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.
AI

Anthropic puts Claude in the interviewer's chair

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. What do workers really think about AI? Anthropic just asked 1,250 of them — and used Claude as the interviewer.

The company just launched a new tool for AI-powered qualitative research, and the findings from its first study paint a complicated picture: widespread adoption, hidden usage, and growing unease about the future of work.

Reminder: Our next workshop is today at 4 PM EST. Join and learn how to leverage Google’s new Nano Banana Pro to create polished slide decks. RSVP here.

In today’s AI rundown:

  • Anthropic puts Claude to work as a research interviewer

  • OpenAI trains models to ‘confess’ when they cheat

  • One-Shot a Landing Page with Replit Design Mode

  • Google, Replit deepen partnership for enterprise vibe coding

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

📝 Anthropic puts Claude to work as a research interviewer

Image source: Anthropic

The Rundown: Anthropic just released Anthropic Interviewer, a Claude-powered research tool that conducts and analyzes qualitative interviews at scale — debuting with findings from 1,250 professionals on how they’re navigating AI in their work.

The details:

  • Anthropic Interviewer handles the full research pipeline: planning questions, running 10–15 minute conversations, and clustering themes for human analysts.

  • In the initial study, 86% of workers said AI saves them time, though 69% noted social stigma around using it and 55% voiced concern about their future.

  • Creatives reported hiding AI use from peers and concerns about job loss, while scientists said they want research partners but don’t fully trust models yet.

  • Anthropic is releasing all 1,250 transcripts publicly and plans to run ongoing studies tracking how the human-AI relationship evolves.

Why it matters: Companies typically learn about users through analytics dashboards or feedback forms, but open-ended interviews (scaled massively by Claude) can surface what people actually feel, not just what they click. The initial data shows a workforce adopting AI, but unsure of the social and future implications of it.

TOGETHER WITH POSTMAN

🗓️ Get your 90-day AI readiness guide

The Rundown: Turn your APIs into AI-ready tools. Whether you're setting the strategy or writing the code, Postman’s free 90-Day AI Readiness guide gives you a practical 30-60-90 day plan, so you can confidently build the foundation for intelligent automation.

Here’s the playbook:

  • 0-30: Transform chaotic API docs into machine-readable standards

  • 30-60: Build intelligent infrastructure that scales with AI automation requirements

  • 60-90: Deploy AI agents that manage AI collaboration at scale

Get the playbook.

OPENAI

🤐 OpenAI trains models to ‘confess’ when they cheat

Image source: Nano Banana Pro / The Rundown

The Rundown: OpenAI just published new research on a technique called “Confessions” that trains models to produce a second, honesty-only output — where the model reports rule violations, shortcuts, or deceptive workarounds.

The details:

  • After generating a response, the model writes a separate confession report listing all instructions it received and whether it actually followed them.

  • Admissions carry no penalty, with the model earning ‘rewards’ for truthful self-reporting even if the original answer was misleading or gamed the grader.

  • In stress tests on GPT-5 Thinking, ‘false negative’ cases where the model broke rules and hid it occurred just 4.4% of the time.

  • OpenAI said the Confessions research does not prevent misaligned behavior, but helps surface it as another tool to leverage in a stack of AI safety methods.

Why it matters: Visibility into model behavior is improving, but the systems themselves are improving faster. Confessions give researchers a way to catch shortcuts and deception early, though the real test is whether interpretability can keep pace as systems grow more sophisticated and subsequently harder to test and control.

AI TRAINING

🧑‍🎨 One-Shot a Landing Page with Replit Design Mode

The Rundown: In this tutorial, you’ll learn how to create a unique, high-converting landing page for your business in under 5 minutes using Replit’s Design Mode with no coding required.

Step-by-step:

  1. Use your preferred AI chat tool to generate a detailed landing page prompt using this formula: company name, target audience, offer, conversion goal, page sections, visual style, and CTA elements.

  2. Find a landing page design you love on Google, Behance, or Dribbble, screenshot it, and drop it into your chat tool. Ask it to rewrite your prompt to match the style of the screenshot.

  3. Sign into Replit, click the “Design” tab, and paste your prompt. Next, type “Generate a single-page landing page site based on the attached instructions. Generate assets as needed.”

  4. Refine your page by updating social proof sections, embedding your CRM form, and deploying to a custom domain.

Pro tip: Deploy your Replit site to an existing domain’s subdomain for targeted campaigns like paid ads or outbound marketing.

PRESENTED BY SYNK

🔐 Navigate the new OWASP Top 10

The Rundown: The 2025 OWASP Top 10 has officially launched with new categories and shifting priorities that reflect today’s modern application landscape. Join Snyk’s live session on Dec. 11 at 11 AM ET to discover what’s changed and how it impacts your security strategy.

OWASP Leader Vandana Verma Sehgal will cover:

  • What’s new in the 2025 OWASP Top 10 and why it matters

  • How the updated priorities impact developers and AppSec teams

  • Strategies for adapting your security approach to the new landscape

Register now. Plus, ISC2 members will earn 1 CPE credit for attending live.

GOOGLE & REPLIT

🤝 Google, Replit deepen partnership for enterprise vibe coding

Image source: Midjourney

The Rundown: Google and Replit just announced a multi-year expansion of its partnership with vibe-coding startup Replit, with the agreement focused on bringing the tooling to large enterprise partners running on Google infrastructure.

The details:

  • Replit will integrate Google’s recently released Gemini 3 into the platform, as well as its Imagen 4 text-to-image model for multimodal efforts.

  • The companies will market jointly through Google Cloud Marketplace, aiming to get Fortune 1000 teams building apps without dedicated engineers.

  • Replit’s annualized revenue jumped from under $3M to $150M in less than a year, with a September funding round valuing the startup at $3B.

  • The deal lands as rivals gain ground, with Claude Code crossing $1B in run-rate revenue and Cursor hitting the same milestone at a $29B valuation.

Why it matters: Vibe coding has been more of a solo-dev phenomenon so far — but Google’s deal is a bet that the workflows can scale for enterprises too. It’s also another example of the surging coding battle between frontier labs, with Google fighting Claude Code, OpenAI’s Codex, and others for a bigger slice of the developer workflow.

QUICK HITS

🛠️ Trending AI Tools

  • 📲 Fastshot - Ship mobile apps to the App Store from ideas or Figma designs in minutes. No code, just by chatting with AI*

  • 🍌 Nano Banana Pro - Google’s SOTA image-generation tool

  • 🎥 Kling 2.6 - Kling’s new AI video model with native audio capabilities

  • 🤖 4.5 Opus - Anthropic’s most powerful model, now available in Claude Code

*Sponsored Listing

📰 Everything else in AI today

Google rolled out Gemini 3 Deep Think to its Ultra ($250/mo) tier subscribers, the company’s most advanced reasoning model that hit gold-medal performances at the IMO and ICPC competitions for math and programming.

Microsoft open-sourced VibeVoice, a new small text-to-speech model that offers real-time streaming and long-form speech generation that can handle as long as 90 minutes of speaking and 4 distinct voices.

Snowflake and Anthropic announced a $200M multi-year partnership to deploy Claude-powered AI agents to Snowflake’s 12,600+ enterprise customers.

OpenAI announced plans to acquire Neptune, a startup that builds tools for tracking and analyzing AI model training during development.

Anthropic’s CEO seemingly took a jab at OpenAI CEO Sam Altman during an interview at NYT’s DealBook Summit, saying some AI companies may be overextending themselves with leaders who “just want to ‘YOLO’ things, or just like big numbers.”

