, you will learn how to use Replit's Tasks feature to improve an existing site without messing up your already working app, making it easier to keep isolated fixes from piling on top of each other and make real progress on your app.

Step-by-step:

1. Go to Replit, drop in your app idea, and hit Plan. If you already have an app, open that instead. You will need at least a Replit core plan to try this feature

2. Once in your project, click the plus button on the left, create a task for a fix like mobile optimization, and prompt: “Make dashboard and all components mobile responsive. Use different components for mobile if not possible”

3. While it's planning, queue a second task in parallel, try something like improving the landing page design, cleaning up the nav bar, or a bug fix

4. Once task planning is done, start them. You can preview each task, and when they finish, click Apply Changes to Main Version to update the production app

Pro tip: Toggle your main agent into plan mode and tell it the improvements you want. It will generate a PRD. Then, divvy up that PRD into new parallel tasks.

PRESENTED BY OPTIMIZELY

🧪 Scale experimentation without scaling headcount

The Rundown: Most experimentation teams have hit a wall — headcount is maxed and testing capacity has stalled. Optimizely's upcoming webinar shows how leading teams are using AI agents across the full experimentation workflow to scale programs without adding to the team.

The April 1st session will cover:

• How AI agents add value across every stage of the experimentation lifecycle

• Real use cases from the leader behind Farfetch's award-winning program

• Practical strategies to scale testing across your organization without growing your team

Register for April 1st at 10 am ET. Can't join live? Register anyway to receive the recording.

MICROSOFT

🎨 Microsoft AI’s image model climbs leaderboards

Image source: Microsoft

The Rundown: Microsoft's AI Superintelligence team just released MAI-Image-2, a text-to-image model that landed at No. 5 on the Arena AI leaderboard — marking the strongest release yet for Mustafa Suleyman’s lab.

The details:

• Arena.ai ranked MAI-Image-2 at No. 5 overall, trailing just Gemini (several variants) and GPT Image-1.5 with strong upgrades in photorealism, 3D, and art.

• The biggest jump from its predecessor came in text rendering, up 115 points, with drastically improved performance on posters, slides, and infographics.

• MAI-Image-2 is free to try in Microsoft's MAI Playground for U.S. users, with Copilot, Bing, and API access on its Foundry platform rolling out soon.

• The release comes amid Microsoft’s AI leadership shuffle, with Suleyman shifting away from Copilot to focus solely on frontier model work.

Why it matters: Microsoft has been signaling its desire to reduce its reliance on OpenAI and truly compete with its own models, and MAI-Image-2 is the strongest step yet in that direction. But the legacy tech giant still has a major uphill battle to gain market share from the already well-entrenched frontier options at the top.

QUICK HITS

🛠️ Trending AI Tools

•  🗣️ Unwrap Customer Intelligence - Turn unstructured customer feedback into data-backed insights that inform your product roadmap*

• ⚙️ Composer 2 - Cursor’s powerful, cost-effective new in-house coding model

• 🎨 MAI-Image-2 - Microsoft's new upgraded text-to-image model

• 🚀 Google AI Studio - Google’s upgraded full-stack vibe coding agent

*Sponsored Listing

📰 Everything else in AI today

Google rolled out upgrades that turn its AI Studio into a one-stop vibe-coding app builder, pairing a new Antigravity coding agent with built-in backends and user login.

Jeff Bezos is reportedly raising a $100B fund to buy chip, defense, and aerospace manufacturers, with plans to use them for his secretive AI startup, Project Prometheus.

Perplexity introduced Health, a new feature allowing users to securely connect health apps, wearables, and data to its Computer agentic system.

DoorDash launched a new ‘Tasks’ app, paying its couriers to capture video and data from everyday tasks and conversations for AI and robotics training.

OpenAI announced the acquisition of open-source developer tool startup Astral, folding the company’s staff into its Codex team.

Meta launched an AI support assistant across FB and IG for 24/7 support, also previewing advanced content enforcement systems that catch 5K daily scam attempts.

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Gerro A. in The Philippines:

“I'm an assistant secretary in my class in high school. The president told me to create a hub where my classmates can share reviews and leave comments. He expected a simple Google Drive or Discord to share reviews. However, I spent one week creating a website from scratch, using ChatGPT as my helper along the process. Take note, I had no idea how to use HTML, JS, or CSS (I only know how to make games in Python).

This led me to explore and learn more about Full-Stack Development and provide my classmates with a productivity tool.”

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Google bets on ‘Vibe Design’ with Stitch

• Read our last Tech newsletter: China greenlights commercial brain implant

• Read our last Robotics newsletter: Nvidia’s move to own the robot future

• Today’s AI tool guide: Use Replit’s Tasks feature to improve your site

• RSVP to next workshop on March 26 @ 2PM EST: Intro to Vibe Coding pt. 3

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, robotics enthusiasts. Nvidia just made its clearest move yet to become the computing backbone of the physical world.

At its annual GTC conference, the $3T chip giant pitched a full-stack future for humanoids, factory bots, and robotaxis, bundling robot brains, simulation tools, and autonomous-driving systems into one sprawling ecosystem. But can startups and rivals still carve out space in a market Nvidia seems determined to lock up?

In today’s robotics rundown:

• Nvidia wants to be the OS of every robot

• Gecko Robotics lands $71M U.S. Navy deal

• Ex-Meta engineers give robots visual memory

• Robot dogs guard billion-dollar AI data centers

• Quick hits on other robotics news

LATEST DEVELOPMENTS

NVIDIA

🤖 Nvidia wants to be the OS of every robot

Image source: Reve AI / The Rundown

The Rundown: Nvidia’s Jensen Huang used the annual GTC conference to pitch the tech giant as the default computing backbone for physical AI, debuting new humanoid software stacks, next-gen hardware platforms, and new robot and robotaxi partners.

The details:

• Nvidia unveiled a plug-and-play “brain” for humanoids, putting GR00T N1.7 into early access and teasing the next-gen N2.

• It’s pairing these models with new Isaac and Cosmos tools, designed to train robots faster in simulation before deploying them in the real world.

• Nvidia is also partnering with chipmakers to ensure that next-gen robot motors, sensors, and safety systems are designed to integrate with its ecosystem.

• Uber will tap Nvidia’s DRIVE to power Level 4 robotaxis starting in LA and San Francisco in 2027, with plans to scale to 28 cities by 2028.

Why it matters: Nvidia wants to sit at the center of a multi‑trillion‑dollar robotics market, getting manufacturers and platforms to standardize on its system. The strategy’s success depends on robots finally graduating from pilots to production, and on how much resistance Nvidia draws from open-source alternatives and rival silicon.

GECKO ROBOTICS

⚓️ Gecko Robotics lands $71M U.S. Navy deal

Image source: Gecko Robotics

The Rundown: Pittsburgh startup Gecko Robotics scored a five‑year, up‑to‑$71M U.S. Navy deal to deploy its wall‑climbing AI inspection robots across warships, creating high-res virtual models to help cut maintenance backlogs.

The details:

• The deal starts as a five-year agreement with an initial $54M award and a ceiling of $71M, making it the Navy’s largest robotics contract to date.

• Gecko’s climbing, flying, and swimming robots will scan critical structures, feeding sensor data into AI models that can flag corrosion and structural issues.

• Gecko says its system can identify repairs up to 50x faster than manual inspection, shrinking months-long work down to a couple of days.

