Get the latest AI news, understand why it matters, and learn how to apply it in your work — all in just 5 minutes a day. Join over 2,000,000+ subscribers.

Join Free
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Robotics

Clone's $20K synthetic human

Jennifer Mossalgue • 6 minutes

Read Online | Sign Up | Advertise

Good morning, robotics enthusiasts. Clone Robotics has always been the humanoid outlier, building eerie fluid-muscle “synthetic humans” in Poland while better-funded rivals pursue the more conventional bots.

Now CEO Dhanush Radhakrishnan is promising a sci-fi roadmap: surgical-tool dexterity by 2026, human-like walking by 2027, and hotel butlers by 2028, all for under $20K a unit. But can Clone turn its wild vision into something that actually works?

In today’s robotics rundown:

  • Clone Robotics unveils $20K android roadmap

  • GEN-1 bot folds boxes, fixes vacs, and rarely misses

  • Hermeus nabs $350M for Mach 5 unmanned aircraft

  • VC Eclipse raises $1.3B to wire AI into robots

  • Quick hits on other robotics news

LATEST DEVELOPMENTS

CLONE ROBOTICS

🤖 Clone Robotics unveils $20K android roadmap

Image source: Clone Robotics / Reve AI

The Rundown: At the 2026 Abundance Summit, Clone Robotics co-founder and CEO Dhanush Radhakrishnan unveiled a “synthetic human” roadmap, announcing sub-$20K androids, a new Silicon Valley hub, and “Robo Butler” deployments by 2028.

The details:

  • Clone is pursuing “synthetic humans,” androids built around an anatomically accurate polymer skeleton actuated by fluid-driven “Myofiber” muscles.

  • The company says it is opening a hub in Mountain View, in the heart of Silicon Valley, to scale R&D, hiring, and partnerships around its android platform.

  • Radhakrishnan says the company can manufacture a full musculoskeletal android for under $20K by producing Myofiber muscles at scale.

  • The roadmap calls for a torso platform capable of using tools like scalpels by late 2026, human-like walking by 2027, and robot butlers in hotels by 2028.

Why it matters: A Mountain View office and a reported $50M raise put Clone in the same conversation as Figure and 1X, but not on the same timeline. It’s 2028 target trails rivals already in factory deployments, and the case for fluid-muscle androids at scale rests entirely on manufacturing claims that have yet to be tested in production.

GENERALIST

🛠️ GEN-1 bot folds boxes, fixes vacs, and rarely misses

Image source: Generalist

The Rundown: San Mateo–based startup Generalist says it just crossed a threshold that robotics has been chasing for years: a general-purpose AI model that performs real-world manipulation tasks with 99% reliability.

The details:

  • The company says its GEN-1 physical AI model can fold boxes, fix vacuums, and pack phones, without fumbling a third of them like its predecessor.

  • On box assembly, GEN-1 reportedly clocks in at 12.1 seconds, compared to 34 seconds for both GEN-0 and Physical Intelligence’s pi-0.

  • The base model is pretrained entirely on human “data hands,” only encountering robot hardware during a final hour of task-specific adaptation.

  • Unlike scripted factory robots, GEN-1 claims to recover from mid-task disruptions and handle objects or situations it was never explicitly trained on.

Why it matters: A one-in-three failure rate is a non-starter on a factory floor, but at 99% and 3x the speed, the game changes. Generalist admits some tasks still can’t hit that bar — but for a model that needs just one hour of robot-specific training to deploy, the lab-to-production gap just got a lot smaller.

HERMEUS

✈️ Hermeus nabs $350M for Mach 5 unmanned aircraft

Image source: Hermeus

The Rundown: Defense startup Hermeus raised $350M to develop what it calls the fastest unmanned aircraft on Earth, a funding round that pushes the LA-based company to a $1B valuation.

The details:

  • The company pulled in $200M in new equity led by Khosla Ventures, plus $150M in debt to scale hardware without diluting control.

  • The startup had been building its own engine from scratch before partnering with RTX subsidiary Pratt & Whitney to modify its F100 to shorten timelines.

  • Hermeus completed its first test flight last May with a subsonic demonstrator, then flew a new iteration nearly 3x larger and 4x heavier a few weeks ago.

  • CEO AJ Piplica models Hermeus on SpaceX — build, test, fail, repeat — with supersonic flight next on the ladder and Mach 5 as the ultimate target.

Why it matters: VC investment in defense tech crossed $9B last year, and Hermeus is positioning itself as the software-era answer to a very hardware problem. If it can actually field a Mach 5 drone fighter on startup timelines, it could reset expectations for how fast an entirely new class of aircraft gets designed, tested, and deployed.

ECLIPSE

💰 VC Eclipse raises $1.3B to wire AI into robots

Image source: Mind Robotics / Reve AI

The Rundown: Eclipse, the Palo Alto hard-tech VC best known for backing AI chipmaker Cerebras, raised $1.3B to wire AI directly into the physical stack of robots, factories, and energy systems.

The details:

  • The raise is reportedly split into $591M for early-stage investments and $720M for later-stage deals, disclosed in SEC filings.

  • The new funding surpasses Eclipse’s previous high of $1.23B in 2023, bringing total assets under management to roughly $10B.

  • The portfolio already includes battery recycling firm Redwood Materials, Arc, Bedrock Robotics, Wayve, and industrial robotics lab Mind Robotics.

  • Partner Jiten BehlBehl told TechCrunch that Eclipse will also incubate some startups in-house, a process he says is already underway.

Why it matters: Most physical AI investors write checks and wait. Eclipse is taking a more hands-on approach, building companies internally while pushing its existing portfolios in adjacent sectors to become each other’s customers and partners, a structure designed to de-risk the long road from prototype to commercial scale.

QUICK HITS

📰 Everything else in robotics today

BMW is turning its historic Munich plant into an iFACTORY where 800 robots and AI will automate about 60% of logistics while building only electric cars by 2027.

Waymo launched its public robotaxi service in Nashville, its 11th U.S. city, and will gradually expand access while partnering with Lyft.

Chinese startup LingXin SmartHand, a leading maker of dexterous robotic hands, is reportedly seeking new funding at a valuation of around $2.9B.

Syncere opened preorders for Lume, a $1,499 robotic floor lamp that doubles as a laundry‑folding home robot and aims to bring “ambient robotics” into bedrooms.

BYD patented an autonomous robot that can navigate to parked EVs, scan their battery and tire pressure, then automatically charge the vehicle and inflate all four tires.

Chinese robotaxi company Pony.ai launched an invitation-only robotaxi passenger service in Singapore’s Punggol district in partnership with ComfortDelGro.

German-Chinese robot maker Kuka said many European factories are too slow to adopt AI, so it plans to prioritize growth in the U.S. and Asia instead, Bloomberg reports.

Chinese startup D-Robotics raised an additional $150M to expand its embodied AI platform and developer ecosystem, taking its total Series B funding to $270M.

The U.S. may bar federal agencies from buying Chinese-made humanoids, a shift that could hand Hyundai’s Boston Dynamics a major opening in U.S. robotics.

U.S. researchers built an AI‑powered robotic guide dog that uses an LLM to talk with visually impaired users, explain route options, and narrate surroundings in real time.

German city mayors are urging a night-time ban on robot mowers, warning that the machines’ blades are injuring and killing hedgehogs and nocturnal wildlife.

Japan opened its first 3D-printed railway station, built from concrete components printed off-site using an ABB industrial robot, then assembled overnight.