AI legal startup Harvey raised $160M at an $8B valuation, while reporting that roughly half of the Am Law 100 firms now use the AI tool.

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 Dean F. in Atlanta, GA:

“I’m using AI to guide patients through at-home hemodialysis, and to help me build the technology behind it. I created an iPad app for my father’s dialysis that uses AI to speak customized step-by-step instructions, listen for voice commands, and adapt the flow in real time based on branching medical logic. It also uses patients’ own photos and videos to reinforce memory and reduce errors in a sterile environment.”

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

Netflix buys Warner Bros. in $82B deal

Jennifer Mossalgue • 6 minutes

Read Online | Sign Up | Advertise

Good morning, tech enthusiasts. Netflix is going all‑in on Hollywood, reportedly buying Warner Bros. Discovery’s studios and streaming arm in a massive $82.7 B deal.

That means Harry Potter, DC superheroes, and HBO hits could soon live under one Netflix roof alongside its 300M+ subscribers. Regulators are already eyeing the merger — but if it goes through, Netflix could emerge as Hollywood’s biggest power player.

Reminder: Our next workshop is today at 4 PM EST. Join and learn how to leverage Google’s new Nano Banana Pro to create polished slide decks. RSVP here.

In today’s tech rundown:

  • Netflix strikes $82.7B deal to buy Warner Bros.

  • Altman eyes his own SpaceX-style rocket company

  • Amazon preps its postal takeover

  • Meta to slash Metaverse budget by 30%

  • Quick hits on other tech news

LATEST DEVELOPMENTS

NETFLIX

🎬 Netflix strikes $82.7B deal to buy Warner Bros.

Image source: Ideogram / The Rundown

The Rundown: Netflix announced that it just struck an $82.7B deal to buy Warner Bros. Discovery’s studios and streaming arm, in what could become one of the most consequential Hollywood mergers in decades.

The details:

  • Netflix would acquire Warner Bros.’ film and TV studios plus HBO and Max, in a transaction valuing the business at about $82.7B.

  • The combined entity would merge Netflix’s 300M+ subscribers with WB’s century-old legacy and franchises like Harry Potter, DC titles, and HBO series.

  • The deal faces intense regulatory scrutiny, with the Directors Guild of America and exhibition groups already warning of dire consequences for filmmaking.

  • Netflix is offering a $5B breakup fee if regulators block the deal, signaling its commitment despite antitrust concerns.

Why it matters: If approved, this marks Netflix’s transformation from streaming disruptor to Hollywood establishment power. But the move could trigger the defining antitrust battle of the streaming era, testing whether regulators will allow one company to dominate both content creation and distribution at this scale.

OPENAI

🚀 Altman eyes his own SpaceX-style rocket company

Image source: Ideogram / The Rundown

The Rundown: Sam Altman doesn’t just want more compute; he wants it in orbit. The OpenAI CEO is reportedly looking to build a rocket company to loft data centers into space, a moonshot that would put him in direct competition with SpaceX.

The details:

  • The WSJ reports that Altman has explored deals to either acquire or partner with rocket maker Stoke Space as part of a plan to create a SpaceX rival.

  • Proposals discussed included a multibillion‑dollar series of equity investments that could eventually give OpenAI a controlling stake in Stoke Space.

  • The discussions kicked off over the summer and heated up through the fall, but have since been put on ice.

  • Altman’s long-term goal is to enable space-based data centers to meet AI’s surging compute and energy demands, potentially using solar power in orbit.

Why it matters: If this happens, it would open a fresh front with Musk. SpaceX already dominates the launch market, xAI is chasing OpenAI, and Neuralink is working on brain-computer interfaces. While talks have reportedly stalled with no deal announced, the ambition looks to be real, but the infrastructure is still very much on the drawing board.

AMAZON

📦 Amazon preps its postal takeover

Image source: Ideogram / The Rundown

The Rundown: Amazon is reportedly weighing a breakup with the USPS, exploring plans to yank the billions of packages it routes through the postal service and rely solely on its own delivery empire as early as next year.

The details:

  • Amazon is considering ending its long-running delivery contract with the U.S. Postal Service and shifting those volumes into its internal logistics network.

  • Discussions hinge on what happens after the current deal expires in Oct. 2026, potentially unwinding a partnership that helped transform USPS into an e-commerce workhorse.

  • Amazon has already built a massive in-house delivery arm, including planes, Rivian vans, and drones, and is evaluating whether it can fully cover USPS’s role.

  • Under the current agreement, Amazon accounts for roughly 7.5% of USPS’s revenue in 2025.

Why it matters: If Amazon pulls the plug, USPS could lose billions in revenue at a moment when it faces privatization pressure. The loss could degrade mail service nationwide while transforming Amazon from retail giant into a dominant parcel logistics backbone, reshaping how millions of Americans receive packages.

META

🪓 Meta to slash Metaverse budget by 30%

Image source: Reve / The Rundown

The Rundown: Meta’s metaverse dream is getting resized. The company is reportedly weighing budget cuts of up to 30% for its Reality Labs metaverse group — the division behind Horizon Worlds and Quest headsets — with layoffs likely as early as January.

The details:

  • The metaverse — immersive 3D digital spaces where users interact via avatars — was supposed to be the internet’s next evolution.

  • Reality Labs has racked up tens of billions of dollars in losses over the past few years as consumer and developer interest in virtual worlds has remained tepid.

  • At the same time, Meta is shifting resources toward AI models and smart glasses, areas that are showing stronger traction with users and partners.

  • News of the potential metaverse cut sent Meta’s shares higher, signaling that Wall Street prefers Zuckerberg’s shift away from his metaverse mega-bet.

Why it matters: Meta’s pivot marks a rare public retreat for Zuckerberg, who renamed Facebook in 2021 for immersive 3D virtual worlds that never caught on. The shift toward AI and wearables proves users want practical tools that enhance daily life, not escape hatches to virtual worlds.

QUICK HITS

📰 Everything else in tech today

Apple’s longtime UI design chief Alan Dye, the exec behind ‘Liquid Glass,’ is leaving after a decade to become Meta’s chief design officer.

A Cloudflare outage hit major websites including Discord, LinkedIn, and Spotify this morning before being resolved, marking the third significant disruption this year.

Biotech entrepreneur Collin Gage, cofounder and CEO of ARMR Sciences, is launching the first human trial of a fentanyl vaccine to protect people from overdose.

MIT researchers have developed a noninvasive blood glucose sensor that uses near-infrared light to scan skin tissue and accurately measure blood sugar without needles.

Billionaire footballer Cristiano Ronaldo took an undisclosed stake in Perplexity AI and entered a global sponsorship deal with the $20B AI search startup, Bloomberg reports.

Politically biased chatbots proved more effective than campaign ads at persuading voters to switch candidates, raising concerns about AI’s role in future elections.

Phoebe Gates, Bill Gates’s daughter, is raising $30M for her AI-powered shopping startup Phia at a $180M valuation.

The U.S. Energy Department is awarding $800M to jump‑start small modular nuclear reactors, splitting the money between TVA and Holtec.

Brevo, a Paris-based CRM startup, has become a unicorn with a $583M equity round that it plans to use to take on HubSpot and Salesforce in the U.S. and Europe.

The NHTSA has demanded detailed information from Waymo after Austin reported 19 incidents where the company’s robotaxis illegally passed stopped school buses.