• The move dovetails with Trump’s drive to rebuild U.S. shipbuilding and close the gap with China by getting more combat-ready hulls back to sea faster.

Why it matters: It shows the Pentagon is ready to buy commercial robotics at scale to fix one of its least glamorous problems: maintenance. If Gecko can turn ships into always-updated digital twins, it could become the playbook for software-first startups looking to modernize everything from depots to airfields and keep the work coming.

MEMORIES AI

📽️ Ex-Meta engineers give robots visual memory

Image source: Memories.ai

The Rundown: A startup founded by former Meta engineers is building what it calls a “visual memory layer” — infrastructure that lets wearables and robots store, index, and search first-person video footage so they can remember what they’ve seen.

The details:

• Memories.ai is partnering with Nvidia, tapping Cosmos-Reason 2 for vision-language reasoning and the Metropolis stack for large-scale video search.

• Founded in 2024 by ex-Meta Ray-Ban glasses engineers Shawn Shen and Ben Zhou, Memories spun out of their work on video-capturing smart glasses.

• The company has raised $16M to date, including an $8M seed round in 2025 and an $8M extension led by Susa Ventures.

• Shen told TechCrunch the company is already working with major wearable makers and sees bigger opportunities in robotics.

Why it matters: AI will need to remember what it sees to really work in the physical world, Memories.ai argues. Its visual memory layer is meant to plug a hole left by largely text-focused memory tools from OpenAI, xAI, and Google, giving robots a way to store and recall what they actually “see” while working in the real world.

ROBOT DOGS

🐕 Robot dogs guard billion-dollar AI data centers

Image source: Boston Dynamics

The Rundown: AI data centers are starting to outsource guard duty to quadruped robot dogs from Boston Dynamics and Ghost Robotics, autonomously patrolling sprawling campuses for threats that fixed sensors routinely miss.

The details:

• Spot units run $175K–$300K, but operators say the ROI math is straightforward: recoup the cost in roughly two years by cutting guard labor.

• Robot dogs patrol perimeters, scan fence lines for intruders, and flag hazards like leaks, heat anomalies, or propped‑open doors that fixed sensors miss.

• Ghost Robotics’ Vision 60s are already active at a handful of facilities, navigating rough terrain while streaming 360-degree video to control rooms.

• The pitch from operators is augmentation, not replacement — human security teams monitor the feeds as robots take care of 24/7 patrol.

Why it matters: Tech companies are pouring $700B into hundreds of new AI facilities, with Meta’s Hyperion alone slated to sprawl to 4x the size of Central Park. Even as operators trim onsite guard headcount, execs insist the robots are there to augment humans, with centralized security teams monitoring their feeds from control rooms.

QUICK HITS

📰 Everything else in robotics today

Unitree founder Wang Xingxing said Chinese humanoids could run a sub‑10‑second 100m dash by mid‑2026, potentially beating Usain Bolt’s 9.58‑second world record.

Time said the U.S. is rushing into a “humanoid soldier” arms race, led by Foundation’s Phantom MK‑1 robots in Ukraine, on the promise of keeping human troops alive.

Renault started using Wandercraft’s Calvin-40 humanoids to haul car tires at its Douai plant with plans to roll out about 350 more units over the next 18 months.

South Korean firm Tesollo unveiled a lightweight robotic hand with what it says is near-human-level dexterity that can grasp delicate objects and use everyday tools.

UK defense startup Cambridge Aerospace, which builds drone interceptor systems, is reportedly in talks to raise about $200M at a valuation above $1B.

Samsung is fast‑tracking humanoids and agentic AI for its own plants, aiming to turn all global factories into AI‑autonomous “smart factories” by 2030.

Researchers developed a fully biodegradable soft robotic finger that can perform over 1M actuation cycles accurately before decomposing into nontoxic compost.

Pokémon Go players have, often unknowingly, reportedly helped Niantic Spatial and Coco Robotics train their delivery bots using more than 30B player-shot images.

Drone startup Seneca will test five AI-guided firefighting drones in Aspen this summer to see if autonomous aircraft can attack wildfires faster than traditional crews.

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Google bets on 'vibe design' with Stitch

• Read our last Tech newsletter: China greenlights commercial brain implant

• Read our last Robotics newsletter: Travis Kalanick’s stealth robot play

• Today’s AI tool guide: Generate an actionable SEO audit with AI

• RSVP to next workshop @ 2 PM EST today: Intro to Vibe Coding pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team

Good morning, AI enthusiasts. Last year, "Vibe coding" changed how the world built. Now, Google is hoping it can do the same for design.

The company’s new Stitch overhaul adds voice editing capabilities, new agentic capabilities, instant prototyping, and more, hoping “vibe design" can do for UI what vibe coding did for development — collapse weeks of work into a single conversation.

Reminder: Our next live workshop is today at 2 PM EST — join part 2 of our Intro to Vibe Coding, where you’ll learn the vocabulary, mental models, and prompting habits that keep your development clean and consistent. RSVP here.

In today’s AI rundown:

• Google brings 'vibe design' to its AI UI canvas

• MiniMax's new M2.7 helped build itself

• Generate an actionable SEO audit with AI

• Microsoft ‘weighing’ legal action over Amazon-OAI deal

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

GOOGLE

🎨 Google brings 'vibe design' to its AI UI canvas

Image source: Google

The Rundown: Google just overhauled Stitch, turning its AI UI design tool into a voice-enabled infinite canvas that takes users from a rough idea to a clickable prototype — and coining the term "vibe design" for its new development workflow.

The details:

• Stitch now runs on an infinite canvas where users feed in images, code, or briefs, plus an agent manager that juggles multiple design directions at once.

• A new voice feature in preview turns the tool into a hands-free design partner, able to take direction and make live edits mid-chat.

• Instant prototyping can turn static screens into interactive prototypes in seconds, while auto-generating logical next screens for the UI flow.

• A new DESIGN.md format lets teams port design rules between Stitch and coding tools, with each project getting a style system out of the box.

Why it matters: Design has already changed completely in the AI era, but agentic capabilities are taking things to new levels. Stitch's upgrades let users move at AI-native speeds with easy integration into existing workflows, and a “vibe design” ethos that puts strong creation in reach in the same way “vibe coding” did for development.

TOGETHER WITH VISA

🤝 Get to Know Visa’s Trusted Agent Protocol

The Rundown: The trust layer for agentic commerce. Visa’s Trusted Agent Protocol allows AI agents to verify who they are, who they represent, and what they’re authorized to do — so merchants can interact with agents with confidence, and humans stay in control.

Visa Trusted Agent Protocol Helps Deliver: 

• Cryptographic verification of AI agent identity and intent

• Clear authorization and accountability for agent‑initiated actions

• Trust signals to help merchants distinguish legitimate agents from bad actors

Learn more here.

MINIMAX

♻️ MiniMax's new M2.7 helped build itself

Image source: MiniMax

The Rundown: MiniMax launched M2.7, what the company calls its “first model which deeply participated in its own evolution" — writing its own training code, running autonomous improvement loops, and matching the scores of top Western models.

The details:

• Early M2.7 versions were put to work on their own training, including writing improvement routines and tuning how the model learns from feedback.

• M2.7 ran 100+ cycles of autonomously analyzing its mistakes, rewriting code, and testing fixes — showing a 30% accuracy boost on internal benchmarks.