U.S. regulators closed their probe into Tesla’s remote parking feature without a recall after finding only rare, mostly minor crashes among millions of uses.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Meta Superintelligence Labs ships its first model

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Last summer, Mark Zuckerberg handed Alexandr Wang the keys to a brand new lab with some expensive, freshly poached talent from top rivals. Today, we got the first look at what that money built.

Meta's Muse Spark isn't topping every benchmark, but it puts the tech giant squarely back in the game — with the resources, data, and 3B+ daily users to build on from here.

In today’s AI rundown:

  • Meta Superintelligence Labs ships first model

  • HeyGen’s Avatar V solves AI’s identity drift

  • Build an automated ad generator with this tool

  • Anthropic simplifies the agent-building system

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

META

🚀 Meta Superintelligence Labs ships its first model

Image source: Meta

The Rundown: Meta’s Superintelligence Labs just rolled out Muse Spark, a multimodal reasoning model that marks the highly anticipated debut release of Alexandr Wang’s high-profile division assembled last summer.

The details:

  • Muse Spark handles voice, text, and image inputs, with a contemplating mode that pits multiple agents against each other on hard problems.

  • The model’s benchmarks are competitive with frontier rivals like Opus 4.6 and GPT 5.4 on reasoning, though it lags in coding and tests like ARC-AGI 2.

  • Muse Spark is particularly strong in health reasoning, with the company prioritizing the area as part of its ‘personal superintelligence’ mission.

  • Unlike the Llama family, Muse Spark is proprietary, with Meta saying it hopes to open-source future versions but has not committed to a timeline.

  • Wang took over Meta Superintelligence Labs 9 months ago after Zuck acquired Scale AI for $14.3B, saying the team “rebuilt our AI stack from scratch".

Why it matters: Meta is back in the game. While still sitting below the top models, Muse Spark is a serious change from where Meta sat with its Llama family. It may not break the internet, but with tons of resources, valuable data across its platforms, and billions of users, Meta’s AI efforts just took a step in the right direction.

TOGETHER WITH BLOTATO

💻 Claude Cowork for beginners

The Rundown: Claude's Cowork tool can dramatically speed up your workflow — and this free beginner course walks you through everything from setup to building a fully functional AI marketing team inside the platform.

In this course, you’ll learn how to:

  • Setup Cowork

  • Create Skills and MCP

  • Build your AI marketing team

  • Manage social media content calendar

Watch the course and 10x your productivity with Claude Cowork.

HEYGEN

🎭 HeyGen’s Avatar V solves AI’s identity drift

Image source: HeyGen

The Rundown: HeyGen released Avatar V, a new model the company calls “the most realistic AI avatar model in the world,” and claims that it can eliminate identity drift — the tendency for AI-generated faces to stop resembling the user over time.

The details:

  • The system builds a full video avatar from a short 15-second phone recording, capturing the user’s real facial details, gestures, and movement patterns.

  • The model also separates identity from appearance for the first time, allowing users to record once, then swap outfits and backgrounds without filming again.

  • HeyGen says Avatar V outperformed Google's Veo 3.1 on accuracy and lip sync in internal tests, while also beating out Kling and Seedance in blind tests.

Why it matters: Just like image and video models, AI avatars have come a ridiculously long way over the last few years, going from simple mouth movements to mimicking a user’s micro-movements for indistinguishable outputs. While some may scoff at the idea of an ‘AI twin’, the content creation landscape is changing with or without them.

AI TRAINING

📺 Build an automated ad generator with this tool

The Rundown: In this guide, you will learn how to turn a product photo into a finished video ad using ElevenLabs Flows. It's a new workflow builder that bundles image, video, voice, and music in one place.

Step-by-step:

  1. Open ElevenLabs, click ElevenCreative > Flows, then hit + New Flow. Name it [product line] ad template so you can reuse it

  2. Add an Image Generation node and upload product 1-3 product shots. Prompt with a scene, like: “The product (see references) on a white pedestal, studio product shoot, soft morning light, photorealistic product shot”

  3. Add a Video Generation node, drag a line from the image node's output into the video node's start frame, and prompt: “Slow cinematic push-in on the product, soft morning light drifting across the scene, shallow depth of field”

  4. Click Run on the video node > Run till here to generate an image and video in one go. Then, swap out image/video prompts, and quickly iterate on creatives

Pro tip: For audio, add a Text-to-Speech or Music node and connect it to a Mix Audio node alongside the video. You can also try this on other products by duplicating the canvas and swapping in new images.

PRESENTED BY VANTA

🔒 A startup guide to SOC 2 in 30 minutes

The Rundown: If customers, investors, or your sales team keep asking about SOC 2 and you haven't started yet, Vanta's upcoming live session on April 22 is designed to help you take the first step with confidence — all in just 30 minutes.

The session will cover:

  • What SOC 2 actually is, when you need it, and the difference between Type I and Type II

  • Where startups lose time and how to avoid common rework during the process

  • A practical framework for evaluating tools and auditors without overbuying

Register now for the April 22nd session. Can't make it? Sign up anyway to get the recording.

ANTHROPIC

⚒️ Anthropic simplifies the agent-building system

Image source: Anthropic

The Rundown: Anthropic opened a public beta for Claude Managed Agents, a new platform that lets developers go from an agent idea to a live product in days — handling all the backend plumbing that used to take engineering teams months to set up.

The details:

  • Users pick the task, tools, and guardrails, with Managed Agents handling running, securing, and controlling what the agentic system can access.

  • Agents can work solo for hours without dropping state, with a coordination mode also in preview, letting one agent farm out subtasks to others.

  • Notion, Rakuten, Asana, and Sentry are early adopters, with Rakuten reportedly setting up agents across five departments in about a week each.

  • Each agent session costs $0.08 per hour on top of the usual AI usage fee, with users paying based on consumption instead of upfront platform fees.

Why it matters: Anthropic continues to roll out features that eat away at the complexities of users getting the most out of their models and tools. Managed Agents now does the same, simplifying the agentic building process and making it possible for anyone to deploy and control agents without the typical backend headaches.

QUICK HITS

🛠️ Trending AI Tools

  • 🤖 Scrunch - See how AI interprets your site, run a free audit, and unlock the new way to reach customers*

  • 🧠 Muse Spark - Meta’s multimodal reasoning AI with multi-agent mode

  • 🎥 Avatar V - HeyGen's AI avatar model that generates studio-quality videos

  • 🕹️ Clicky - Open-source AI teacher that lives next to your cursor

*Sponsored Listing

📰 Everything else in AI today

Elon Musk amended his OAI lawsuit to redirect all damages to the nonprofit arm and push Altman off its board, with OAI calling it "a harassment campaign."

Perplexity hit $450M in estimated annual recurring revenue after a 50% monthly jump, driven by its Computer agentic system and usage-based pricing model.

Elon Musk revealed that xAI has seven new models currently in training on its Colossus 2 supercomputer, including massive 6T and 10T parameter systems.

Canva acquired Simtheory and Ortto, adding agentic AI workspace tools and marketing automation to its platform as it pushes end-to-end campaign workflows.

Jeff Bezos’ secretive AI startup Prometheus poached Kyle Kosic, a former xAI co-founder who led the infrastructure team before leaving the startup for OAI in 2024.

OpenAI published a child safety policy blueprint pushing for updated U.S. laws on AI-generated CSAM, stronger reporting, and built-in safeguards to prevent exploitation.