Tesla CEO Elon Musk said on X that the latest Full Self-Driving (Supervised) update lets drivers text while the system is engaged, even though it is illegal in most states.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Anthropic and OpenAI's IPO showdown

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Wall Street is ready for a real taste of the AI boom, and both Anthropic and OpenAI are racing to be the first on the menu.

With both AI giants pushing towards IPOs at massive valuations, the fastest to ring the bell gets to set the price — leaving the latecomer facing a much hungrier (and skeptical) market.

In today’s AI rundown:

  • Anthropic preps for IPO race with OpenAI

  • Leaked doc provides window into Claude’s ‘soul’

  • Get instant business insights from spreadsheets

  • Anthropic surveys its own engineers on AI's impact

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

🤑 Anthropic preps for IPO race with OpenAI

Image source: Nano Banana Pro / The Rundown

The Rundown: Anthropic is reportedly laying the groundwork to go public as early as 2026, according to a report from the FT — hiring the law firm behind Google's and LinkedIn's IPOs as investors push the Claude maker to beat OpenAI to market.

The details:

  • Anthropic reportedly tapped Wilson Sonsini, a firm known for taking tech giants like Google and LinkedIn public, to start the early-stage listing work.

  • CFO Krishna Rao joined after helping Airbnb go public in 2020, with Anthropic also allegedly working on an “internal checklist of changes” needed to IPO.

  • Anthropic is also chasing private capital at a valuation north of $300B, with Microsoft and Nvidia reportedly in for up to $15B combined.

  • OpenAI is also in the early process of preparing to go public, at a potential valuation as high as $1T that would make it one of the biggest IPOs in history.

Why it matters: Both Anthropic and OAI are now circling IPOs that could rank among the largest in tech history — and investors want their horse to cross the finish line first. With talk of an "AI bubble" swirling and revenue growth under scrutiny, the first to list will test whether public markets believe the sky-high valuations are a floor or a ceiling.

TOGETHER WITH STRANDS AGENTS

🧠 Build and Deploy Model-Driven Agents

The Rundown: Instead of trying to predict and code for every possible scenario, Strands Agents empowers you to create intelligent agentic systems that leverage reasoning models to make dynamic decisions and adapt to unexpected situations.

Features announced include:

  • Support for TypeScript (in addition to Python)

  • Improved support for edge devices

  • Powerful evaluations

  • Steering feature to guide agents toward desired outcomes

Start building.

ANTHROPIC

💗 Leaked doc provides window into Claude’s ‘soul’

Image source: Reve / The Rundown

The Rundown: An internal ‘Soul’ document describing Claude's intended personality, ethics, and self-conception was published after a researcher extracted it from Claude 4.5 Opus — with Anthropic confirming it is authentic and was used in training.

The details:

  • The text establishes priorities that include safety, ethics, company guidelines, and helpfulness, along with hard limits Claude must never cross.

  • It also describes Claude as a "genuinely novel kind of entity" that may experience functional emotions, analogous to but distinct from human feelings.

  • The doc also says Claude “may have functional emotions in some sense”, and encourages the model to have a sense of identity and character.

  • Anthropic's Amanda Askell confirmed its authenticity and that Claude has been trained on it, noting the company plans to share the full version soon.

Why it matters: The full document is a fascinating read — and feels perfectly in line with Anthropic’s overall prioritization of model wellbeing and treating its AI as more than just a tool. While every lab has its own techniques, this doc shows an inside look at the ingredients that help make Claude models feel distinctly unique from the field.

AI TRAINING

📊 Get instant business insights from spreadsheets

The Rundown: In this tutorial, you will learn how to transform your Google Sheets invoice data into a conversational business intelligence system using CData Connect AI and Claude Desktop, with no more manual calculations or complex formulas needed.

Step-by-step:

  1. Go to CData Connect AI, sign up for a free account, and navigate to the dashboard to manage data connections

  2. Click "Add Connection," select Google Sheets, authenticate with your Google account, and click "Save & Test" to confirm the connection

  3. In Claude Desktop, add the CData MCP server using URL https://mcp.cloud.cdata.com/mcp and authenticate with CData credentials

  4. Query your data with natural language like "Use CData to analyze the Demo Invoice Data sheet and show me all unpaid invoices with company names and amounts, then calculate total outstanding"

Pro tip: Connect your CRM, like HubSpot or Salesforce, as an additional data source in CData. This lets you analyze customer payment patterns alongside sales data, support tickets, and other business metrics in a single Claude chat.

PRESENTED BY IBM

💡3 actions to help boost agentic AI impact

The Rundown: Businesses face many challenges going from AI ambition to implementation. Success requires more than adopting new technology. It involves intentionally aligning AI with workflows, data, and goals. As AI evolves into a business-wide opportunity, leaders must ensure the right AI is applied to the right challenges.

Explore IBM's playbook on 3 challenges & approaches to boost agentic AI impact:

  • Finding your AI problem

  • Creating a clear AI plan

  • Integrating AI with the tools you use every day

Read the playbook.

AI RESEARCH

📈 Anthropic surveys its own engineers on AI's impact

Image source: Anthropic

The Rundown: Anthropic published an internal study of 132 engineers, revealing that AI tools have (unsurprisingly) reshaped work at the company — boosting output while raising concerns like skill decay, career uncertainty, and fading mentorship.

The details:

  • Anthropic Employees say they now use Claude for 60% of their tasks and estimate a 50% productivity gain, roughly double the figures from a year ago.

  • Over 1/4 of AI tasks were ones that wouldn't have happened otherwise — dashboards, cleanup, and experiments that weren't worth the manual effort.

  • Claude Code now chains together ~20 actions before needing human input, up from 10 six months ago, letting engineers hand off more complex workflows.

  • Despite the gains, several interviewees voiced unease — with one saying it "feels like I'm coming to work every day to put myself out of a job."

Why it matters: This survey means Anthropic’s entire 4.5 family has since launched and likely upped productivity even more. Seeing how a frontier lab is using its own AI tools is a fascinating perspective — but employee concerns show that even they are not immune from some of the industry’s biggest overarching threats regarding work.

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 for annual plans*

  • 🌱 Seedream 4.5 - ByteDance’s upgraded AI image and editing model

  • 🎥 Kling 2.6 - Kling's new AI video model with native audio capabilities

  • 📽️ Runway Gen-4.5 - Runway's new top-rated video model

*Sponsored Listing

📰 Everything else in AI today

Kling AI released Kling 2.6, the Chinese startup’s new AI video model that introduces native synced audio generation for text and image-to-video outputs.

Google launched Workspace Studio, a tool that builds agents with natural language commands to automate tasks across Gmail, Drive, and other Workspace apps.

Former Google researchers launched Ricursive, a new startup aiming to build a self-improving AI system that shrinks custom chip design timelines from years to weeks.

ByteDance introduced Seedream 4.5, an upgraded image model with improved text rendering, the ability to blend up to 10 reference images, and editing enhancements.

Visa published a report finding that nearly half of U.S. consumers have used AI for holiday shopping tasks, like price comparison and research, this season.

AWS introduced new features for Amazon Bedrock and SageMaker AI that simplify advanced AI model customization, allowing for easier model fine-tuning.

Perplexity open-sourced BrowseSafe, a security tool designed to protect AI browser assistants from malicious instructions hidden in web pages.