• On coding, M2.7 hit 56.2% on SWE-Pro and 55.6% on VIBE-Pro, putting it near Opus 4.6 and GPT-5.3-Codex for agentic engineering work.

Why it matters: Self-evolving AI is one of the bigger next steps forward, and while we’ve heard rumblings from OAI, Anthropic, Google, and xAI, MiniMax is one of the first labs out of China to openly make the claim. Future models will all likely be training and improving themselves, but for now, we’re just watching the feature emerge.

AI TRAINING

📝 Generate an actionable SEO audit with AI

The Rundown: In this guide, you will learn how to run a full SEO audit on any website using AI and turn the results into a slide deck and a prioritized task list. We tested this method in Manus, Claude, ChatGPT, Gemini, and NotebookLM.

Step-by-step:

1. Open any AI tool and prompt: “Perform a comprehensive SEO audit of [website URL]. Analyze SEO, on-page optimization, content quality, and backlink profile. Give me the top 10 highest-impact fixes ranked by effort vs. results. Export as a PDF report with sections for each category. Cite your data sources.”

2. Review the PDF, then prompt: “Turn this audit into a slide deck I can present to a client or stakeholder. Keep it under 10 slides. Include an executive summary, top issues by category, and a recommended action plan with priorities.”

3. Finally, fix the issues by prompting: “Create an SEO workbook as a CSV with every issue from the audit. Columns: Issue, Category, Priority (High / Medium / Low), How to Fix, Affected Page URL. Sort by priority.”

Pro tip: Give tools browser access so they can gather SEO data from free online tools. Manus did the best job on this without any configuration.

PRESENTED BY GLEAN

🔒 AI moves fast, and security hasn't caught up

The Rundown: AI agents now make decisions, take actions, and touch sensitive data across organizations — but most security policies weren't built for that. The AWARE Framework, developed by the Work AI Institute with Palo Alto Networks and Databricks, gives security and IT leaders a practical guide for governing AI agents at scale.

Download the AWARE Framework to learn:

• Why AI agents create risks most security tools weren't designed to catch

• Where organizations have blind spots and how to close them before they become a problem

• A practical governance model that any team can act on without slowing AI adoption

Get the framework here.

MICROSOFT & OPENAI

⚖️ Microsoft ‘weighing’ legal action over Amazon-OAI deal

Image source: Lovart / The Rundown

The Rundown: Microsoft is reportedly considering legal action against Amazon and OpenAI, with sources telling the FT it will sue if a new $50B cloud deal between the two companies violates its exclusive contract to host the startup's models on Azure.

The details:

• The dispute is over Frontier, OpenAI's new enterprise agent platform — also the anchor of a broader deal committing $138B in cloud spending to AWS.

• Microsoft dropped its exclusive hosting lock on OAI in October, but kept a clause that forces all developer access to OAI models to run through Azure.

• The FT’s source said “We know our contract… We will sue them if they breach it”, with another source adding “the last thing OAI needs is another court case”.

• OAI reportedly signed a new deal with AWS last week, which opened the door for the company’s deployment with the Pentagon.

Why it matters: Just when you thought the Microsoft-OAI relationship couldn't get more awkward, now there's a potential lawsuit to worry about. The FT source is right that OAI can't afford another legal fight with an IPO looming and the Musk trial, but the Microsoft partnership itself also continues to feel like a headache for the AI giant.

QUICK HITS

🛠️ Trending AI Tools

• 🚀 Stitch - Google’s newly updated UI creation tool for “vibe design”

• 📱 Dispatch - Control Claude Cowork on your computer via mobile

• ♻️ MiniMax M2.7 - 'Self-evolving' AI with strong coding, agentic benchmarks

• 🦞 GLM-5-Turbo - Z AI's high-speed agentic model built for OpenClaw

📰 Everything else in AI today

ASAPP's Nirmal Mukhi and special guest Forrester’s Kate Leggett explore how to plan, staff, and operationalize the new roles AI agents bring to the customer service workforce.*

The U.S. Dept. of Defense filed a 40-page rebuttal to Anthropic's lawsuits, arguing its safety limits make it an “unacceptable risk to national security” during war operations.

Xiaomi released MiMo-V2-Pro, a model that topped OpenRouter's charts under an ‘Hunter Alpha’ codename and excels in agent-related tasks and OpenClaw usage.

Microsoft acquired the full team behind Cove, a collaborative AI interface startup, with the company saying its “ideas will live on” at the tech giant.

Midjourney rolled out a preview of its new V8 image model, coming with improved speed, detail, and text rendering, garnering mixed reactions in early testing.

*Sponsored Listing

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Sofia F. in London, UK:

"I've recently started a wine subscription, and while I know which wines I prefer, I've never been great at pairing them with food. I’ve finally solved this by setting up an AI workflow to manage my wine pairings. Every time my subscription arrives, I snap photos of the labels so the AI can index the vintage and tasting notes.

Now, when I finish my weekly shop, I just upload the receipt; the AI parses the ingredients and cross-references them against my wine inventory to generate a 7-day pairing schedule. It’ll even warn me if I’m about to waste a high-tannin Bordeaux on a mid-week stir-fry, suggesting a cheap bottle of plonk from the corner shop instead! My husband, a total wine connoisseur, says it hasn't been wrong yet."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Simo sounds alarm on OpenAI’s ‘side quests’

• Read our last Tech newsletter: China greenlights commercial brain implant

• Read our last Robotics newsletter: Travis Kalanick’s stealth robot play

• Today’s AI tool guide: Generate an actionable SEO audit

• RSVP to next workshop @ 2 PM EST today: Intro to Vibe Coding pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, AI enthusiasts. While OAI was launching a video app, building a web browser, designing hardware, and adding shopping to ChatGPT, Anthropic quietly took control of the enterprise market.

Now, Applications CEO Fidji Simo is reining the company back in — telling staff they "cannot miss this moment because we are distracted by side quests" and refocusing on two key pillars: coding tools and business customers.

In today’s AI rundown:

• OpenAI scrapping ‘side quests’ to catch Anthropic

• Mistral opens its model-training playbook

• Generate a cohesive e-commerce product shoot

• Microsoft redraws its AI org chart

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

OPENAI

🎯 OpenAI scrapping ‘side quests’ to catch Anthropic

Image source: Lovart / The Rundown

The Rundown: OpenAI is overhauling its product strategy to focus on coding tools and businesses after CEO of Applications, Fidji Simo, called Anthropic's enterprise dominance a "wake-up call" in a company-wide meeting, according to the WSJ.

The details:

• Powerful Claude Code and Cowork releases grabbed the lead with business customers, with Simo telling staff OAI is treating the gap as a "code red."

• Simo said OAI “can’t miss the moment because we are distracted by side quests”, coming amid efforts including hardware, adult mode, ads, and more.

• OAI’s 2025 launches included Sora, the Atlas browser, e-commerce features, and more, which insiders said led to confusion and constant compute shuffling.

• It did claw back in coding, with Codex quadrupling its weekly users to 2M+ since January — alongside a new GPT 5.4 model targeting business workflows.

Why it matters: The Pentagon drama may still be fresh in consumer minds in the battle between OpenAI and Anthropic, but where the real war is being fought is on the enterprise side. OAI is pulling in a million different directions, and Simo saying so out loud to the whole company tells you how real the Anthropic gap has gotten.