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 Lynne B. in the United Kingdom:

"I am a serious amateur photographer. I use in-camera techniques to create expressionist images, and I process the images in Photoshop. I also provide Zoom sessions on my methods for others. Many photographers do not have the facility to create multiple exposures in camera.

I am using ChatGPT to write code for Scripts within Photoshop. These scripts simulate the in-camera multiple exposure process and speed it up. It has taken a while to stop the error codes as the Script area of Photoshop can be temperamental, but the great thing about AI is, it matches my own tenacity with a challenging task."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Anthropic's new AI is too powerful for the world

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Anthropic employees are calling it a "turning point in history." A Cisco exec says a "threshold has been crossed." That's the talk around the new Claude Mythos Preview — but you're not getting access to it.

The model is instead being deployed through Project Glasswing, a new defensive cybersecurity coalition with major tech partners tasked with securing the world's most critical software before similar capabilities end up in the wrong hands.

In today’s AI rundown:

  • Anthropic’s Project Glasswing shows off Mythos AI

  • Open-source AI pushes forward with Z AI’s GLM-5.1

  • Get to inbox zero with this Claude prompt

  • Anthropic continues to rise, locks in 3.5GW compute

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

😱 Anthropic’s Project Glasswing shows off Mythos AI

Image source: Anthropic

The Rundown: Anthropic introduced Project Glasswing, a cybersecurity coalition with AWS, Apple, Google, Microsoft, Nvidia, and 7 other partners built around Claude Mythos Preview, a new unreleased frontier AI with extremely powerful capabilities.

The details:

  • Mythos flagged thousands of security flaws across every major OS and browser, including bugs that survived 27 years of review and millions of scans.

  • Its benchmarks show big improvements over both Opus 4.6 and other frontier rivals across coding, reasoning, and nearly every other domain.

  • The model will not be released publicly, instead limiting access to 12 launch partners and 40+ other orgs for defensive security backed by $100M in credits.

  • Anthropic's Sam Bowman called it “an uneasy surprise” after Mythos emailed him from a test instance that wasn't supposed to have internet access.

  • Mythos was the subject of leaks after a blog draft was found in unpublished files last week, with Anthropic using the model internally since February.

Why it matters: If you ever wonder what type of models the top labs have under wraps, Mythos is a nice preview of the answer. Anthropic thinks it’s so powerful it won’t even release it publicly, instead giving time for the company (and its group of partners) to work on cybersecurity and safety rollouts for future Mythos-level general models.

TOGETHER WITH LAMBDA

📶 Optimize your AI training runs

The Rundown: Most AI training runs use less than half their available compute. Lambda's team found the root causes and built a tested playbook that boosted efficiency and cut costs by over 25%, without changing the model.

Learn how to:

  • Find the memory inefficiencies quietly inflating your costs

  • Optimize your training configuration to make full use of your hardware

  • Address the bottlenecks slowing down GPU communication

Get the optimization guide.

ZHIPU AI

🚀 Open-source AI pushes forward with Z AI’s GLM-5.1

Image source: Zhipu AI

The Rundown: Chinese AI lab Z AI just released GLM-5.1, a new open-source coding model that competes with frontier rivals on coding benchmarks and is built for marathon autonomous sessions of up to 8 hours straight.

The details:

  • GLM-5.1 hit 58.4 on SWE-Bench Pro, topping both GPT-5.4 and Opus 4.6 and marking a rare moment for open source at No. 1 on a top coding benchmark.

  • Z AI also said the model can “stay effective on agentic tasks over much longer horizons”, showing strong results over longer, complex problems.

  • In tests, Z AI had GLM-5.1 build a working Linux desktop as a web app over 8 hours, including a file browser, terminal, and games, without human guidance.

  • The model also shows top performance in Arcada Labs’ Design Arena, coming in second for creative web design after Claude Opus 4.6.

Why it matters: Top Chinese labs continue to be on the tail of the frontier, with GLM-5.1 showing the strongest coding yet — along with long-horizon task capabilities that the company said are the “most important curve after scaling laws”. An open-source model with this coding performance says a lot about how fast the gap is closing.

AI TRAINING

📧 Get to inbox zero with this Claude prompt

The Rundown: In this guide, you will learn how to run email triage once in Claude, then have Claude write a recurring task prompt for your inbox. You will end up with a daily cleanup workflow based on your real rules, not a generic prompt.

Step-by-step:

  1. Open Claude with Gmail access and run one triage session on your unread inbox. The goal is to show Claude what matters before automating the process

  2. Prompt: “Generate an interactive email triage report for the last 24 hours. Sort each email into exactly one of these buckets: Needs response, Needs attention, Archive, Archive and unsubscribe. For each email, include the sender, subject, a one-line reason for the category, a direct link to the email, and the item number. Add labels to the approved emails”

  3. Review, correct misfires, and prompt: Turn this workflow into a recurring task prompt for my inbox, with my common senders, archive, and unsubscribe rules

  4. Save that prompt as a Claude Cowork scheduled task so it can run every morning without rebuilding the logic

Pro tip: Set up Gmail rules around the labels Claude adds. It can apply labels through the connector, so “Needs Response” and “Needs Attention” can be auto-starred, and Archive emails can be auto-archived.

PRESENTED BY OPTIMIZELY

🎓 Build 3 AI agents in 5 days for free

The Rundown: It’s time to put AI to work inside your marketing; embedded in your workflows and eliminating the drudge work. Consider this your official invite to Opal U: AI Marketing University from the team at Optimizely Opal.

Here’s what you’ll get:

  • Live workshop (free) with 50 senior marketers - 1 hour a day (Mon-Fri)

  • 3 working AI agents you build and take back to your team

  • 5 hours invested to save 10+ hours every week

Join the next Opal U session starting Monday.

ANTHROPIC

💰 Anthropic continues to rise, locks in 3.5GW compute

Image source: Anthropic

The Rundown: Anthropic signed a multi-gigawatt compute deal with Google and Broadcom, locking in 3.5GW of TPU capacity for 2027, while also sharing new surging revenue numbers and enterprise growth despite its battle with the U.S. government.

The details:

  • Since January, Anthropic's run-rate revenue tripled to $30B, and its $1M+ enterprise customer base doubled to 1,000+, forcing the compute expansion.

  • Broadcom will supply 3.5GW of Google's TPUs starting in 2027, nearly all US-based — adding to the $50B Anthropic pledged for domestic AI buildout.

  • The revenue projections put the company ahead of rival OpenAI’s recent report of $2M / month in revenue, while both race towards an IPO.

  • The growth also comes despite the Pentagon labeling Anthropic a supply-chain risk, a move the company says rattled over 100 enterprise clients.

Why it matters: Tripling run-rate revenue while facing the Pentagon is quite the move, and shows demand for Claude is still off the charts, even if the U.S. government is blacklisting it. But given the recent rate limit issues, more compute is certainly a welcome sight — especially with behemoth models like Mythos waiting in the wings.

QUICK HITS

🛠️ Trending AI Tools

  • 👨‍💻 The Box Agent - Put your enterprise content to work, securely and at scale*

  • 🚀 GLM-5.1 - Z AI’s new open-source model for long-horizon agentic coding

  • 🔎 Harrier - Microsoft Bing’s SOTA embedding model for search and RAG

  • ⚙️ SWE-1.6 - Cognition's updated coding AI for speed and smoother agent UX

*Sponsored Listing

📰 Everything else in AI today

You.com's new guide shows you how to find the highest-impact AI use cases, from internal workflows to customer experiences. Get the guide here.*

A new mystery model named ‘HappyHorse-1.0’ debuted at No .1 on Artificial Analysis’ video leaderboards, surpassing ByteDance’s viral Seedance 2.0.