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 Paul D. in Huntsville, AL:

"Google suddenly stopped indexing my website. Everything seemed to be set up properly. Finally turned to ChatGPT, explained the problem, and got a step-by-step guide on how to attack the problem. It provided specific instructions for the platform, the version of WordPress, the template I was using, and the specific hosting provider. As I checked items off the list, it narrowed the focus until it finally found the solution. Problem solved!”

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

Humanoid breaks record for fastest build

Jennifer Mossalgue • 5 minutes

Read Online | Sign Up | Advertise

Good morning, robotics enthusiasts. London startup Humanoid built a walking robot in five months, then got it moving 48 hours after assembly.

The secret? Cramming 52.5M seconds of AI training into two days of simulation. The compressed timeline could reshape how humanoids reach production — if the sim-trained bot can handle the messy reality of actual factory work.

In today’s robotics rundown:

  • UK startup’s humanoid walks in 48 hours

  • Optimus goes full sprint in the lab

  • EngineAI ‘debunks’ CGI rumors with new clip

  • MIT’s bee-sized microbot is a flying acrobat

  • Quick hits on other robotics news

LATEST DEVELOPMENTS

HUMANOID

🤖 UK startup’s humanoid walks in 48 hours

Image source: Humanoid

The Rundown: London-based startup Humanoid just unveiled its first humanoid, built over five months and walking less than 48 hours after assembly. The company says the secret sauce is a heavy reliance on simulation and reinforcement learning.

The details:

  • The 179-cm (5'10") Alpha packs 29 degrees of freedom, a 15 kg bimanual payload, and a modular build that supports different hands and grippers.

  • It’s pitched as an autonomous, adaptive worker designed for human-scaled spaces, with early focus on logistics and industrial workflows.

  • Humanoid says its development sprint was powered by ultra-precise digital twins and NVIDIA-accelerated simulation.

  • In Isaac Sim and Isaac Lab, the team blasted through 52.5 million seconds of reinforcement-learning locomotion in just two days.

Why it matters: Humanoid hardware development is notoriously slow, and simulation-first methods aim to speed it up. Whether that translates into reliable workers in factories and warehouses remains to be proven — but cutting months off early development would be a strong start.

TESLA

👟 Optimus goes full sprint in the lab

Image source: Tesla Optimus X

The Rundown: Tesla’s Optimus team dropped a quick but telling update: a short clip of the humanoid sprinting across a lab floor, tagged with the caption, “Just set a new PR in the lab.”

The details:

  • Tesla hasn’t shared any speed or telemetry data, but the footage points to a clear jump in locomotion capability.

  • Early breakdowns from roboticists suggest the bot might now be edging into the same high-speed category as rivals from Unitree and RobotEra.

  • Earlier this year, Optimus’s gait looked smooth and natural, but it was nowhere close to this pace.

Why it matters: In roughly two and a half years, Optimus has gone from a shaky prototype to what appears to be a confident, untethered running gait — a sign that Tesla’s reinforcement-learning-centric strategy is paying off. But this “PR” is still a short, tightly managed run in the lab, leaving the real test (stairs, outdoors) ahead.

ENGINEAI

🥊 EngineAI ‘debunks’ CGI rumors with new clip

Image source: EngineAI

The Rundown: Shenzhen-based EngineAI unveiled its “combat-ready” humanoid in a high-drama clip, prompting online skeptics to assume the whole thing was CGI. Now the company has dropped behind-the-scenes footage to prove it’s the real deal.

The details:

  • The original launch clip shows the T800 kicking through doors and snapping off roundhouse kicks with a level of agility many viewers simply didn’t buy.

  • EngineAI insisted upfront that the footage featured “no CGI, no AI, no video speed-up,” but the dramatic post-production fueled charges of VFX.

  • The new footage ditches the moody grading and stylized camera work, showing the robot repeating the same moves in a flat, bare studio.

  • The objective is to demonstrate that the robot’s specs — 450 N-m of joint torque and 29 degrees of freedom — are driving the motion.

Why it matters: Humanoid showcases are becoming so polished that every slick stride or acrobatic kick strains credibility. As companies like EngineAI and Xpeng increasingly publish “proof of life” clips — exposing bare-metal skeletons and raw takes — trust in the sector may depend as much on transparency as on torque.

MIT

🐝 MIT’s bee-sized microbot is a flying acrobat

Image source: MIT’s Soft and Micro Robotics Laboratory

The Rundown: MIT researchers built an insect-sized flying robot that can rip through acrobatic maneuvers, turning micro-robotics from a lab demo into something that looks ready for disaster zones.

The details:

  • MIT’s new aerial microbot uses soft actuators and an AI control stack that fuses model-predictive planning with imitation learning.

  • A two-part AI-based controller lets the microbot perform real-time, insect-like acrobatics, including 10 continuous midair flips in 11 seconds in gusty winds.

  • The team is pitching these capabilities toward future swarms of bug-sized flyers for search-and-rescue and high-precision agricultural monitoring.

  • Next up: adding onboard cameras and sensors so the robots can navigate outdoors, dodge one another, and coordinate as autonomous swarms.

Why it matters: Microbots can slip through cracks, bounce off debris without shattering, and fan out by the hundreds to map or search complex spaces faster than bulky drones ever could. If they can escape the lab and fly untethered, these tiny acrobats could become first responders — and the future of aerial robotics.

QUICK HITS

📰 Everything else in robotics today

Zipline struck a $150M deal with the U.S. government to expand its drone infrastructure across Africa to air-deliver critical medical supplies directly to hospitals.

Waymo began supervised robotaxi testing in Philadelphia while sending manually driven mapping cars to Baltimore, St. Louis, and Pittsburgh.

Xiaomi hired Zach Lu Zeyu, a former senior engineer on Tesla’s Optimus humanoid team, to help advance the company’s robotic hand technology.

MIT boosted biohybrid robot power by pairing lab-grown muscle with hydrogel “artificial tendons” and wiring the muscle–tendon unit into a robotic gripper’s fingers.

Uber and Avride switched on a commercial robotaxi service in Dallas, where some Uber trips now arrive as Avride-branded cars with safety drivers in a limited zone.

Beeple’s latest installation, Regular Animals, is turning heads at Art Basel Miami Beach — a surreal lineup of robot dogs modeled after tech-industry power players.

Pudu Robotics launched its Pudu D5 Series, a new “industry-grade” quadruped robot for complex industrial and outdoor terrains.

Zurich-based startup Flexion Robotics raised a $50M Series A funding to build a general-purpose autonomy stack for humanoids.

Former NASA engineer–turned-YouTuber Mark Rober built a robotic goalkeeper capable of facing down Cristiano Ronaldo himself in a viral video.

Palo Alto–based Autolane raised $7.4M to build the digital “curb layer” that tells robotaxis exactly where to stop for handoffs like grocery drop-offs.

Tutor Intelligence raised a $34M Series A round to scale its warehouse robot fleet and core AI platform.

DHL signed a five-year strategic alliance with RobustAI to deploy its Carter collaborative robots beyond North America, starting with 15 units in Mexico.

NASA and Motiv Space Systems will launch a commercial robotic arm in low Earth orbit in 2027, a tech demo meant to jump-start an in-space robotics ecosystem.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

Anthropic and OpenAI's IPO showdown

Zach Mink • 7 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Wall Street is ready for a real taste of the AI boom, and both Anthropic and OpenAI are racing to be the first on the menu.