TOGETHER WITH HUBSPOT

🧠 100+ ChatGPT prompts to revolutionize your workflow

The Rundown: HubSpot’s free, comprehensive  “How to Use ChatGPT at Work” guide provides 100+ ready-to-use prompts to help professionals boost efficiency and adopt AI-driven workflows.

Inside, you’ll find:

• A quick crash course to master ChatGPT in under 30 minutes

• Practical industry use cases to spark real-world inspiration

• 100+ prompts to streamline tasks and accelerate productivity

• Expert tips to tackle common AI roadblocks with confidence

Get your free copy and join 10,000+ professionals leveling up with AI.

MISTRAL

🏗️ Mistral opens its model-training playbook

Image source: Mistral

The Rundown: Mistral launched Forge, a platform that hands enterprises the same training recipes and infrastructure the French AI lab uses internally — allowing companies to build custom models on proprietary data without ever sharing it.

The details:

• Rather than basic fine-tuning, Forge offers full pre-training, post-training, and RL pipelines that mirror how Mistral builds its own flagship models.

• Training can run entirely on a company's own servers with zero data exposure to Mistral, a hard requirement for defense, finance, and government buyers.

• Early partners include ASML, Ericsson, and the European Space Agency, with use cases from legacy code migration to ancient manuscript restoration.

• Forge comes during a busy week of Mistral releases that includes Small 4 and Leanstral, with the French startup also joining Nvidia’s Nemotron Coalition.

Why it matters: Most major enterprise AI boils down to the same thing: take a general model and hope it's close enough. Mistral is making a different bet — that companies sitting on tons of proprietary data, compliance rules, and internal codebases need models trained on that knowledge, not just prompted with it.

AI TRAINING

📸 Generate a cohesive e-commerce product shoot

The Rundown: In this guide, you will learn how to generate a full set of product photos using Grok's Aurora model, building each image off the last one so they all look like they came from the same session.

Step-by-step:

1. Go to grok.com, upload a product photo, and prompt: “Product photo of [your product] on a matte black surface, soft box lighting from the upper left, shot on 85mm lens, f/2.8, shallow depth of field, commercial product photography”

2. Build more shots from the first output by keeping the same studio description and changing only the angle or crop — a 3/4 angle, a close-up, or a white BG

3. Add a lifestyle shot. Describe the product in use: being worn, held, or placed in a real environment. Include the setting, lighting, and mood

4. Lay out all your images. They should look like they came from the same shoot

Pro tip: Try picking a themed studio shoot, like basketball-themed, nature-themed, etc. Grok might surprise you here.

PRESENTED BY UNWRAP

💬 Powerful insights for powerful brands

The Rundown: Unwrap’s customer intelligence platform brings all your customer feedback (surveys, reviews, support tickets, social comments, etc.) into a single view, then uses AI + NLP to surface the most actionable insights and deliver them straight to your inbox.

With Unwrap, you get:

• A clear, single view of customer sentiment

• The ability to ask questions about your feedback using natural language with Assistant

• Real-time alerts from your feedback as they arise

• Tools trusted at scale by Perplexity, Oura, Stripe, Clay, DoorDash, lululemon, Southwest Airlines, and others

Connect with Unwrap to get a free trial of the tools, exclusive to The Rundown AI readers.

MICROSOFT

♻️ Microsoft redraws its AI org chart

Image source: Microsoft AI

The Rundown: Microsoft just overhauled its AI org chart, announcing the merge of its fragmented Copilot teams and shifting Microsoft AI CEO Mustafa Suleyman’s focus squarely towards its five-year mission to build superintelligence in-house.

The details:

• Former Snap exec and new Microsoft AI EVP Jacob Andreou will run the combined Copilot org, which will span across design, product, and engineering.

• Suleyman said the move will “enable me to focus all my energy on our Superintelligence efforts”, with a focus on enterprise systems.

• A reworked OAI partnership cleared the way for Microsoft to build toward AGI on its own, lifting a ban on solo development that ran through 2030.

• Copilot is still struggling for traction, with 6M daily users in February vs. ChatGPT's 440M — and its enterprise add-on reaching just 3% of Office subs.

Why it matters: Microsoft stock is down this year, the legacy software companies are under pressure to prove AI ROI, and Copilot adoption is a fraction of the big players. This reorg is Nadella betting that the fix starts at the model layer just as much as the product one — and that the company needs its own frontier systems to compete.

QUICK HITS

🛠️ Trending AI Tools

• 🧠 Adapt - Your company’s AI computer. Learns your business and autonomously reasons and acts across your stack in Slack*

• 🤏 GPT-5.4 Mini & Nano - OAI’s fast, cheap small models for coding agents

• 🦞 NemoClaw - Nvidia's open-source security layer for OpenClaw agents

• 🧠 Mistral Small 4 - Mistral's new AI combining reasoning, coding, and vision

*Sponsored Listing

📰 Everything else in AI today

OpenAI launched GPT-5.4 mini and nano, two smaller, faster versions of its flagship model built for coding assistants and multi-agent systems.

Mistral released Small 4, an open-source model that merges its reasoning, coding, and vision capabilities into one system.

Anthropic unveiled Dispatch, a Claude Desktop feature that lets users message the assistant from a phone as it works on a PC, running code, browsing, and managing files.

Sam Altman's proof-of-personhood company World launched AgentKit, a tool that lets websites verify a real human is behind an AI shopping agent's purchases.

Google announced that its Personal Intelligence feature is now rolling out to free-tier users across its AI Mode in Search, the Gemini app, and Chrome in the U.S.

Gamma introduced Imagine, an AI design tool baked into its presentation platform that generates logos, infographics, and social graphics with automatic brand styling.

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Dmitry K. in Livonia, MI:

"I have two vehicles and found it challenging to keep meticulous maintenance records over the years. Recently, I set up a Claude project to create a Vehicle Maintenance Logbook. Every time I have a service appointment, I take a picture of the receipt, including the work performed, and upload it to Claude.

It then enters all relevant information into my logbook, while keeping me informed of manufacturer-recommended maintenance and giving me an overall health score for each vehicle. Now I have better maintenance clarity and a fully digital record, which will come in handy when selling either car!”

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Nvidia’s big AI day at GTC

• Read our last Tech newsletter: China greenlights commercial brain implant

• Read our last Robotics newsletter: Travis Kalanick’s stealth robot play

• Today’s AI tool guide: Generate a cohesive e-commerce product shoot

• RSVP to our next workshop on Thursday: AI coding bootcamp pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, tech enthusiasts. China has issued the world’s first commercial approval for an invasive brain-computer interface, marking a regulatory breakthrough that rivals Neuralink and Synchron have yet to reach.

The implant pairs a neural chip with a robotic glove, letting paralyzed patients grip objects by thought — and now, by prescription.

In today’s tech rundown:

• China approves world’s first commercial BCI

• Starcloud wants 88K AI satellites in orbit

• New blood test may predict how long you live

• Samsung kills its trifold after three months

• Quick hits on other tech news

LATEST DEVELOPMENTS

BCI

🧠 China approves world’s first commercial BCI

Image source: Reve AI / The Rundown

The Rundown: China just approved the world’s first invasive brain-computer interface (BCI) for commercial use, letting Shanghai-based Neuracle Medical Technology bring its neural implant to market ahead of Neuralink and every other rival.