OpenAI, Google, and Anthropic are cooperating on identifying and limiting Chinese rivals from distilling their systems, sharing info via a “Frontier Model Forum” non-profit.

Microsoft’s Bing team open-sourced Harrier, a SOTA embedding model for search and retrieval that supports 100+ languages and powers its AI agent grounding service.

Intel announced that it is joining Elon Musk’s recently unveiled Terafab project, saying the company will “help accelerate Terafab's aim to produce 1 TW / year of compute”.

*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 Beth T. in Belmont, NJ:

"I'm taking pickleball lessons, so I created a custom GPT to help me learn more in between sessions. I use different prompts for scoring, strategy, rules, and quizzes, depending on what I want to practice.

It's a simple way to reinforce what I'm learning, ask follow-up questions, and get clear explanations. It's like having a personalized pickleball coach between lessons."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

See you soon,

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

Tech

This startup wants to hack the night sky

Jennifer Mossalgue • 5 minutes

Read Online | Sign Up | Advertise

Good morning, tech enthusiasts. A California startup wants to turn satellites into giant mirrors and aim them at Earth after dark, lighting up everything from construction sites to public events.

Reflect Orbital is seeking FCC approval to launch thousands of light-redirecting spacecraft to illuminate paying customers on the ground. But top scientists around the globe warn that “daylight on demand” shouldn’t be up for sale.

In today’s tech rundown:

  • This startup plans to light up the night

  • Apple’s foldable iPhone hits engineering snag

  • Netflix launches ad-free gaming app for kids

  • The smart glasses without ‘creepy’ vibes 

  • Quick hits on other tech news

LATEST DEVELOPMENTS

SPACE

🪩 This startup plans to light up the night

Image source: Reflect Orbital

The Rundown: California startup Reflect Orbital is seeking FCC approval to launch thousands of orbital mirrors that would redirect sunlight for paying customers on Earth after dark — and top scientists are sounding the alarm.

The details:

  • Reflect Orbital, founded in 2021 with $35M in funding, is building satellites with large mirrors designed to redirect sunlight onto the Earth’s surface after dark.

  • Earendil-1, its demo satellite, would deploy 60-foot mirrors from a 625 km orbit to illuminate 5 km ground targets, with a target launch this year.

  • Presidents of four international scientific societies, representing 2,500 researchers in 30+ countries, have sent letters of concern to the FCC over this.

  • The company says it has received more than 260K service requests for uses including construction, public events, and a $1.25M Air Force contract.

Why it matters: Critics warn that a single company, with one federal agency’s approval, could reshape the night sky for everyone on Earth. With Reflect Orbital aiming for 50K satellites by 2035, scientists warn of “major adverse health consequences” for humans and massive disruption for hundreds of species.

APPLE

🍎 Apple’s foldable iPhone hits engineering snag

Image source: Lovart / The Rundown

The Rundown: Apple’s rumored foldable iPhone has hit unexpected hinge and display durability snags that could delay its planned 2026 debut, according to a report first broken by Nikkei Asia.

The details:

  • Apple’s first foldable iPhone has run into tougher‑than‑expected hinge and display issues in early test production, raising the risk for a delayed launch.

  • Suppliers have reportedly been warned that mass production and initial shipments may be pushed back if engineering fixes take longer.

  • Earlier reports said that the Cupertino giant plans to anchor a 2026 lineup around the foldable plus two iPhones with bigger screens.

  • Apple is experimenting with advanced hinge designs and new materials like liquid‑metal components to tame creasing and stress on the ultra‑thin glass.

Why it matters: Apple hopes its foldable iPhone will jolt a slowing premium smartphone market, but engineering snags show that even its famously controlled hardware machine can struggle when it tries to reinvent the form factor. Samsung Display has meanwhile locked in orders for up to 20M foldable OLED panels.

NETFLIX

🖍️ Netflix launches ad-free gaming app for kids

Image source: Netflix

The Rundown: Netflix is turning its kids’ tab into a training ground for the next generation of streamers with the launch of Netflix Playground, an ad‑free mobile gaming app for kids aged eight and under.

The details:

  • The standalone gaming app comes bundled with every Netflix subscription at no extra cost, but requires parental sign-in.

  • It launches first in the U.S., Canada, the UK, Australia, the Philippines, and New Zealand on iOS and Android, with global rollout scheduled for April 28.

  • Every title in Playground is playable offline, with parental controls in place and no ads, in‑app purchases, or additional fees of any kind.

  • At launch, the catalog centers on games based on familiar brands like Peppa Pig, Sesame Street, StoryBots, Dr. Seuss, and coloring or puzzle apps.

Why it matters: Playground is Netflix’s first real hit at Apple Arcade and Amazon Kids+, folding an ad‑free kids’ game bundle into a subscription that rivals still upcharge for. If it hooks young kids on playing inside the same franchises they watch, Netflix tightens its grip on family time.

EVEN REALITIES

👓 Meet the smart glasses without ‘creepy’ vibes

Image source: Even Realities

The Rundown: Chinese upstart Even Realities is taking aim at Meta’s Ray‑Ban smart glasses by selling camera‑free specs that promise all the AI assistance with none of the “creepy lens on your face” surveillance vibes, the Financial Times reports.

The details:

  • Even’s $600 G2 glasses skip the front‑facing camera entirely, using a mic and a floating 3D heads-up display for email, maps, and real-time translation.

  • The company positions the glasses as a direct foil to Meta’s Ray-Bans, arguing most people don’t want “a camera on face” in everyday use, even if creators do.

  • It also launched Even Hub, an app store that turns the G2 into an open platform, with 50+ third‑party apps and an SDK used by 2K developers.

  • Meta is meanwhile working to scale its AI glasses production toward 20M pairs a year by 2026 while packing them with camera‑driven Meta AI features.

Why it matters: Smart glasses are having a genuine breakout moment — Meta’s Ray-Ban, Chinese rival Rokid, and a wave of Android XR devices are all competing to build the next iPhone, but one you’d wear. Even is making a different bet: that most people want a quiet AI assistant on their face, not a surveillance device.

QUICK HITS

📰 Everything else in tech today

NASA’s Artemis II crew flew the Orion to 252,700 miles from Earth, setting a new record for the farthest distance humans have ever traveled from the planet.

Elon Musk is reportedly requiring banks and other advisers working on SpaceX’s planned IPO to purchase subscriptions to his Grok AI chatbot service.

Oracle reportedly laid off 30K employees by email as part of a cost-cutting push despite reporting a 95% surge in profit and heavy investment in an AI data center.

Maine is poised to become the first U.S. state to temporarily ban the construction of large new data centers to study their environmental and power-grid impacts.

Amazon and the US Postal Service reached a deal that will cut Amazon’s USPS package deliveries by 20%.

High gas prices in the U.S. are making EVs more attractive again, potentially helping Tesla reverse its recent sales slump, Axios reported.

Apple is again asking the U.S. Supreme Court to review its App Store fight with Epic Games over commission limits.

Scientists built a tiny implant that keeps drug‑producing cells alive for weeks to deliver controlled treatments inside the body.

Satellite startup Impulse Space is partnering with Anduril to develop experimental space-based interceptor tech for Trump’s planned Golden Dome missile defense shield.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Sam Altman's new 'social contract' for AI

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Sam Altman wants the U.S. government to tax robots, create a national wealth fund, implement a 4-day workweek, and start planning for AI that can't be shut off. He also wants you to know this is urgent.