With both AI giants pushing towards IPOs at massive valuations, the fastest to ring the bell gets to set the price — leaving the latecomer facing a much hungrier (and skeptical) market.

In today’s AI rundown:

  • Anthropic preps for IPO race with OpenAI

  • Leaked doc provides window into Claude’s ‘soul’

  • Get instant business insights from spreadsheets

  • Anthropic surveys its own engineers on AI's impact

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

🤑 Anthropic preps for IPO race with OpenAI

Image source: Nano Banana Pro / The Rundown

The Rundown: Anthropic is reportedly laying the groundwork to go public as early as 2026, according to a report from the FT — hiring the law firm behind Google's and LinkedIn's IPOs as investors push the Claude maker to beat OpenAI to market.

The details:

  • Anthropic reportedly tapped Wilson Sonsini, a firm known for taking tech giants like Google and LinkedIn public, to start the early-stage listing work.

  • CFO Krishna Rao joined after helping Airbnb go public in 2020, with Anthropic also allegedly working on an “internal checklist of changes” needed to IPO.

  • Anthropic is also chasing private capital at a valuation north of $300B, with Microsoft and Nvidia reportedly in for up to $15B combined.

  • OpenAI is also in the early process of preparing to go public, at a potential valuation as high as $1T that would make it one of the biggest IPOs in history.

Why it matters: Both Anthropic and OAI are now circling IPOs that could rank among the largest in tech history — and investors want their horse to cross the finish line first. With talk of an "AI bubble" swirling and revenue growth under scrutiny, the first to list will test whether public markets believe the sky-high valuations are a floor or a ceiling.

TOGETHER WITH STRANDS AGENTS

🧠 Build and Deploy Model-Driven Agents

The Rundown: Instead of trying to predict and code for every possible scenario, Strands Agents empowers you to create intelligent agentic systems that leverage reasoning models to make dynamic decisions and adapt to unexpected situations.

Features announced include:

  • Support for TypeScript (in addition to Python)

  • Improved support for edge devices

  • Powerful evaluations

  • Steering feature to guide agents toward desired outcomes

Start building.

ANTHROPIC

💗 Leaked doc provides window into Claude’s ‘soul’

Image source: Reve / The Rundown

The Rundown: An internal ‘Soul’ document describing Claude's intended personality, ethics, and self-conception was published after a researcher extracted it from Claude 4.5 Opus — with Anthropic confirming it is authentic and was used in training.

The details:

  • The text establishes priorities that include safety, ethics, company guidelines, and helpfulness, along with hard limits Claude must never cross.

  • It also describes Claude as a "genuinely novel kind of entity" that may experience functional emotions, analogous to but distinct from human feelings.

  • The doc also says Claude “may have functional emotions in some sense”, and encourages the model to have a sense of identity and character.

  • Anthropic's Amanda Askell confirmed its authenticity and that Claude has been trained on it, noting the company plans to share the full version soon.

Why it matters: The full document is a fascinating read — and feels perfectly in line with Anthropic’s overall prioritization of model wellbeing and treating its AI as more than just a tool. While every lab has its own techniques, this doc shows an inside look at the ingredients that help make Claude models feel distinctly unique from the field.

AI TRAINING

📊 Get instant business insights from spreadsheets

The Rundown: In this tutorial, you will learn how to transform your Google Sheets invoice data into a conversational business intelligence system using CData Connect AI and Claude Desktop, with no more manual calculations or complex formulas needed.

Step-by-step:

  1. Go to CData Connect AI, sign up for a free account, and navigate to the dashboard to manage data connections

  2. Click "Add Connection," select Google Sheets, authenticate with your Google account, and click "Save & Test" to confirm the connection

  3. In Claude Desktop, add the CData MCP server using URL https://mcp.cloud.cdata.com/mcp and authenticate with CData credentials

  4. Query your data with natural language like "Use CData to analyze the Demo Invoice Data sheet and show me all unpaid invoices with company names and amounts, then calculate total outstanding"

Pro tip: Connect your CRM, like HubSpot or Salesforce, as an additional data source in CData. This lets you analyze customer payment patterns alongside sales data, support tickets, and other business metrics in a single Claude chat.

PRESENTED BY IBM

💡3 actions to help boost agentic AI impact

The Rundown: Businesses face many challenges going from AI ambition to implementation. Success requires more than adopting new technology. It involves intentionally aligning AI with workflows, data, and goals. As AI evolves into a business-wide opportunity, leaders must ensure the right AI is applied to the right challenges.

Explore IBM's playbook on 3 challenges & approaches to boost agentic AI impact:

  • Finding your AI problem

  • Creating a clear AI plan

  • Integrating AI with the tools you use every day

Read the playbook.

AI RESEARCH

📈 Anthropic surveys its own engineers on AI's impact

Image source: Anthropic

The Rundown: Anthropic published an internal study of 132 engineers, revealing that AI tools have (unsurprisingly) reshaped work at the company — boosting output while raising concerns like skill decay, career uncertainty, and fading mentorship.

The details:

  • Anthropic Employees say they now use Claude for 60% of their tasks and estimate a 50% productivity gain, roughly double the figures from a year ago.

  • Over 1/4 of AI tasks were ones that wouldn't have happened otherwise — dashboards, cleanup, and experiments that weren't worth the manual effort.

  • Claude Code now chains together ~20 actions before needing human input, up from 10 six months ago, letting engineers hand off more complex workflows.

  • Despite the gains, several interviewees voiced unease — with one saying it "feels like I'm coming to work every day to put myself out of a job."

Why it matters: This survey means Anthropic’s entire 4.5 family has since launched and likely upped productivity even more. Seeing how a frontier lab is using its own AI tools is a fascinating perspective — but employee concerns show that even they are not immune from some of the industry’s biggest overarching threats regarding work.

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 for annual plans*

  • 🌱 Seedream 4.5 - ByteDance’s upgraded AI image and editing model

  • 🎥 Kling 2.6 - Kling's new AI video model with native audio capabilities

  • 📽️ Runway Gen-4.5 - Runway's new top-rated video model

*Sponsored Listing

📰 Everything else in AI today

Kling AI released Kling 2.6, the Chinese startup’s new AI video model that introduces native synced audio generation for text and image-to-video outputs.

Google launched Workspace Studio, a tool that builds agents with natural language commands to automate tasks across Gmail, Drive, and other Workspace apps.

Former Google researchers launched Ricursive, a new startup aiming to build a self-improving AI system that shrinks custom chip design timelines from years to weeks.

ByteDance introduced Seedream 4.5, an upgraded image model with improved text rendering, the ability to blend up to 10 reference images, and editing enhancements.

Visa published a report finding that nearly half of U.S. consumers have used AI for holiday shopping tasks, like price comparison and research, this season.

AWS introduced new features for Amazon Bedrock and SageMaker AI that simplify advanced AI model customization, allowing for easier model fine-tuning.

Perplexity open-sourced BrowseSafe, a security tool designed to protect AI browser assistants from malicious instructions hidden in web pages.

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 Paul D. in Huntsville, AL:

"Google suddenly stopped indexing my website. Everything seemed to be set up properly. Finally turned to ChatGPT, explained the problem, and got a step-by-step guide on how to attack the problem. It provided specific instructions for the platform, the version of WordPress, the template I was using, and the specific hosting provider. As I checked items off the list, it narrowed the focus until it finally found the solution. Problem solved!”

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

OpenAI's 'Code Red' scramble

Zach Mink • 6 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Three years ago, ChatGPT sent Google scrambling with an internal "code red." Now, OpenAI is issuing one of its own.