The details:

• The system pairs a brain implant with a robotic glove, enabling adults with spinal cord injuries to grip and hold objects via thought-driven signals.

• It’s narrowly targeted: designed for 18–60-year-olds with stable upper-limb paralysis who retain arm movement but can’t grasp.

• The coin-sized wireless implant sits on the surface of the brain’s outer membrane — above the tissue, not inside it, to reduce damage risk.

• Musk’s Neuralink and Synchron are still in trial or demo mode, while Musk says Neuralink will reach “high-volume production” this year.

Why it matters: China’s approval makes this the first invasive BCI cleared for commercial medical use, while rivals, including Neuralink and Synchron, remain in trials. Beijing designated BCI a national "future industry," weaving it into its economic planning, all while moving faster through regulatory channels than the FDA.

SPACE

🛰️ Starcloud wants 88K AI satellites in orbit

Image source: Starcloud

The Rundown: Nvidia-backed startup Starcloud filed plans to build a mega-constellation of 88K satellites designed to host AI workloads in space rather than expand the world’s already strained data center footprint.

The details:

• Starcloud has asked regulators to approve its “orbital data center,” a satellite network built around AI accelerators and cloud servers in space.

• The Redmond-based startup argues the setup could lower cooling costs, cut latency, and offer a credible alternative to land-based server farms.

• The filing drops Starcloud into direct competition with Starlink and Amazon’s Project Kuiper for a finite slice of orbital real estate.

• A planned fleet of 88K spacecraft would dwarf today’s constellations; SpaceX’s Starlink, currently the world’s largest, has about 10K satellites in orbit.

Why it matters: The proposal lands as governments are still sorting out how much of the sky a single private operator can claim. It also puts a finer point on an ongoing debate — whether the promise of space-based AI is compelling enough to justify packing an already crowded, light-polluted orbit even further.

BIOTECH

🩸 New blood test may predict how long you live

Image source: Ideogram / The Rundown

The Rundown: A routine blood test may soon do more than flag high cholesterol — it may forecast who is likely to be alive two years from now, thanks to a newly identified RNA signal that outperforms traditional health markers in predicting short‑term survival.

The details:

• Duke researchers found that levels of six tiny RNA fragments (piRNAs) predict whether people over 70 survive the next two years with up to 86% accuracy.

• The team analyzed 828 small RNAs in blood plasma, alongside health indicators from medical records, assessments, and self-reported lifestyle data.

• People who lived longer consistently showed lower levels of the nine piRNAs linked to aging, with six of them forming the strongest survival predictor.

• In computer simulations, adjusting patients’ piRNA levels to optimal ranges pushed predicted two-year survival from roughly 47% to nearly 100%.

Why it matters: A blood test that forecasts short-term survival is still years from clinical use, but piRNAs represent a class of biomarkers standard panels have never captured. Up next, the research team is testing younger people and probing whether common drugs like metformin or GLP-1s can shift the signal.

SAMSUNG

💀 Samsung kills its trifold after three months

Image source: Samsung

The Rundown: Samsung is pulling the plug on its $2,899 Galaxy Z TriFold less than three months after launch, winding down the dual-hinged, 10-inch phone-tablet hybrid in Korea first, then the U.S. as remaining inventory clears.

The details:

• Released in tiny online drops that sold out in minutes, the TriFold functioned more as a proof-of-concept than a mainstream product, Samsung says.

• Samsung reportedly made little to no profit per unit, squeezed by high manufacturing costs and elevated prices for memory and storage components.

• Foldables remain a small fraction of the overall phone market, but one of the few segments still posting growth as traditional slab phones plateau.

• The exit hands Apple a clean opening: when its long-rumored foldable iPhone eventually arrives, it can position itself as the measured, refined alternative.

Why it matters: Samsung killing its TriFold is a reality check for luxury foldables, evidence that even the category’s dominant player can’t sustain a three-hinge gamble in a market this thin. With Huawei shipping more conventional foldables and Apple lining up its first folding iPhone, Samsung is retreating to safer, higher‑volume designs.

QUICK HITS

📰 Everything else in tech today

Meta’s stock jumped about 3% after reports that the company is weighing layoffs of 20% or more of its roughly 79K employees to rein in soaring AI infrastructure costs.

Apple is buying longtime Final Cut Pro plug‑in maker MotionVFX in a bid to lock more video creators into its subscription ecosystem and fend off Adobe.

The Trump administration is reportedly set to collect an unprecedented $10B “transaction fee” from investors in the new U.S.-controlled TikTok.

Apple rolled out the AirPods Max 2, a $549 refresh of its over-ear headphones that adds an H2 chip with stronger noise cancellation, USB‑C, and live translation.

Dell Technologies disclosed in its latest annual report that its workforce fell about 10% as of January 2026, shrinking by roughly 11K jobs to around 97K employees.

Angel Protection debuted AI software at SXSW that scans existing security cameras for brandished firearms and delivers a human-verified alert to police in seconds.

More than 10 biotech firms, including AI-powered drug discovery platforms, have filed for Hong Kong IPOs this year.

Geely-backed Zeekr is targeting a 2026 South Korea launch with its premium EV lineup, turning Hyundai and Kia’s turf into the next frontline for China’s EV expansion.

New Peloton CEO Peter Stern, a former Apple exec, is steering the company toward GLP-1 weight-loss drug users and a cheaper, mass-market treadmill lineup.

Amazon is hiking the price of its ad-free Prime Video add‑on in the U.S. from $2.99 to $4.99 a month starting April 10, rebranding the tier as “Prime Video Ultra.”

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Nvidia's big AI day at GTC

• Read our last Tech newsletter: Apple’s foldable iPhone leaks

• Read our last Robotics newsletter: Travis Kalanick’s stealth robot play

• Today’s AI tool guide: How to use Grok for free automated research

• RSVP to our next workshop on March 19: AI coding bootcamp pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team

Good morning, AI enthusiasts. Jensen Huang has called OpenClaw "the single most important piece of software, probably ever," and this year's GTC showed just how serious the Nvidia CEO is about adopting the viral agentic tool.

Its NemoClaw security stack for OpenClaw was one of dozens of new announcements, including next-gen chips, AI game graphics, and a wave of enterprise and robotics partnerships that show the chipmaker is spreading far beyond its hardware roots.

In today’s AI rundown:

• Nvidia unloads at GTC with NemoClaw, and more

• Exposed AI band becomes real in Japan

• How to use Grok for free automated research

• Manus brings its AI agent to the desktop

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

NVIDIA

🦞 Nvidia unloads at GTC with NemoClaw and more

Image source: Nvidia

The Rundown: Nvidia CEO Jensen Huang made a wave of announcements at GTC 2026, including an open-source NemoClaw for agents, next-gen Vera Rubin platform, DLSS 5 for photorealistic game graphics, and new enterprise and robotics tools.

The details:

• NemoClaw brings security and privacy guardrails to OpenClaw agents, with an emphasis on growing the agentic tech’s use across enterprises.

• The Vera Rubin platform puts seven new chips into production to power AI training and agents, with Huang also teasing space-based data centers.

• DLSS 5 uses AI to add photorealistic lighting and materials to games in real time, with Bethesda, Capcom, and Ubisoft among the first studios on board.

• A new open-source Agent Toolkit lets enterprises build secure AI agents, along with new AI platforms and partnerships for vehicles, robots, and more.