OpenAI's 13-page policy document lays out ideas behind the "new social contract," which Altman says is needed for a world adjusting to superintelligence — a transition he says has already begun.

In today’s AI rundown:

  • OpenAI’s new ‘social contract’ ideas for society, ASI

  • New Yorker surfaces secret memos behind Altman's firing

  • Stress test business ideas with Perplexity

  • Wang's first Meta models getting ready to ship

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

OPENAI

📜 OpenAI’s new ‘social contract’ ideas for society, ASI

Image source: OpenAI

The Rundown: OpenAI just published a 13-page policy document with ideas to help society navigate superintelligence and its societal impacts, asking Washington to tax AI-driven profits, create a wealth fund, implement a 4-day workweek, and more.

The details:

  • The proposal said we are “beginning a transition toward superintelligence”, with Altman telling Axios the moment requires a new “social contract” for society.

  • The most aggressive idea: a sovereign-style fund seeded by AI firms that would pay dividends to every American, as Alaska does with oil revenue.

  • Other ideas include taxes on robot labor, a 4-day workweek, "Right to AI" access for all, and containment playbooks for rogue autonomous AI.

  • Axios called it "the most detailed blueprint any tech titan has ever published for how to tax, regulate, and redistribute wealth from the technology he's building."

Why it matters: The CEO of an $852B company is asking the U.S. to prepare for a future where his own tech breaks the economic system — and you don't make that pitch unless you believe it's actually coming. But with the way things are moving, coupled with the slow-moving gears of the government, the clock is ticking.

TOGETHER WITH YOU.COM

🤑 Get the most out of your AI investment

The Rundown: Successful AI transformation starts with deeply understanding your organization’s most critical use cases. This practical guide from You.com walks through a proven framework to identify, prioritize, and document high-value AI opportunities.

In this AI Use Case Discovery Guide, you’ll learn how to:

  • Map internal workflows and customer journeys to pinpoint where AI can drive measurable ROI

  • Ask the right questions when it comes to AI use cases

  • Align cross-functional teams and stakeholders for a unified, scalable approach

Get the Guide.

SAM ALTMAN

🔍 New Yorker surfaces memos behind Altman's firing

Image source: The New Yorker

The Rundown: The New Yorker published an investigation into Sam Altman, drawing on 100+ interviews, unseen memos from ex–chief scientist Ilya Sutskever, and notes from Dario Amodei — alleging a long-running pattern of deception at the top of OpenAI.

The details:

  • The reporting spans Altman's full career arc, including conflicts at his startup Loopt, Y Combinator partners trying to push him out, and the OAI board drama.

  • Sutskever's memos, built from 70 pages of Slack messages and HR docs, allege Altman misrepresented safety protocols to the board.

  • Amodei's private notes, kept for years, reach the same conclusion as Sutskever independently: "The problem with OpenAI is Sam himself.”

  • A Microsoft exec told the reporters there's "a small but real chance" Altman is "remembered as a Bernie Madoff, Sam Bankman-Fried-level scammer.”

Why it matters: While there is no ‘smoking gun’ in this piece, there is a vastly detailed and concerning pattern of deception that seems to span across Altman’s career. But for every detractor, you also have fiercely loyal supporters and coworkers – making the CEO of the nearly trillion-dollar AI giant one of the most polarizing figures in the world.

AI TRAINING

🤔 Stress test business ideas with Perplexity

The Rundown: In this guide, you will learn how to use Perplexity Deep Research to stress test any business idea. Save the prompt below once and rerun it on every idea you have to see what’s feasible to build.

Step-by-step:

  1. Open Perplexity and switch to Deep Research mode. This works on the free plan (5 queries/day) and is basically a hidden version of Perplexity Computer

  2. Paste this prompt with your idea in the chat, hit run, and walk away for 5 to 6 minutes. Perplexity does the research and builds the slide deck in the same run

  3. Save the prompt somewhere you will actually use it again, like in a dedicated Perplexity space

  4. Then, every Saturday morning, take one idea off your list and run it. You will burn through a year of half-evaluated ideas in a month

Pro tip: Build variants. A 6-slide version for a co-founder pitch, a version that compares two ideas, or a 90-day MVP plan for ideas that already cleared validation.

PRESENTED BY IBM

⛳ IBM helps the Masters bring the fairway to every fan

The Rundown: There’s so much more to watching sports than watching. When every shot, by every player, on every hole produces more than 30 data points, the Masters and IBM turn golf into a data-rich, AI-powered immersive experience.

In the series, you'll explore:

  • Shots matter: Enhanced insights tell you why

  • Landing forecast: Score outcomes predicted by data and AI models

  • Round in 3 Minutes: Recap reels produced minutes after play

Get the full story.

META

 🚀 Wang's first Meta models getting ready to ship

Image source: Lovart / The Rundown

The Rundown: Meta is set to release the first AI models developed by Alexandr Wang’s Superintelligence team, with Axios reporting the company will make some of them available as open source — though the largest models will reportedly stay closed.

The details:

  • Meta and Wang’s codenamed ‘Avocado’ model was delayed in March over benchmark performances that fell short of rival models across the board.

  • The company is reportedly planning a consumer-focused, hybrid approach that includes both open and closed models for broad distribution across its apps.

  • Axios said Meta “knows its new models may not be competitive across the board” but “believes it will have areas of strength that appeal to consumers.”

Why it matters: This report doesn’t inspire confidence for a release that has already been delayed for poor performance, and after all the money spent and the high-profile new team brought in, another flop would be a painful one for a tech giant pushing desperately to enter the frontier AI race.

QUICK HITS

🛠️ Trending AI Tools

  • 🔒 Incogni - remove your personal data from the web so scammers and identity thieves can’t access it. Use code RUNDOWN to get 55% off.*

  • 🎥 VOID - Netflix’s open-source, physics-aware AI for video editing

  • 🗣️ AI Edge Eloquent - Google’s free voice dictation app that runs fully offline

  • 🎆 MAI-Image-2 - Microsoft’s image AI with upgraded realism and creativity

*Sponsored Listing

📰 Everything else in AI today

The Information reported that Sam Altman and OpenAI CFO Sarah Friar are not on the same page regarding IPO timing, though OpenAI denied the claim.

Iran's military singled out the $30B Stargate data center in Abu Dhabi as a target, publishing satellite footage and vowing to destroy U.S. infrastructure across the region.

OpenAI Head of Business Finance Chengpeng Mou posted new stats on ChatGPT’s use for healthcare questions, with the platform getting 2M insurance messages weekly.

Google released AI Edge Eloquent, a free iOS dictation app that cleans up raw speech into polished text entirely on-device.

Legion Health won approval to let its AI app directly refill psychiatric medications, the first time a state has greenlit AI to do the process without clinician oversight.

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 Wayne F. in the United Kingdom:

"I use AI as my "Workhorse" to act as a one-person eLearning agency. I host SME lessons in NotebookLM as a reference library. Using Kolb’s Learning Cycle, I map complex 5-part Pentad scenarios on a digital canvas, replacing physical rooms once filled with hundreds of post-its and flip charts.