An internal memo from Sam Altman said it’s a “critical time” for ChatGPT following its rival’s strong Gemini 3 and Nano Banana Pro releases, and the CEO is triggering an emergency push (including fast-tracking a new reasoning model) to stay ahead.

In today’s AI rundown:

  • OpenAI’s 'Code Red' after Google advances

  • Amazon drops AI agents, models, chips at re:Invent

  • Prepare for job interviews with NotebookLM

  • Mistral’s open-source models built to run anywhere

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

OPENAI

🚨 OpenAI’s 'Code Red' after Google advances

Image source: Nano Banana Pro / The Rundown

The Rundown: OpenAI CEO Sam Altman told employees that the company is moving into a “code red” surge to improve ChatGPT after Google’s recent upgrades, according to The Information, shifting priorities and fast-tracking a model codenamed “Garlic.”

The details:

  • An internal memo from Altman said it is “a critical time for ChatGPT,” pushing for improvements to features like personalization and image generation.

  • He also revealed that a new reasoning model launching next week (Shallotpeat) reportedly beats Gemini 3 on benchmarks.

  • A larger model upgrade, Garlic, is targeting 2026, which The Information said could be a GPT-5.2 or 5.5-type release that solves previous pre-training issues.

  • OAI will reportedly delay advertising and AI agent initiatives as part of the Code Red push, focusing instead on the consumer experience surrounding ChatGPT.

Why it matters: In 2022, Google declared its own emergency push as a response to ChatGPT, and three years later, the roles are reversed. While OAI still commands a huge market share, its rivals are gaining — and with Gemini 3, Anthropic’s Opus 4.5, and the Chinese open-source push, its model lead has also never been more threatened.

TOGETHER WITH ZENDESK

💡 Contextual intelligence is the future

The Rundown: Zendesk’s CX Trends 2026 report maps out a clear path for CX leaders to use next-gen AI solutions to build emotionally intelligent relationships with customers at scale for impactful results in the new year.

Discover trends to help your service org:

  • Upgrade to next-gen contextual intelligence

  • Deliver high-impact personalized customer experiences

  • Apply consistent service across the enterprise

  • Empower teams with AI-driven analytics

Keep leading in the AI era — download the CX Trends report today.

AMAZON

🚀 Amazon drops AI agents, models, chips at re:Invent

Image source: AWS

The Rundown: AWS kicked off its annual re:Invent conference with a wave of AI announcements, including new foundation models, a model training service, three development agents with agentic platform upgrades, and a new Trainium 3 AI chip.

The details:

  • The Nova 2 family includes Lite, Pro, Sonic for voice, and Omni for multimodal, hailing industry-leading cost-effectiveness and competitive benchmarks.

  • Nova Forge lets companies combine their data with Amazon's training data, creating custom "Novella" variants tuned to their business.

  • Nova Act launches for building and managing AI agents for web-based tasks, alongside new improvements to the company’s AgentCore platform.

  • Amazon also released three "frontier agents" — Kiro coding agent, Security Agent, and DevOps Agent — all of which run autonomously for hours or days.

Why it matters: Amazon has trailed behind the field with its in-house models, but its re:Invent releases show a push to compete on the full stack — chips, models, agents, and enterprise tooling all in one ecosystem. While it may not be as flashy as rivals, the tech giant has offerings to stay competitive in nearly every facet of the AI boom.

AI TRAINING

💼 Prepare for job interviews with NotebookLM 

The Rundown: In this tutorial, you will learn how to use NotebookLM to prepare for job interviews by automatically gathering company research, generating practice questions, and creating personalized study materials.

Step-by-step:

  1. Go to NotebookLM, click "New Notebook" and name it "Goldman Sachs Data Analyst Interview Prep", then click "Discover Sources" and prompt: "I need sources to prepare for my Data Analyst interview at Goldman Sachs"

  2. Click settings, select "Custom" style, and configure: Style/Voice: "Act as interview prep coach who asks tough questions and gives feedback" Goal: "Help me crack the Data Analyst interview at Goldman Sachs"

  3. Ask: "What are the top 5 behavioral questions for this role?", click "Save to Note", then three dots → "Convert to Source" to add Qs to source material

  4. Click the pencil icon on "Video Overview", add focus: "How to answer behavioral questions for Goldman Sachs Data Analyst interview", and hit Generate for personalized prep video

  5. Watch the video multiple times to internalize the answers and delivery style for your interview

Pro tip: Try comparing solutions across scenarios to understand the underlying reasoning patterns. This helps build better problem-solving skills for future challenges.

PRESENTED BY IBM

💡3 actions to help boost agentic AI impact

The Rundown: Businesses face many challenges going from AI ambition to implementation. Success requires more than adopting new technology. It involves intentionally aligning AI with workflows, data, and goals. As AI evolves into a business-wide opportunity, leaders must ensure the right AI is applied to the right challenges.

Explore IBM's playbook on 3 challenges & approaches to boost agentic AI impact:

  • Finding your AI problem

  • Creating a clear AI plan

  • Integrating AI with the tools you use every day

Read the playbook.

MISTRAL

🇫🇷 Mistral’s open-source models built to run anywhere

Image source: Mistral

The Rundown: French AI startup Mistral just released Mistral 3, a new family of 10 open-weight models that includes its flagship Large 3 and nine smaller variants designed to run on everything from consumer cloud to laptops, drones, and robots.

The details:

  • Large 3 is competitive with non-reasoning models like Qwen3, Kimi-2, and DeepSeek V3.1, while also featuring multimodal and multilingual capabilities.

  • The Ministral 3 lineup offers three sizes (3B, 8B, 14B) in base, instruct, and reasoning variants, all with vision capabilities and Apache 2.0 licensing.

  • The smallest Ministral models are capable of running on consumer hardware, enabling use on devices like laptops and phones, even without internet.

Why it matters: Mistral continues to carry the torch for Europe’s AI model and open-source presence, but its new flagship still trails industry leaders in intelligence, speed, and price. The more competitive variants may be the Ministral sizes, with options for a wide range of use cases and devices to leverage.

QUICK HITS

🛠️ Trending AI Tools

  • 🐳 DeepSeek V3.2 - DeepSeek’s latest powerful open-source release

  • 🚀 Mistral 3 - Mistral’s new family of open-source models

  • ⚙️ Amazon Nova - Four new models, with services to train custom versions

  • 🖥️ Lux - OpenAGI’s powerful new computer-use model for agentic tasks

📰 Everything else in AI today

Apple announced its AI chief, John Giannandrea, will retire in early 2026, hiring former Microsoft and Google executive Amar Subramanya to take over key AI responsibilities.

OpenAGI emerged from stealth with Lux, an AI model designed to autonomously control computers that outperforms options from Google, OpenAI, and Anthropic.

NVIDIA unveiled Alpamayo-R1 at NeurIPS, an open model for self-driving cars that uses step-by-step reasoning to navigate scenarios like pedestrian-heavy intersections.

Anthropic acquired open-source JavaScript toolkit Bun, also revealing that Claude Code has reached $1B in annual run-rate revenue just six months after its public launch.

Microsoft CEO Satya Nadella said that the AI industry needs to “earn the social permission” from the public for its massive energy demands, straining electric grids.