Why it matters: Huang pitched Nvidia as 'the first vertically integrated but horizontally open company,' and GTC makes that case hard to argue. Chips, agents, game graphics, robotics — every announcement pointed to the same play: own the infrastructure layer beneath all AI workloads, and let everyone else build openly on top.

TOGETHER WITH YOU.COM

🤔 Keep your LLMs from making stuff up

The Rundown: It happens — LLMs hallucinate. Grounding your LLM, however, can help dramatically improve accuracy. In this guide, You.com explains what AI grounding is and how organizations can implement it to achieve more reliable outputs.

The playbook covers:

• A three-part approach that outperforms RAG alone

• Why grounding isn't set-and-forget, and how to build audit trails

• The open vs. closed platform trade-off (and what it means for your next model switch)

Get the guide.

AI & MUSIC

🎸 Exposed AI band becomes real in Japan

Image source: Spotify

The Rundown: A pseudonymous producer called "Kage" built a fictional Japanese metal band called ‘Neon Oni’ with Suno that pulled in 80k+ monthly listeners before fans uncovered it was AI, then hired real Tokyo musicians to perform the AI tracks live.

The details:

• Neon Oni's Spotify page listed fictional member bios and a Tokyo location, with AI-generated music videos, merch, and a growing fanbase.

• Reddit users eventually spotted AI-generated hands in the music videos and traced the creator to Europe, exposing the band as entirely fabricated.

• The creator recruited seven musicians from Tokyo bands to perform the AI music live, with three shows now complete and a headline gig set for March 29.

• From an interview with Kage: "In an age where AI is taking everyone's jobs, this has actually created jobs. It's done the complete opposite."

Why it matters: Initial deception aside, musicians play cover songs and hit songwriters produce tracks for other artists every day. Neon Oni might be the strange, AI-future version of that model: one creator builds the brand and writes the music with AI, then real performers bring it to the stage when a concept or particular sound gains traction.

AI TRAINING

🔎 How to use Grok for free automated research

The Rundown: In this guide, you will learn how to use Grok's Tasks feature. If you have a free X.com account, you get 2 automated tasks each day — use them to build automated daily research briefings pulled from live X data.

Step-by-step:

1. Go to Grok and sign in with your X account. Click on your profile picture in the bottom left, then click Tasks in the pop-up

2. Click New Task, give it a name, set your schedule (daily, weekly, or specific days), and prompt: “Search X for the top trends in [your niche] from the last 24 hours. Summarize the top 3 and flag anything gaining traction”

3. Once you save the task, Grok will run the search on schedule and send you the results via email and as a push notification in the mobile app

4. To view run results on desktop, navigate back to grok.com/tasks and click on the task.

Pro tip: In addition to your two daily tasks, you can also stretch your free Grok account further by scheduling up to 10 research tasks per week or month.

PRESENTED BY LAMBDA

📊 Open-weight arrivals: Lambda's benchmark results

The Rundown: Lambda model cards are clean, single-page reports built around the metrics that actually matter in production, so you can evaluate open-weight models without digging through scattered benchmarks.

Each card includes:

• Model overview and specs to help you evaluate fit

• Real throughput and latency numbers on NVIDIA A100, H100, and B200 GPUs

• Side-by-side comparisons for vLLM and SGLang serving frameworks

• Launch commands and benchmark scripts to reproduce results on your own cluster

Explore the LLM index.

MANUS

🖥️ Manus brings its AI agent to the desktop

Image source: Manus

The Rundown: Manus just launched My Computer, a new desktop app that moves its cloud-based AI agent onto users' local machines to manage files, run terminal commands, build apps, and more.

The details:

• My Computer works through the local terminal, giving the agent direct access to read, sort, and edit files stored on a user's machine.

• Use cases range from organizing unsorted photos into labeled folders, batch-renaming invoices, to building and packaging apps autonomously.

• Meta acquired the Chinese agentic startup in December for $2B, with its team joining the company, and CEO Xiao Hong coming in as a VP.

• The agent can also tap into a machine's hardware when it's sitting idle, running jobs in the background, or completing tasks assigned remotely from a phone.

Why it matters: Manus was already one of the more capable AI agents in the cloud, and now it’s making the desktop move we’ve seen from OpenClaw, Perplexity, and others. The race to be the orchestrator of users’ computers is on, and Manus is a good opportunity for Meta to gain a foothold without a current frontier model of its own.

QUICK HITS

🛠️ Trending AI Tools

• 🗣️ Unwrap Customer Intelligence - Turn unstructured customer feedback into data-backed insights that inform your product roadmap*

• 💻 My Computer - Manus’ AI agent for local machines

• 📐 Get Physics Done - Open-source AI for end-to-end physics research

• 📢 AI CMO - Okara’s agent-powered marketing suite

*Sponsored Listing

📰 Everything else in AI today

Microsoft AI Envisioning Days – Free video series helping software companies build, deploy, and monetize AI apps and agents. Watch now.*

Encyclopedia Britannica and Merriam-Webster sued OpenAI, alleging it scraped their articles, produced competing outputs, and attributed hallucinations to them.

Physical Superintelligence dropped Get Physics Done, an open-source AI agent that can scope research problems, run experiments, verify results, and draft papers.

Moonshot AI published Attention Residuals, a technique that lets models pull from their earlier layers instead of stacking them, delivering 1.25x more compute efficiency.

OpenAI is reportedly restructuring its Stargate computing team, coming alongside a shift in strategy to renting more AI servers rather than building its own data centers.

Meta signed a $27B deal with Nebius to deploy AI cloud infrastructure, including one of the first large-scale deployments of Nvidia's Vera Rubin platform.

*Sponsored Listing

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Barry L. in Houston, TX:

"As a stand-up comedian, I’m constantly having to market and promote both myself and my shows. AI has become my go-to marketing and promotions strategist, helping me grow online and drive more ticket sales.

One of the best prompting tips I’ve learned is ending prompts with: “Ask me clarifying questions until you’re 95% confident you can complete the task successfully.” That tip alone dramatically improved my results.

AI has helped me develop better content ideas, sharpen my messaging, and build a strong Meta ad strategy for ticket sales. I produce a twice-monthly comedy show, and we’ve sold out every show but one.

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Musk takes xAI into a full rebuild

• Read our last Tech newsletter: Apple’s foldable iPhone leaks

• Read our last Robotics newsletter: Travis Kalanick’s stealth robot play

• Today’s AI tool guide: How to use Grok for free automated research

• RSVP to our next workshop on March 19: AI coding bootcamp pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, robotics enthusiasts. Travis Kalanick is back — and he’s got robots. The Uber cofounder has spent years building Atoms, a stealth robotics venture born from his ghost-kitchen empire and now aimed at food, logistics, and even mining.

Forget the bipedal humanoids: Kalanick says wheels will win in the real world.

In today’s robotics rundown:

• Uber’s Travis Kalanick launches robotics startup

• California’s Sunday hits $1.15B valuation

• Rivian’s robot startup just raised $500M

• Unitree’s tennis bot can actually rally

• Quick hits on other robotics news

LATEST DEVELOPMENTS

ATOMS

☄️ Uber’s Travis Kalanick launches robotics startup

Image source: Ideogram / The Rundown

The Rundown: Uber cofounder and former CEO Travis Kalanick is launching Atoms, a stealthy robotics venture that aims to turn his CloudKitchens ghost‑kitchen empire into a fleet of wheeled industrial robots for food, logistics, and mining.