I maintain strict control via a command-driven workflow, only generating assets when I trigger "VEO" or "create-image" using Nano Banana and Veo 3.1. Finally, I use Perplexity with NotebookLM to audit my work against the QM Rubric, Moore’s Theory of Transactional Distance, and Kolb’s cycle. AI allows me to collapse an entire production team into a single pipeline where I am simultaneously Creator, Designer, and Lead Auditor."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

See you soon,

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

Robotics

UBTech offers $18M a year for AI scientist

Jennifer Mossalgue • 5 minutes

Read Online | Sign Up | Advertise

Good morning, robotics enthusiasts. Chinese robotics firm UBTech is offering up to $18M a year for a single AI scientist to help power its humanoid ambitions.

In a talent market where Silicon Valley giants have reportedly put $20M annual pay packages — and even $100M-plus incentives — on the table for elite AI researchers, the price of winning is starting to look absurd. Is this hype, or is the humanoid race getting very real?

In today’s robotics rundown:

  • UBTech offers $18M a year for one AI scientist

  • This tiny bot grows its own nervous system

  • Japan’s new workforce: robots wanted

  • New gig economy teaches humanoids how to work

  • Quick hits on other robotics news

LATEST DEVELOPMENTS

UBTECH

🤖 UBTech offers $18M a year for one AI scientist

Image source: UBTech

The Rundown: Chinese humanoid maker UBTech is offering up to $18M annually for a single chief AI scientist, turning one job listing into a neon sign for just how extreme the global humanoid race has become.

The details:

  • The role will lead “embodied intelligence” research, translating VLA and robotics models into dependable software for full-size industrial humanoids.

  • In January, Airbus deployed UBTech’s Walker S2 robots on aircraft production lines, proving these machines can handle real factory floors, not just demos.

  • The company says full-size humanoid revenue has surged, with sales climbing more than 50% and claiming a growing slice of total income.

  • By dangling CEO-level pay, UBTech looks to be turning a single AI scientist role into a spectacle meant to signal dominance while attracting talent.

Why it matters: UBTech is racing against OpenAI, Google DeepMind, Tesla’s humanoid push, and fast‑moving rivals like Geekplus-W — backed by Beijing’s 2026 robotics priority and Geekplus-W’s February hire of Tsinghua’s Dr. Zhao Hao — making its blockbuster talent grab a matter of survival.

BIOTECH

🔬 This tiny bot grows its own nervous system

Image source: Haleh Fotowat

The Rundown: Researchers at Tufts and Harvard just created “neurobots,” microscopic living machines built from frog cells that develop their own nervous systems, reorganizing their bodies and behavior in ways evolution never produced.

The details:

  • These engineered neurobots are tiny living robots assembled from frog cells that autonomously grow rudimentary nervous systems from scratch.

  • As neurons spread through each bot, they wire directly into the outer cell layer and begin influencing how the organism moves.

  • Unlike earlier frog-cell bots, neurobots swim with greater intensity and display varied, unpredictable movement patterns rather than looping the same motion.

  • Once the nervous system emerges, gene activity shifts — switching on pathways associated with brain formation and even eye development.

Why it matters: Neurobots demonstrate that a cluster of frog cells can self-organize not only a functional body but a nervous system capable of reshaping that body’s behavior. The implications are early-stage but could hint at a coming class of engineered life that thinks and repairs like tissue but deploys like hardware.

JAPAN

💼 Japan’s new workforce: robots wanted

Image source: Ideogram / The Rundown

The Rundown: With a working-age population that makes up just 59.6% of the country’s total, Japan is reportedly running out of humans to do the work. Robots are filling roles no one is left to take, Techcrunch reports.

The details:

  • Japan faces such an extreme labor shortage that companies are rolling out robots not to cut jobs, but because there are no workers left to hire.

  • Robots are filling frontline roles in convenience stores, logistics, and hospitality, from stocking shelves and cleaning floors to delivering room service.

  • Elderly care facilities are adopting robotic assistants to lift patients, monitor vital signs, and provide companionship in a rapidly aging society.

  • Policymakers and business leaders are reframing automation as critical economic infrastructure rather than a threat to employment.

Why it matters: Japan is becoming a real-time lab for how advanced economies can survive when demographics make full employment mathematically impossible, with nearly a third of citizens already over 65. How it balances robots, regulation, and human dignity could set expectations for aging countries everywhere.

HUMANOIDS

🤖 New gig economy teaches humanoids how to work

Image source: Micro1

The Rundown: Around the world, gig workers are strapping cameras to their chests and filming themselves doing dishes, folding laundry, and stocking shelves for roughly $15 an hour — all to train the robots that may one day replace them.

The details:

  • Palo Alto startup Micro1 has recruited thousands of "robotics generalists" across 50+ countries to film tasks for humanoid training datasets.

  • Encord and Scale AI run robotics programs that pay people to record manipulation tasks, claiming up to 100K hours of robot-training video.

  • DoorDash has also paid drivers to film themselves doing chores, turning its workforce into an embodied-AI data pipeline.

  • Stealth robotics startups are reportedly posting Craigslist-style ads offering $10–$20 an hour for people to record everyday tasks on their phones.

Why it matters: Training humanoids on human movement is the robotics equivalent of what scraped text did for LLMs: source the raw material cheaply, capture the value at the top. Meanwhile, workers in Nigeria and India filming themselves folding laundry are training machines that could one day automate the very jobs they rely on.

QUICK HITS

📰 Everything else in robotics today

Rivian spinoff, Also, raised $200M at a $1B valuation and partnered with DoorDash to build autonomous last‑mile delivery vehicles.

Waymo began offering robotaxi pickups and drop-offs at San Antonio International Airport — its first airport service in Texas and fourth major airport overall.

Uber and WeRide began fully driverless, fare‑charging robotaxi operations in several districts of Dubai via the Uber app.

San Francisco startup NomadicML raised $8.4M at a $50M valuation to grow its platform that converts autonomous‑vehicle video into training data for fleet monitoring.

Researchers developed pasta-shaped, air-powered artificial muscles that let robots lift up to 100x their own weight.

A fleet of Maximo construction robots installed 100 megawatts of utility-scale solar capacity, marking one of the largest robotic solar deployments to date.

Robot “police dogs” are patrolling Atlanta streets to deter crime, sparking debate over surveillance, civil liberties, and the privatization of law-enforcement tech.

A homemade solar-powered quadcopter set an unofficial multirotor endurance record by flying for over five hours using 28 solar panels and a small backup battery.

Georgia Tech researchers built tiny vibration-powered robot swarms that can latch, release, and reorganize without any electronics or batteries.

Cornell researchers developed MirrorBot, a small robot with dual mirrors that uses shared reflections and eye contact to spark conversations and social connections.

U.S. researchers created a new silicone actuator that stays resilient in extreme cold, heat, and near-vacuum, showing promise for soft robots in space missions.

Chinese researchers developed a new soft optical sensor that lets a robot hand sense finger positions precisely enough to perform delicate, human-like tasks.

U.S.-based entertainment startup Dollhouse unveiled Belmont, a cute robot butler that can be rented to roam parties serving guests snacks and drinks.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Anthropic tells OpenClaw users to pay up

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. First came tighter rate limits. Then came the backlash. Now Anthropic is going a step further, cutting off third-party agent platforms like OpenClaw from Claude's subscription plans entirely.

It's a pricing correction the company says is about sustainability, but the timing couldn't be worse — with OpenAI aggressively courting the same developer audience Anthropic has built its reputation on.

In today’s AI rundown:

  • Anthropic boots third-party agents from Claude plans

  • The Rundown Roundtable: Our AI use cases

  • How to take AI notes on phone calls

  • Netflix opens physics-aware AI for video editing

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

🦞 Anthropic boots third-party agents from Claude plans

Image source: Lovart / The Rundown

The Rundown: Anthropic just blocked agent platforms like OpenClaw from running on Claude plans, requiring users to pay separately via usage add-ons or API keys, as the company confronts agent-driven demand its flat-rate pricing was never built to absorb.

