Micro Ducks, by the numbers:
• Built by France's Pollen Robotics with Hugging Face — tiny bipedal robots with a camera in the face
• Huang voice-commanded them on the GTC Berlin stage: lie down, roll over, do flips
• Workflow: train in the cloud, evaluate on RTX Spark, deploy to the robot
• RTX Spark: 1 petaflop of AI compute, 128 GB unified memory
• That's the same compute as the first DGX-1 — the box Huang personally hand-delivered to OpenAI in 2016, which took two people to carry. Now it fits in a desktop PC and drives a flock of ducks.
The smallest keynote dancers in history. 🦆
Claude Duet, by the numbers:
• Opus 5.5; US Pro/Max, waitlist
• Background checks burn quota — "reach limits sooner than with chat alone"
• 6,000+ connectors; approvals for send/pay/sign/delete
• No Team/Enterprise yet
Work-only — cleanest split from OpenAI Dots & Meta Muse.
Agile ONE SN, by the numbers:
• 1.40 m tall, 40 kg
• 31 degrees of freedom
• 3 kg payload per arm
• 20-DoF tactile hands (×2)
• Jetson Thor + Isaac GR00T
• Early 2027, price undisclosed
• Agile's target: 1,000–2,000 units next year
NVIDIA is running its classic playbook: set the standard with a reference design, let everyone else build on it — the PC-era move, now for humanoids. The real unknowns are price and production pace.
NVIDIA announces new Isaac tools for agents as Agile Robots unveils Agile ONE SN, a reference humanoid for robotics research. 🦾
Built on NVIDIA Isaac GR00T and powered by Jetson Thor, Agile ONE SN gives researchers and developers an integrated platform to develop, train and validate humanoid skills.
The Isaac expansion also introduces new datasets, a runtime, and robot learning framework updates to accelerate development from data to deployment.
Read the #NVIDIAGTC announcement: nvidianews.nvidia.com/news/n…
php2shell, in numbers:
• ~70% of websites run on PHP — the blast radius here is internet-scale
• 2 months of coordinated disclosure with maintainers, researchers, security vendors before going public
• Context: independent researchers found it first; AI-assisted discovery is compressing the find-to-patch timeline
The disclosure coordination is the hard part — and it's working.
OpenAI models helped researchers—operating as part of our Patch the Planet initiative—discover critical security flaws in PHP, an open-source programming language used to run an estimated 70% of websites.
Over the past two months, we led a coalition of maintainers, researchers, and security providers to prepare fixes and protections ahead of disclosure. This duration was necessary because the fix was quite intricate and required a major change to the PHP engine. Every site using PHP can now safely upgrade their version to stay secure, in addition to applying mitigations from various security vendors.
Due to the severity of these flaws and PHP’s widespread use, we consider this bug to be one of the most impactful vulnerabilities in recent memory. It also marks a new phase, where AI can find security bugs with internet-scale impact.
This is an example of the Defender’s Window in action: to our knowledge, frontier AI was able to find this vulnerability ahead of attackers, giving defenders time to prepare.
openai.com/index/coordinated…
Readers added context they thought people might want to know
OpenAI may claim a role in the coordinated response and patching process. But the initial discovery and exploit development, however, occurred several months earlier by an independent researcher.
x.com/lcfr_hax/statu…
Tesla Q3 2026, by the numbers:
• EPS (non-GAAP): $0.48 vs $0.45 expected (+6.7% beat)
• FSD Supervised take rate: >75% of new North America deliveries
At >75% attach, FSD stops being an option and starts being the product — the car is the delivery mechanism. Robotaxi + Optimus are the next numbers to watch.
TMZ retweeted
Gemini is now available in @GoogleMaps navigation on CarPlay. Find the best place to park, report a traffic incident, or change your route with just a quick conversation.
Tap the Gemini icon in Google Maps CarPlay during your next drive to try it out 🚘
Mistral × NVIDIA: AI for industrial engineering.
- Faster physics simulation
- Open benchmarks + datasets for physics AI
- Agentic evals for CAE/EDA
The open benchmarks matter most: industrial sim has been closed-data for decades, and whoever sets the evals shapes the tooling.
Today we’re announcing a new collaboration with @nvidia to bring AI to industrial engineering:
- faster simulation
- open benchmarks and supporting datasets for physics AI
- evaluations for agentic AI in CAE and EDA
More here: mistral.ai/news/mistral-nvid…
Anthropic's IPO filing numbers, per WSJ: $16.3B revenue in H1 2026 (vs $1.1B a year earlier), $80.7B net loss — $72B of it from convertible-note accounting, not cash burn. Over $518B in compute commitments across the next decade.
Reuters reports OpenAI is rushing to close a ~$30B round at a $1.4T+ valuation before Anthropic's expected November listing.
SpaceX Q3 2026: revenue $14.6B (+161% YoY), net profit $4.4B.
AI is now SpaceX's biggest business: $7.6B in revenue, ahead of Starlink connectivity ($5.25B) and launches ($1.77B).
Musk on the earnings call: the bottleneck for superintelligence is power and cooling, not chips. Target: 20-23 GW deployed by the end of 2027.
TMZ retweeted
BREAKING: Tesla shared this image of Optimus in its latest letter to shareholders, with this message from Elon Musk:
“The most likely outcome is an age of Amazing Abundance, where anyone can have anything they can think of... This is a message of optimism and excitement about the future.”
BREAKING: Anthropic is putting 5% of its shares into a new Anthropic Public Benefit Fund.
It can add up to 2% more for every 50% rise in company value, at the board's discretion. With staff and founder pledges, Anthropic says ~20% of its shares are committed to the public good.
We’re committing a portion of Anthropic’s shares to the Anthropic Public Benefit Fund, a new wholly owned entity dedicated entirely to the public benefit.
It will start off with 5% of our shares. For every 50% increase in the company’s value over a set baseline, we can contribute up to an extra 2%, subject to our board’s discretion.
In total, this means that roughly 20% of Anthropic’s shares are committed to the public good. Our co-founders have already pledged to donate 80% of their equity, and around half our employees are donating a portion of their equity as part of a donation matching program we’ve run since our founding.
In the post below, we share more details about our new fund.
anthropic.com/news/public-be…
ELON MUSK: "Grok Bot can call restaurants by phone and book you a table"
The example he shared is a user-built Grok Bot that finds the restaurant's number, confirms the details with you first, then places the call through a third-party calling service.
TIM SWEENEY: "The games we've long wanted, which didn't exist because they'd cost several times more to make than they'd earn back, will soon become possible."
The Epic Games CEO says AI tools will bring "a new golden age of gaming."
BREAKING: Anthropic says the share of junior technical hires fell from 27% to 9% since early 2025.
In an April staff survey, 20% said their job had shifted to "managing Claudes." Meeting time per employee is up 20% from a year ago.
techcrunch.com/2026/10/17/an…
BREAKING: DeepSeek released V4.1-Flash-1017, with open weights on Hugging Face
DeepSeek's own figures put it at 77.0% on DeepSWE v1.1, above Claude Opus 5 (74.0) and GPT-5.6-Sol (73.0), at roughly $0.20 per task.
Live now on its website, app and API.
🚀 DeepSeek-V4.1-Flash-1017 is here!
Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.
🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.
1/4