The details:

• CloudKitchens is Kalanick’s delivery‑only restaurant business, which rents out commercial kitchens to virtual restaurant brands accessible via delivery apps.

• Atoms has absorbed CloudKitchens and now pitches itself as a universal wheeled robot base for task‑specific machines, not humanoids.

• The company is in talks to acquire autonomous trucking startup Pronto to accelerate its push into mining and industrial transport.

• In a 1,600‑word manifesto, he cast Atoms as the dawn of a robotics “golden age,” with cheap software and energy fueling swarms of task‑specific robots.

Why it matters: Atoms is pitching a universal wheeled “robot base” for task‑specific machines, with Kalanick insisting wheels will beat humanoid legs in the real world. He says the company has been building in stealth for eight years and is finally ready to put robots to work across delivery, food service, mining, and other logistics‑heavy jobs.

TOGETHER WITH ROBOFLOW

🤖 Robots that See: Live Roboflow with Peer Robotics

The Rundown: Visual AI company Roboflow goes live with Vishrut Kaushik, Senior Robotics Engineer at Peer Robotics, on how vision-enabled mobile robots are transforming factory floors, reducing physical demands and making facilities safer for human workers.

In this session, you’ll learn:

• How Peer Robotics shifted from LIDAR to intelligent vision systems

• How computer vision unlocks #D<-- remove digital twins and advance safety monitoring

• The engineering behind their custom vision models and data pipelines

Register Free.

SUNDAY ROBOTICS

🦄 California’s Sunday hits $1.15B valuation

Image source: Sunday Robotics

The Rundown: California robotics startup Sunday just closed a $165M Series B round, hitting a $1.15B valuation, as the Silicon Valley startup races to put its general-purpose humanoid, Memo, in the home.

The details:

• Founded by former Google DeepMind and Airbnb engineers Tony Zhao and Cheng Chi, Sunday is building Memo, a waist-high household humanoid.

• Memo’s autonomy leans on Sunday’s “Skill Capture Glove” system, which has collected nearly 10M real-home chore demos from hundreds of households.

• Sunday says it raised this money “to stop doing demos” and will use it to ship a beta fleet of Memo robots to early users later this year.

• The company says it already has roughly 1K people on a waitlist for Memo and is leaning hard on modern AI techniques and large-scale manipulation data.

Why it matters: Sunday’s new funding drops it deep into the race to build general-purpose home robots, alongside Optimus, Figure, and Apollo. At the same time, it’s rubbing shoulders with a different wave of bots — like Amazon’s Astro — that want to be the “helpful robot in your living room,” just with very different shapes and skill sets.

MIND ROBOTICS

🦾 Rivian’s robot startup just raised $500M

Image source: Stripe (YouTube)

The Rundown: Rivian spin-out Mind Robotics has raised $500M at a $2B valuation to automate the factory tasks that conventional robots still can’t do — and its founder, Rivian CEO RJ Scaringe, thinks that humanoids are solving the wrong problem.

The details:

• Founded in November 2025 by Scaringe, Mind Robotics has now banked $615M in total funding.

• The startup is building a full-stack industrial robotics platform to automate dexterous factory tasks that conventional industrial arms cannot handle.

• Rivian, a major shareholder, is Mind’s launch partner, opening its factories, engineering talent, and production data as a training testbed for the robots.

• Mind wants robots on factory lines by the end of this year, carving out space between legacy industrial vendors and humanoid startups.

Why it matters: Humanoids are grabbing headlines, but Scaringe insists the real factory value is in hands — manipulation and physical reasoning, not backflips. That pits Mind against both legacy industrial giants and humanoid players like Tesla and Figure, which he argues add complexity and power draw without adding line value.

ROBOTICS RESEARCH

🎾 Unitree's tennis bot can actually rally

Image source: Github

The Rundown: A Unitree G1 humanoid trained by Chinese researchers can now hold its own on a tennis court — returning balls, tracking footwork, sustaining multi-shot exchanges with human opponents.

The details:

• The system, called LATENT, was trained on just five hours of motion-capture data — short, imperfect clips of human swings and movement.

• It returns balls traveling over 15 m/s with roughly 90% accuracy, enough to keep rallies going rather than just reacting to staged shots.

• The team has open-sourced the approach on GitHub, framing LATENT as a reusable template for teaching athletic skills from messy real-world data.

Why it matters: The G1 retails for around $16K — cheap by humanoid standards — and it learned fluid, unscripted court play without pristine training data or choreographed setups. If the same pipeline transfers to other high-speed physical tasks, “five hours of motion capture” could become the new baseline for robot athletic training.

QUICK HITS

📰 Everything else in robotics today

Uber started offering Hyundai-owned Motional’s self-driving Ioniq 5 robotaxis, with safety drivers for now, for rides between select Las Vegas hotspots via its app.

Lucid Motors unveiled a steering-wheel-free “Lunar” robotaxi concept built on its upcoming mid-size EV platform, while outlining future plans with Uber and Nuro.

Engineers built a wearable two-legged robot that attaches behind a person, walking in sync to carry loads and effectively turning the user into a cyborg “centaur.”

Samsung just created a dedicated “Hand Lab” inside its robotics division to develop highly dexterous, tendon‑driven robotic hands for future humanoids.

Uber, Wayve, and Nissan plan to pilot a Nissan Leaf–based robotaxi service on Uber’s app in Tokyo in late 2026, marking Uber’s first robotaxi partnership in Japan.

Two UK brothers built a 3D‑printed robot that uses custom algorithms to solve a 4×4×4 Rubik’s Cube in 45.3 seconds, earning a new Guinness World Record.

A Unitree humanoid in Macau was removed from the street by police after it startled a 70‑year‑old woman, prompting a hospital visit and an official warning to its operator.

Researchers built a bicycle‑style robot that balances on two wheels at high speed while using AI to dodge obstacles in real time.

Chinese startup XGSynBot unveiled the Z1, a wheeled humanoid for factories that can swap modular tools like grippers and welders in under six seconds.

Researchers built a silicon photonics chip that lets robots and drones generate 3D maps while simultaneously measuring the speed of moving objects.

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Musk takes xAI into a full rebuild

• Read our last Tech newsletter: Apple’s foldable iPhone leaks

• Read our last Robotics newsletter: Figure’s humanoid cleans living rooms

• Today’s AI tool guide: Create an intro animation in 15 minutes

• RSVP to our next workshop on March 19: AI coding bootcamp pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team

Good morning, AI enthusiasts. Three years ago, Elon Musk assembled 11 co-founders to take on OpenAI and Anthropic with a new AI startup. Nine of them are now gone.

With Musk admitting xAI is now "being rebuilt from the foundations up," and raiding Cursor for coding talent, it is clear that the billionaire is suddenly running a startup reset while preparing for one of the largest public listings in history.

In today’s AI rundown:

• xAI down two more co-founders as Musk rebuilds

• The Rundown Roundtable: Our AI use cases

• Create an intro animation for your brand

• AI-designed vaccine shrinks rescue dog's tumor

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

XAI

🚪 xAI down two more co-founders as Musk rebuilds

Image source: Lovart / The Rundown

The Rundown: Elon Musk just posted that xAI "was not built right" and needs a ground-up rebuild, with 9 of 11 original co-founders now gone and the company actively hiring major coding talent to help Grok catch up to the industry’s frontier.