The details:

  • Agent tools hit Claude with nonstop requests that exceed what its normal plans typically cover, despite Anthropic models being the leading driver for the tech.

  • Anthropic’s Boris Cherny announced the change, saying it is a step towards “managing growth to continue to serve our customers sustainably long-term”.

  • Anthropic is handing out credits worth a month's subscription, discounting add-ons up to 30%, and offering refunds amid cancellation requests.

  • OpenClaw creator Peter Steinberger criticized the step, saying, “First they copy popular features into their closed harness, then they lock out open source."

Why it matters: Anthropic was already catching heat over tighter rate limits, and walling off its agentic power-user community won't help the goodwill problem. It’s a tough situation with Anthropic’s agent usage likely playing a role in degrading normal user experience, but OAI is now there as the alternative at a crucial time in the rivalry.

TOGETHER WITH UNWRAP

⚡ See how Perplexity automates customer feedback

The Rundown: Unwrap is a customer intelligence platform that pulls your feedback – surveys, reviews, support tickets, social comments – into one view, using AI and NLP to surface actionable insights and deliver them straight to your inbox.

With Unwrap, you get:

  • All customer feedback automatically categorized

  • Query feedback using Unwrap Assistant, or in your favorite tools using Unwrap's MCP

  • Real-time alerts from feedback as they arise

  • A clear view of customer sentiment

Unwrap is offering a trial of its tools to Rundown AI subscribers, grab time with the team to get set up.

THE RUNDOWN ROUNDTABLE

💡The Rundown Roundtable: Our AI use cases

Image source: Ideogram / The Rundown

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.

Joey, Head of Partnerships: I have set up Claude to give me a daily outfit recommendation based on the local weather and the type of brands that fill 70% of my closet. I've provided examples and feedback, so suggestions become more and more accurate to my style and what I would normally wear without needing to think about it.

Nate, University Educator: Opposite of high school science class, I love that I never have to worry about being embarrassed to ask Claude when I don't understand something. The icing on the cake is to have it use the new chat visualizations feature.

After watching (the wonderful) Hail Mary Project Movie with my kids, I needed some help trying to explain the science of Time Dilation. It helped me get a little closer.

Billy, University Educator: I used AI (Perplexity Computer) to buy jeans for myself because I hate online shopping. It hunted for deals on the brands I like and even double checked each link was in stock for my size. Great success.

AI TRAINING

📞 How to take AI notes on phone calls

The Rundown: In this guide, you will learn how to set up an AI notetaker that works with any call on your iPhone. You probably know about AI notetakers for web, but most people don't know that you can also do it on both outbound and inbound phone calls.

Note: Check the recording consent laws for your state before recording your calls.

Step-by-step:

  1. Install Granola from the App Store, open it, tap the phone icon in the bottom left to set up. Enter your number, then follow the verification instructions

  2. To make a call, tap the phone button, pick a contact/type a number, and call. The call works like a normal one, but Granola is listening in the background

  3. After hanging up, wait a minute. Granola will give a summary with action items and anything worth remembering. No need to stay on screen during processing

  4. If you want to use Granola for an inbound call, you can open the app and create a new note. It will only be able to transcribe your voice though

Pro tip: Name notes in a format like [name] @ [company], create folders for work and personal calls. You can also connect Notion, Zapier, Slack, HubSpot, or your CRM.

PRESENTED BY GOOGLE CLOUD

⚡️ Ship your AI agent before your competitors do

The Rundown: The AI race is won by shipping fast and scaling seamlessly. Google Cloud’s Startup technical guide cuts through the noise, giving technical founders pre-built frameworks to design, build, and deploy intelligent autonomous systems in record time.

Inside the updated guide, you’ll discover:

  • Pre-built frameworks to accelerate autonomous agent design

  • Streamlined prompt engineering workflows to save developer hours

  • Frictionless deployment strategies using Google Cloud infrastructure

Get the guide.

AI RESEARCH

🎬 Netflix opens physics-aware AI for video editing

Image source: Netflix Research

The Rundown: Netflix just released VOID, an open-source framework built to erase video objects while rewriting the physics associated with them, instead of typical erasing and inpainting tools.

The details:

  • Existing removal tools just paint over backgrounds, without actually reasoning about the cause-and-effect those edits introduce across the broader scene.

  • VOID uses a mask that maps what to erase, what's physically affected, and what to keep, with a judge model then charting the consequences.

  • VOID can handles physics it never trained on, with demos like a balloon floating when a holder is removed or blocks not falling when one in the chain is erased.

  • 25 evaluators compared VOID against six baseline models including Runway, preferring Netflix’s results nearly 2/3 of the time.

Why it matters: This is Netflix Research’s first public AI release, and its a sign of where the video space is heading — intuitive systems that don’t just erase objects in footage like an image editor, but can actually simulate and alter the physics of the scene based on the changes for more controllability and real production use.

QUICK HITS

🛠️ Trending AI Tools

  •  🔀 Merge Gateway - Control plane for production AI: routing, cost, and reliability in one API*

  • ⚙️ Cursor 3 - Cursor's agent-first interface for parallel coding agents

  • 🎥 PikaStream 1.0 - Pika's video chat AI that gives any AI agent a face, voice

  • 💎 Gemma 4 - Google's new open-weight AI with SOTA intelligence for its size

*Sponsored Listing

📰 Everything else in AI today

OpenAI is navigating a leadership change, with Fidji Simo on medical leave, COO Brad Lightcap on special projects, and CMO Kate Rouch stepping down for cancer recovery.

Anthropic acquired startup Coefficient Bio for roughly $400M, folding the team into its healthcare and life sciences group focused on drug discovery.

Mercor confirmed a data breach tied to an attack on open-source library LiteLLM, with hackers claiming access to up to 4 TB of data from the $10B AI training startup.

Pika Labs released PikaStream 1.0 in beta, a real-time model that lets AI agents join Google Meet calls as video avatars with voice cloning and live conversation.

OpenAI rolled out ChatGPT in CarPlay, allowing users to access Voice Mode in their supported vehicle for hands-free use.

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 Kevin R. in Terre Haute, IN:

"My main passion is building and modifying RC cars. I've been in the hobby for 40+ years. I am also a moderator on one of the longest running RC forums.

A radio system came out from Radiomaster with a very powerful open source firmware by EdgeTX. This radio has a steep learning curve, so the surface based RC guys and gals are not too keen to learn it. So I decided to build a wizard to program various car models, but I am definitely not a coder.

I have been using Claude, Gemini, and Grok to help me. When Claude generated the first working version of the app, it blew me away. Finally, I could bring my ideas to life. Now I am completely hooked, and learning so much. Your email sub has been a HUGE help! So thank you!"

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

See you soon,

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

Tech

Google's Texas-sized data center problem

Jennifer Mossalgue • 6 minutes

Read Online | Sign Up | Advertise

Good morning, tech enthusiasts. Google is reportedly tying its AI expansion in Texas to a giant gas plant that could belch 4.5M tons of CO₂ a year — more than some U.S. cities.

That’s a jarring turn for the company that spent years evangelizing 24/7 carbon-free energy, especially since this plant appears to have no carbon capture at all. But Google, like the rest of Big Tech, says that AI’s appetite for power is growing faster than the clean grid can supply it.

In today’s tech rundown:

  • Google to power Texas AI data center on gas

  • Artemis II astronauts head to the moon

  • Amazon is coming for Walmart — with robots

  • Whoop is now a $10B fitness tracker

  • Quick hits on other tech news

LATEST DEVELOPMENTS

GOOGLE

😷 Google to power Texas data center on gas

Image source: Ideogram / The Rundown

The Rundown: Google reportedly plans to power a new AI data center in Texas with a gas power plant that could emit about 4.5M tons of CO₂ a year, in what critics say is a major rollback from its earlier 2030 carbon‑free energy goals.

The details:

  • Google confirmed it is partnering with Crusoe on the Goodnight data center campus in Texas, where Crusoe has filed for a 933 MW gas plant.

  • The data center could cost nearly $30B, and the gas plant could emit roughly 4.5M tons of CO₂ annually — more yearly emissions than San Francisco.

  • Unlike Google’s recent gas deal in Illinois, the Goodnight plant reportedly has no carbon capture technology whatsoever.

  • Google confirmed the partnership but says no offtake agreement for the gas plant has been signed.

Why it matters: Google built its brand on climate leadership — it pioneered 24/7 carbon-free energy and has signed more than 22 gigawatts of clean energy power purchase agreements. A bare-gas, no-capture plant of this scale is a different animal entirely, but Google says surging AI demand is outpacing the clean energy buildout.

TOGETHER WITH KESTRA

⚒️ Fix your broken automation stack

The Rundown: Kestra is an open-source workflow orchestration platform that replaces scattered scripts and cron jobs with one unified layer. With 26k+ GitHub stars, it’s already trusted by Apple, JPMorgan Chase, Toyota, and BHP for mission-critical workflows.

Do more with Kestra:

  • Orchestrate any language, tool, or service

  • Deploy on-prem, hybrid, or any cloud

  • Build from the UI or code, fully in sync

Explore Kestra.

NASA

🌝 Artemis II astronauts head to the moon

Image source: NASA

The Rundown: NASA’s Artemis II mission has just launched four astronauts into a looping flyby of the Moon, rebooting crewed deep‑space exploration more than half a century after Apollo’s final flight.

The details:

  • The Space Launch System mega-rocket lifted off on April 1, sending four astronauts aboard Orion on a 10-day test flight around the Moon and back.

  • A roughly six-minute translunar injection burn broke the crew free of Earth orbit — the first time humans have departed Earth orbit since Apollo 17 in 1972.

  • Updated trajectory data puts the crew’s maximum distance from Earth at 252,021 miles — surpassing the Apollo 13 distance record by 3,366 miles.

  • The lunar flyby is scheduled for Monday, April 6, when the crew will photograph areas of the far side never directly seen by human eyes.

Why it matters: What’s learned on this flight is critical to future Artemis missions — NASA is targeting Artemis III for lunar technology demonstrations in 2027 and a crewed surface landing with Artemis IV in 2028. Every telemetry point from this mission could write the rulebook for what comes next. 

AMAZON

🛒 Amazon is coming for Walmart — with robots

Image source: Reve AI / The Rundown

The Rundown: Amazon has been developing a network of massive, robot-heavy hybrid supercenters under an internal initiative called Project Kobe — and leaked documents obtained by Business Insider reveal just how serious the bet is.

The details:

  • Each store would clock in at 225K square feet, with nearly half the floor plan dedicated to back-of-house robotics and fulfillment infrastructure.

  • AutoStore robotic systems handle warehouse operations; a future in-house platform called Orbital is also in the pipeline.

  • An AI tool named Frida is designed to help category managers automate inventory decisions at the local level.

  • The first approved site is in Orland Park, Illinois (late 2027 opening), with additional locations in New Jersey and Illinois on the table.

Why it matters: Amazon and Whole Foods hold just 3% of the U.S. grocery market versus Walmart’s 21% — and Project Kobe is Amazon’s most ambitious attempt yet to close that gap, by collapsing the e-commerce fulfillment center and the big-box store into one. If the pilots work, Amazon is prepared to roll the format out at scale.

WHOOP

🏋🏽‍♂️ Whoop is now a $10B fitness tracker

Image source: Whoop

The Rundown: Whoop just closed a massive $575M Series G round that nearly triples its valuation to $10.1B — a sign the market is backing its pivot from elite fitness tracker to full-blown health platform.

The details:

  • Diagnostic device maker Abbott and Mayo Clinic joined as strategic investors, alongside athletes Cristiano Ronaldo, LeBron James, and Rory McIlroy.

  • The platform already includes FDA-cleared ECG, blood pressure insights, and Advanced Labs blood biomarker analysis, with Whoop promising “more to come.”

  • Abbott’s move mirrors Dexcom’s 2024 investment in Oura’s smart ring, a pattern of medtech players buying strategic footholds in consumer biometric platforms.

  • Whoop now counts 2.5M members and exited 2025 with a $1.1B annualized bookings run rate, up 103% year over year.

Why it matters: Whoop’s new backers aren't typical venture money — Abbott makes diagnostic devices, Mayo Clinic runs hospitals. Whether that translates into actual regulated products or just credibility remains to be seen. The FDA’s 2025 warning letter to Whoop over its blood pressure claims is also a hint of the hurdles ahead.

QUICK HITS

📰 Everything else in tech today

Amazon is in talks to acquire satellite telecoms group Globalstar in a deal worth about $9B, aiming to build a rival to Elon Musk’s Starlink satellite internet network.

Microsoft will invest about $10B in Japan from 2026 to 2029 to expand AI infrastructure, train 1M tech workers, and deepen cybersecurity cooperation.

Google is now rolling out a feature that lets users in the U.S. change their existing Gmail address without creating a new account.

SpaceX confirmed that one of its Starlink satellites suffered an unexplained anomaly that caused it to break apart into debris fragments in low Earth orbit.

London-based hardware company Nothing is reportedly planning to launch AI-powered smart glasses in 2027 and AI earbuds in 2026.

Amazon will start adding a 3.5% “fuel and logistics” surcharge to fulfillment fees it charges many third‑party sellers, blaming higher fuel costs linked to the Iran war.

China’s cyberspace regulator issued draft rules to tightly control “digital humans,” requiring clear labeling and banning features that could addict children.

The New York Times dropped freelance critic Alex Preston after he admitted using an AI tool that inserted language from a Guardian review into his own book review.

Alexa+ subscribers with Echo Show displays can now link their Uber Eats or Grubhub accounts and order delivery in a natural, back‑and‑forth conversation.

Lucid Motors is recalling more than 4K Gravity SUVs because a supplier improperly welded some of their second-row seat belt anchors.

Chinese researchers developed a new electrolyte for lithium batteries that more than doubles energy density and EV range while still working reliably in extreme cold.

Renewable power made up nearly half of the world’s electricity capacity in 2025, reaching 49.4% after solar additions drove renewable capacity to 5,149 GW.

Newsletter platform Beehiiv is launching podcast hosting so creators can produce podcasts alongside newsletters on one platform, to challenge Patreon and Substack.

China’s CAS Space successfully launched its new Kinetica‑2 Y1 rocket at a cost reportedly comparable to SpaceX’s Falcon 9.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

No matching search results

Try using different keywords, double-check your spelling, or explore related categories.

Clear Search

Stay Ahead on AI.

Join 2,000,000+ readers getting bite-size AI news updates straight to their inbox every morning with The Rundown AI newsletter. It's 100% free.

Subscribe