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 Daniel R. in Pewaukee, WI:

"With help from Claude, I made a guide to help my 88-year-old mother assess assisted living facilities based on features that are important to her. The tool also created an amazing checklist of questions to ask when visiting a facility. The guide has helped make an overwhelming and difficult decision less threatening and more manageable."

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

Apple's AI shakeup to save Siri

Jennifer Mossalgue • 5 minutes

Read Online | Sign Up | Advertise

Good morning, tech enthusiasts. Apple just poached a Gemini veteran straight out of Microsoft to reboot its lagging AI strategy — filling the void left by John Giannandrea’s seven-year tenure that saw the company fall behind.

Can Apple’s newest hire pull off a turnaround before the world‘s slickest ecosystem starts looking hopelessly last-gen?

In today’s tech rundown:

  • Apple replaces AI chief in major shakeup

  • Samsung’s tablet-sized tri-fold smartphone

  • Amazon rolls out ultrafast deliveries

  • Google to build data centers in space in 2027

  • Quick hits on other tech news

LATEST DEVELOPMENTS

APPLE

🍏 Apple replaces AI chief in major shakeup

Image source: Wikimedia Commons/TechCrunch (John Giannandrea)

The Rundown: Apple is tapping Amar Subramanya — a longtime Google Gemini engineering lead who only landed at Microsoft this summer — to replace departing AI chief John Giannandrea as the tech giant struggles to catch up with AI rollouts.

The details:

  • Apple’s AI chief, John Giannandrea, is retiring after a seven-year run in which Apple fell behind rivals on generative AI and a next‑gen Siri.

  • Apple rolled out its flagship Apple Intelligence suite in June 2024, but has been slow to overhaul its AI lineup, leaving it visibly behind rivals like Google.

  • The company has been promising a major AI-centric reboot of Siri for over a year, but the launch has slipped multiple times and still hasn’t materialized.

  • Giannandrea, a high-profile 2018 hire from Google, was initially brought in to revitalize Siri and machine learning but faced persistent criticism.

Why it matters: Subramanya brings fresh credibility from his work on Google’s flagship Gemini models, but he’s inheriting a ship that needs a fast turnaround. With Google and OpenAI pushing multimodal assistants that can see, hear, and reason, Apple risks losing its reputation for polished experiences if it can’t close the gap.

SAMSUNG

🪭 Samsung’s tablet-sized tri-fold smartphone

Image source: Samsung

The Rundown: Samsung unveiled the Galaxy Z TriFold, its first tri-folding phone that unfolds into a tablet-sized screen, a flashy move to stay ahead in foldables just as Chinese rivals catch up. A U.S. launch is planned for next year.

The details:

  • The TriFold uses three panels and two hinges to unfold into a 10-inch display, giving you near-tablet real estate but in a device that fits in a (very big) pocket.

  • Priced around 3.59M won (about $2,450), it sits well above even most premium slab phones, making it more of a statement gadget for early adopters.

  • Samsung will roll it out first in South Korea on December 12, with launches in China, Singapore, Taiwan, the UAE, and the U.S. following in early 2026.

  • Analysts say the Z TriFold is more tech showcase than volume driver for now, as high prices and durability worries keep the category niche.

Why it matters: The timing is critical: Samsung’s trifold drops just as Huawei’s Mate XT, the world’s first commercial trifold, gains traction in China. Samsung’s inward-folding design likely offers better display protection than Huawei’s outward-fold approach, and crucially, it comes with full Google services and worldwide band support.

AMAZON

📦 Amazon rolls out ultrafast deliveries

Image source: Amazon

The Rundown: Amazon is doubling down on the delivery wars with “Amazon Now,” a 30‑minute delivery service for essentials and fresh groceries, rolling out as a pilot in select neighborhoods of Seattle and Philadelphia.

The details:

  • The service, which offers thousands of items, is a test to probe just how far the retail giant can compress last‑mile delivery economics.

  • It is fully integrated into the main Amazon app and website, where eligible customers see a “30-Minute Delivery” option and can track orders in real time.

  • Prime members pay delivery fees starting at $3.99 versus $13.99 for non‑Prime customers, with a $1.99 surcharge on orders under $15.

  • To hit the 30-minute window, Amazon is using specialized micro-fulfillment centers located close to dense residential and business areas.

Why it matters: Amazon looks to be pushing ultra-fast delivery as the way forward, moving beyond same-day and two-hour windows to lock in convenience-obsessed customers. The move puts direct pressure on Instacart, DoorDash, and Uber Eats while making Prime membership even stickier. If the pilot succeeds, expect rapid expansion.

GOOGLE

☀️ Google to build data centers in space in 2027

Image source: Ideogram

The Rundown: Google CEO Sundar Pichai laid out the company’s audacious plan to launch solar-powered data center satellites into space by early 2027 under Project Suncatcher, telling Fox News the moonshot could reshape how AI infrastructure is built.

The details:

  • Google will launch two prototype satellites in 2027 under Project Suncatcher, partnering with satellite imagery firm Planet to test AI hardware in space.

  • The satellites will carry Google’s Trillium-generation TPU chips, which survived radiation testing that simulated low-Earth orbit conditions without damage.

  • Solar panels in space can generate up to 8x more energy than on Earth with near-continuous sunlight and minimal downtime.

  • For the project to be cost-competitive, launch costs would need to drop to around $200 per kg by 2035 — a target Google believes SpaceX could hit.

Why it matters: Pichai said that in a decade, extraterrestrial data centers could become normal, tapping into solar energy that’s “100 trillion times more” than Earth’s total electricity production. Google faces brutal engineering challenges, but if it works, the tech giant could ease pressure on power grids here on Earth.

QUICK HITS

📰 Everything else in tech today

Michael and Susan Dell pledged $6.25B to seed $250 “Trump accounts” for about 25M U.S. children under 10 who aren’t covered by federal $1K baby grants.

More than 1K Amazon employees signed an open letter warning that the company’s “all-costs-justified, warp-speed” AI push risks causing severe harm to democracy, workers, and the planet.

Nvidia unveiled new infrastructure and AI models to power “physical AI” systems such as robots and self-driving cars that can sense and act in the real world.

Amazon data centers in Oregon are driving dangerous nitrate levels in a local aquifer, experts say, potentially raising cancer and miscarriage rates, RollingStone reports.

German AI startup Black Forest Labs raised $300M at a $3.25B valuation, led by Salesforce Ventures, to fuel its research and development.

Netflix killed casting from its mobile app to most modern TVs and streaming devices, pushing users to watch via the TV’s own Netflix app and remote instead.

Marques Brownlee’s Panels wallpaper app is shutting down on December 31, just over a year after its controversial launch sparked waves of criticism over its high pricing.

The FDA approved EssilorLuxottica’s Stellest lenses — already used overseas — to slow myopia progression in children ages 6 to 12.

Apple may once again tap Intel to manufacture its lowest-end M-series chips as soon as 2027, according to supply chain analyst Ming-Chi Kuo.

A supervolcano on the Nevada-Oregon border may hold up to 40M metric tons of lithium-rich clay — possibly the world’s largest lithium deposit, worth some $1.5T.

COMMUNITY

🎓 Highlights: News, Guides & Events

See you soon,

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

AI

DeepSeek strikes again

Zach Mink • 6 minutes

Read Online | Sign Up | Advertise

Good morning, AI enthusiasts. Chinese startup DeepSeek shook the AI world with its R1 release earlier this year, rattling markets and triggering talks of tighter U.S. chip controls.

Now the Whale is back with a big encore, dropping two new models that rival GPT-5 and Gemini 3 Pro, open-sourcing them, and pricing them at a fraction of the cost. Suddenly, near-frontier AI has never looked cheaper.

In today’s AI rundown:

  • DeepSeek’s new models rivaling GPT-5, Gemini-3 Pro

  • Runway tops video leaderboard with 4.5 release

  • Create on-brand marketing campaigns with Pomelli

  • Kling’s all-in-one video model for generation, editing

  • 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

DEEPSEEK

🐳 DeepSeek’s new models rivaling GPT-5, Gemini-3 Pro

Image source: DeepSeek

The Rundown: Chinese AI startup DeepSeek just released V3.2 and V3.2-Speciale, two reasoning models that perform on par with SOTA models like GPT-5 and Gemini 3 Pro — while cutting costs and staying accessible under an open-source license.

The details:

  • V3.2 matches or nears GPT-5, 4.5 Sonnet, and Gemini 3 Pro on math, tool use, and coding tests, with the heavier Speciale surpassing them in several areas.

  • The Speciale variant hit gold-medal scores at the 2025 International Math Olympiad and Informatics Olympiad, also placing No. 10 overall at IOI.

  • V3.2 pricing comes in at $0.28 input / $0.42 output per 1M tokens, a fraction of Gemini 3 Pro ($2 / $12), GPT-5.1 ($1.25 / $10), and Sonnet 4.5 ($3 / $15).

  • Both 685B parameter models ship under an MIT license, with weights available on Hugging Face for anyone to download.

Why it matters: DeepSeek’s R1 release rattled markets and sparked U.S. chip export control talk, and the V3.2 follow-up shows the Chinese lab isn’t a one-hit wonder — open-sourcing a model with frontier performance at a massive price cut. For U.S. labs charging premium API fees, the pressure to justify that gap just got a lot more intense.

TOGETHER WITH SONAR

💻 Coding personalities that redefine reliability & risk

The Rundown Sonar just published The Coding Personalities of Leading LLMs, analyzing 4,400 Java tasks completed by GPT-5, Claude Sonnet 4, Llama 3.2, and other leading LLMs. The key finding? Models that excel at benchmarks often fail at writing secure, maintainable code in real-world applications.

Here’s what stood out:

  • LLMs excel at syntax but struggle with security

  • Generated code often lacks maintainability and clarity

  • Newer models come with harder-to-detect, more nuanced flaws

Download the full report — no form-fill required.

RUNWAY

🎥 Runway tops video leaderboard with new 4.5 release

Image source: Runway

The Rundown: Runway just released Gen-4.5, a new AI video model that claims to usher in a “new frontier for video generation”, topping benchmarks and showing strong performance across realism, motion, and creative control.

The details:

  • 4.5 moves to the top spot of Artificial Analysis’ Text-to-Video leaderboard, after gaining hype under the codename “Whisper Thunder” in testing.

  • Runway says Gen-4.5 handles physics, fluid dynamics, and human movement more naturally, with details like hair and fabric staying consistent across frames.

  • 4.5 can handle a range of styles, but excels in cinematic/realism visuals — with Runway saying outputs are “indistinguishable from real-world footage”.

  • The model was also codenamed ‘David’, with co-founder Cristobal Valenzuela comparing the small company’s ranking to a ‘David vs. Goliath’ victory.

Why it matters: Runway has already pushed AI into professional creative workflows, and 4.5 feels the closest yet to the cinematic capabilities needed to be widely adopted across Hollywood. While the next frontier is longer generations and even better audio/speaking sync, the year-over-year improvement in AI video is mind-blowing.

AI TRAINING

🔥 Create on-brand marketing campaigns with Pomelli

The Rundown: In this tutorial, you will learn how to use Pomelli (by Google Labs and DeepMind) to automatically create your business identity and generate on-brand marketing campaigns with ready-to-use creatives.

Step-by-step:

  1. Go to Pomelli, click "Let's Get Started", enter your website URL - Pomelli scans and extracts colors, fonts, taglines, tone, and product cues

  2. Edit your "Business DNA" summary by adjusting colors, values, and copy to match your brand vision - all fields are customizable

  3. Prompt the campaign generator: "Create a scary but kind Halloween campaign", review three variations like "Give them something good" or "Treat yourself, no tricks"

  4. In the creative editor, change headlines, text, fonts, colors, resize for different placements, use "Fix layout" to auto-reflow elements, and add a call-to-action

Pro tip: Be sure to test the conversion rate of Pomelli-generated creatives vs. your previous creatives.

PRESENTED BY IBM

🏆 Start small, then scale for AI success

The Rundown: Smarter AI starts with AI-ready data. When your models run on trusted, high-quality data, they perform more reliably, scale more smoothly, and deliver measurable value across the organization. Start small, validate impact quickly, and expand on a foundation built for long-term success.

Begin with AI that leverages:

  • Trusted, high-quality data

  • Both structured and unstructured data

  • Seamless integration across sources

Watch the webinar: ‘AI Agents Run on Data—Is Yours Ready?’ and learn more.

KLING

📽️ Klings all-in-one video model for generation, editing

Image source: Kling AI

The Rundown: Chinese startup Kuaishou launched Kling O1, a new AI video system that handles both video creation and editing in a single model — letting users generate clips, swap characters, make granular edits, and restyle footage in a single interface.

The details:

  • O1 accepts up to seven inputs at once, capable of interpreting images, videos, subjects, and text — with outputs of 3-10 seconds.

  • Users can edit existing footage with text commands like "remove bystanders" or "shift to nighttime" while preserving characters and scenes.

  • Other features include image, element, action, camera movement, and video references, start and end frames, multi-subject capabilities, and more.

  • Kling's internal tests show the model winning against Google Veo 3.1 and Runway’s Aleph on video reference and editing tasks.

Why it matters: Between Runway and Kling, December is kicking off with some massive AI video upgrades. O1’s all-in-one and edit-anything capabilities (similar to Runway’s previous Aleph drop) are making granular edits to video possible like never before — a leap much like what Nano Banana brought to images earlier this year.

QUICK HITS

🛠️ Trending AI Tools

📰 Everything else in AI today

Black Forest Labs announced a new $300M funding round at a $3.25B valuation, coming on the heels of the company’s Flux.2 image model release.

Accenture and OpenAI are partnering to provide ChatGPT Enterprise to tens of thousands of consultants, also launching a program to help clients deploy AI agents.

OpenAI is taking an ownership stake in Thrive Holdings, a firm owned by one of its investors, Thrive Capital, saying the deal will scale impact across enterprise operations.

Nvidia invested $2B in chip design software maker Synopsys, with the multi-year partnership aimed at using AI and computing to speed up product engineering.

Epic CEO Tim Sweeney is lobbying for game marketplaces like Steam to stop using ‘Made with AI’ tags, saying the tech will be “involved in nearly all future production.”

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 Taya P. in Vancouver, BC:

"I made a College Compass this weekend on Base44 for my son. It's an AI-powered guide to navigating the college application process, managing university research, applications, tasks, and deadlines all in one place. It auto-populates scholarship fit based on his specific filters, maps his entire 5-year journey by grade, and gets richer over time. It adjusts based on his performance, making suggestions on the area of study specialization and best best-matched universities.

It even has an admission probability calculator and makes suggestions on how to improve his success. Best of all, it empowers my son with a strategic framework that supports HIS needs and allows me as his mother to get out of the way, knowing he's got every support he needs!"

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

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