The details:

• Zihang Dai and Guodong Zhang are the latest departures, leaving just two of the original 11 co-founders (Manuel Kroiss and Ross Nordeen) at xAI with Musk.

• Zhang, who led Grok Code and reported directly to Musk, was reportedly blamed by Musk for Grok’s coding shortfalls before departing.

• Musk said xAI is “being rebuilt from the foundations up”, which follows a major reorg that saw dozens of employees leave the company.

• xAI hired senior Cursor leaders Andrew Milich and Jason Ginsberg last week, coming after Musk's admission that Grok is "currently behind" on coding.

Why it matters: It’s been a year of highs and lows for xAI, but it appears Musk is tearing things down for a major rebuild. With splashy new coding hires and a nearly complete exodus of co-founders, xAI has some serious work to do to hit Musk’s lofty goals — especially while trying to navigate an upcoming IPO in the process.

TOGETHER WITH BLAND AI

📞 Voice AI for every conversation

The Rundown: Bland AI automates phone calls for over 250+ enterprise customers. No phone trees. No hold music. Just faster, smarter customer conversations.

Here are some of the outcomes they've driven for businesses:

• Idaho Finance saved $750k/yr by replacing their IVR with AI Voice Agents

• MyPlanAdvocate added $40M/yr by automating their inbound lead qualification

• And Needle saves $1M/yr by automating outbound calls

Book a demo today to see how they can work for your business.

THE RUNDOWN ROUNDTABLE

💡The Rundown Roundtable: Our AI use cases

The Rundown: The Rundown Roundtable is a weekly feature where we poll members of The Rundown staff about how we use AI in our work and daily lives.

Billy, University Educator: “Each day my OpenClaw checks the local inventory of Toyota dealerships within 250 miles for black or white RAV-4's with heated seats because my wife wants to buy the most popular car in the world and they don't stay on the lots very long.”

Zach, AI Writer: I have been testing Perplexity Computer, and it really makes you rethink a lot of workflows. It has been particularly useful for surfacing interesting and viral social content across X and Reddit, with internet search abilities that seem much more capable for quick canvassing of different platforms.

I have also used it for clipping and captioning video content for social, with Computer able to find, transcribe, give recommendations on the most viral quotes/sections, and caption in just minutes — a wild process that would normally take me (with little video experience) much longer and have to leverage a ton of different apps to make happen.

AI TRAINING

🎥 Create an intro animation for your brand

The Rundown: In this guide, you will learn how to use Manus to generate branded intro, transition, and outro bumper videos for your brand with no motion designer or video editor required.

Step-by-step:

1. Open a Manus task, upload logo/wordmark, and prompt: “I want a branded intro bumper, transition, and outro bumper video based on these logos. Ask 3-5 MCQs about style and vibe before starting. Use the video generator skill.”

2. Answer the questions Manus asks about vibe, animation style, and colors. When it shows options, ask it to visualize them so you can pick the best one.

3. Type “/video-generator” and tell it to generate the chosen concept — Manus will generate keyframes using Nano Banana and then animate them with Veo 3.

4. Use the default mode to save on tokens. Once it’s done, you can download the finished clips and drop them into your video editor.

Pro tip: Manus will try to find and add royalty-free audio to your clips. We recommend you tell it to skip this. It doesn’t usually find the best audio and ends up eating tokens.

PRESENTED BY BLOTATO

💻 Build your first AI marketing team in Claude Code

The Rundown Publish 100+ on-brand social media posts per week with your AI marketing team. This simple guide shows you how to use Claude Code + Blotato MCP to automate your brand's growth, with no coding required.

In this course, you’ll learn how to:

• Install Claude Code

• Teach Claude to write like you

• Make infographics, carousels, and videos

• Schedule & publish directly to 9 social media platforms

Watch the course and start building your AI marketing team with Blotato MCP.

AI & MEDICINE

🧬 AI-designed vaccine shrinks rescue dog's tumor

Image source: Paul Conyngham / The Australian

The Rundown: Sydney AI consultant Paul Conyngham built a custom mRNA cancer vaccine for his dog Rosie by chaining ChatGPT, Grok, DeepMind’s AlphaFold, and UNSW's genomics lab to turn 350 GB of tumor data into a treatment with real results.

The details:

• Rosie was diagnosed with mast cell cancer in 2024 and given months to live despite chemo and surgery.

• Conyngham used ChatGPT to map the research, paid $3K for genomic sequencing, then fed the data through AlphaFold to model Rosie's mutations.

• The UNSW RNA Institute helped turn the formula into a custom vaccine, with Conyngham also revealing the “final vaccine construct was designed by Grok”.

• One tumor shrank by half after her December injection, with Conyngham now working to create a second vaccine for her other non-responding tumors.

Why it matters: A year ago, a pet owner with no biology training couldn't turn a cancer diagnosis into sequenced DNA data, modeled proteins, and a working vaccine blueprint. Rosie isn't completely cured, but the fact that AI provided the tools to take a real swing at a previously hopeless situation with real impact is a powerful thing.

QUICK HITS

🛠️ Trending AI Tools

•  🔄 Crafting for Agents - Promote your enterprise coding agents to senior software engineers and ship code safely with closed-loop validation*

• 🧠 Nemotron 3 Super - Nvidia's 120B reasoning AI with 1M token context

• 🚀 Claude - Anthropic’s AI, now able to create charts and diagrams in chat

• 🗣️ TADA - Hume's TTS AI that syncs text and audio for no hallucination speech

*Sponsored Listing

📰 Everything else in AI today

The global launch of Seedance 2.0, ByteDance’s viral video AI, is reportedly being suspended, with the delay coming after a major copyright backlash from Hollywood.

Elon Musk revealed that Tesla’s Terafab semiconductor manufacturing facility is launching in a week, aiming to create custom silicon chips for use across its tech.

A Florida man used ChatGPT to help handle selling his home, including pricing, marketing, scheduling, and contracts, closing in five days and saving 3% in agent fees.

Meta is reportedly planning layoffs that could cut 20% of its nearly 79K workforce, as the company looks to offset $600B in planned AI infrastructure spending.

Ex-Anthropic researchers are set to raise $175M for Miraendil, a new AI startup building specialized AI for scientific R&D in biology and materials science.

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Adam B. in New York City, NY:

"I’m looking for a business to buy, so I used ChatGPT and Claude to develop a personal OS to help analyze potential deals. I started by telling ChatGPT my personal preferences, which included my financial overview, personality traits, business types, and then took it to Claude and asked it what it thought, and asked it to analyze my OS.

I took its advice and brought it back to Chat, and went back and forth until I got to v31. Now, when I receive a business’s CIM and any other docs, I upload them to Claude alongside my OS, and it generates a 20+ page full evaluation, not only of the business, but also of how the business is for me specifically. In 5 minutes, I get to know whether to pursue or not.

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Google brings Gemini to the road

• Read our last Tech newsletter: Apple’s foldable iPhone leaks

• Read our last Robotics newsletter: Figure’s humanoid cleans living rooms

• Today’s AI tool guide: Create an intro animation in 15 minutes

• RSVP to our next workshop on March 19: AI coding bootcamp pt. 2

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown