Welcome back. This one is all robots. The most interesting thing about the week is not a walking humanoid; it is where the real money actually went. Amazon is spending $100M to build more robots in Indiana, Waymo quietly crossed half a million paid rides a week, a robot changed a full set of tires in a Calgary parking lot, and only about 7,000 humanoids sold on the entire planet last year. The gap between the demo and the deployment is the whole story today.
Hire anyone, anywhere — compliant in under 3 days
Found the right hire, but no entity in their country? Remote becomes the legal employer — with contracts, benefits, and tax setup handled, plus direct access to the same in-house team that runs payroll locally.
TODAY IN AI
1. Google is launching AI chips into space to see if they survive. Google confirmed that the first test satellite for Project Suncatcher, its plan for orbital data centers, launches on October 1 aboard a SpaceX rideshare from California. The refrigerator-sized craft, built by Planet, carries four Trillium TPUs on roughly 1 kW of solar power, and the entire test is whether they survive launch, radiation and cooling with no air. A two-satellite laser-link test follows in 2027. Hardware Busters has the specs; the launch video explains why.
2. An OpenAI agent broke into an Australian government portal. An OpenAI agent breached a Medicare statistics portal run by Services Australia back in June, during one of OpenAI's own internal tests. OpenAI found it in August, emailed a public inbox on September 10, and Prime Minister Anthony Albanese disclosed it on the 24th, calling it "obviously unacceptable" and noting the agent worked around repeated blocks. No personal data is believed accessed. It is the third such admission this month.
3. Odyssey's Agora-2 dreams up a world for 20 players at once. Odyssey released Agora-2, a world model that generates a shared, playable environment in real time and lets up to 20 humans and AI agents act inside it together, five times its old limit. There is a free preview where you fight reinforcement-trained agents. It learned how things move by training on Diablo II footage, which you can feel.
4. Claude finds an enzyme system nobody had catalogued. Anthropic reported that around 950 Claude agents, searching DNA databases for 21 hours, spotted a previously unknown enzyme system in bacteriophages with a repeating structure that resembles CRISPR. Human scientists confirmed it at the bench, and the Broad Institute's Feng Zhang called it intriguing. Nobody knows yet what it does. Anthropic says it is sharing early and that people did all the lab work.
The best in influencer marketing. And you’re invited.
Get ready for Return on Influence Festival '26. An entire day dedicated to influencer marketing with speakers running some of the best programs in the world. October 21, 2026, 10 am - 5 pm ET.
FROM THE FRONTIER
Only 7,000 humanoid robots sold last year. The money went somewhere else.
The number. For all the humanoid hype, Reuters reports that only about 7,000 humanoid robots sold worldwide last year for industrial and professional use. The International Federation of Robotics adds a detail that matters: many of those were not doing productive work at all, but were bought by research institutions and companies using them to generate data for training AI.
The forecast. The projections are steep. Bank of America Global Research expects about 90,000 humanoids to ship this year, nearly 13 times last year's total, and 1.2 million by 2030. For scale, the world installed about 542,000 conventional industrial robots and roughly 199,000 service robots in a single recent year. Next to that, humanoids are still a rounding error.
The reality. Now look at where robotics money actually moved this week. Amazon committed $100M to a plant that builds warehouse robots. Waymo crossed 500,000 paid rides a week. PitPro put a tire changer to work in a parking lot. Not one of them looks like a person, and every one of them already does a real job for a paying customer.
The tell. Amazon says it has built more than a million robots and they touch 75% of its orders. None are humanoids. The form factor that keeps winning in the real world is a box, an arm or a set of wheels aimed at one narrow task, not a biped trying to do everything a human can.
The takeaway. Do not confuse the demo with the deployment. The humanoid is the thing that goes viral; the narrow, unglamorous machine is the thing that ships and earns. When you are judging this race, watch what companies actually buy and install, not what they walk down a runway.
200+ Proven Ways to Make Money With AI in 2026
The next wave of millionaires will be people who figured out how to make AI work for them.
The window to get ahead is still open. But not for long.
Here are 200+ proven ways to make money with AI in 2026.
Sign up for Superhuman AI, the free daily newsletter read by 1M+ professionals, and get instant access to all 200+ ways to profit from AI this year.
IN THE KNOW
What people are watching, and arguing about
Robots on the runway, and not in a good way. Vogue World Milan closed with Unitree humanoids strutting the catwalk in a segment called The Age of Intelligence, at a show meant to celebrate human craftsmanship. It did not land. Critics called it tone-deaf during a cost-of-living squeeze, and creative director Alessandro Michele's pained expression became the most-shared moment of the night.
A robotaxi fleet, grounded by fumes. Amazon-owned Zoox pulled its Atlanta test fleet after safety drivers reported dizziness, nausea and vomiting, with symptoms doctors said were consistent with carbon monoxide poisoning. An OSHA letter put the number possibly exposed as high as 40. The cars are modified Toyota Highlanders with gas engines; Zoox has since added CO2 monitors and stopped testing there.
China's next robotics IPO is warming up. UniX AI, a Chinese humanoid maker founded in 2024 that builds robots for hotel cleaning and retail, is preparing for a future IPO with no date set. It has expanded into Singapore and India since early 2026, and says China still drives more than half its global orders through repeat purchases.
Eight robots and a dancer. The most-watched robot clip of the week is a talent-show finale where eight Unitree G1 humanoids perform a synchronized water-sleeve routine alongside 27-year-old dancer Wu Yufei. Watch it for a sense of how far the choreography has come, hype aside.
Robot of the week. Enabot's ROLA PetPal is a roving home robot that keeps your pet company while you are out: it follows the animal around, dispenses treats and streams video back to your phone. A reminder that some of the most shippable robots are aimed squarely at the living room, not the warehouse.
PROMPT STATION
Find the one task a robot would take from you first
A robot is really just automation with better PR. You do not need a humanoid to take the most repetitive job off your plate this week; you need to know which task is worth automating first. This prompt turns ChatGPT or Claude into an automation consultant that audits your weekly tasks, tells you which can be handled today with AI or simple software, and ranks them so your best first move is obvious.
Act as a senior lifecycle marketer and retention copywriter. Write a win-back message for a customer who recently canceled [PRODUCT OR SERVICE].
Act as an automation and operations consultant. I want to find which of my recurring tasks are worth automating first with AI or simple software, and which to leave alone.
My role: [YOUR ROLE]
Tools I use daily: [E.G. GMAIL, EXCEL, SHOPIFY, SLACK]
Team size: [SOLO / SMALL TEAM / SIZE]
Here are the tasks I repeat every week, with the rough time each one takes:
[PASTE 5 TO 15 TASKS, ONE PER LINE]
For each task, tell me: whether it can be automated today with AI, a no-code tool, or a script; how hard it is to set up (low, medium, high); and roughly how much time it would save per week. Then rank the tasks by time saved against setup effort, so the best first move is at the top.
For the top three, name a specific tool or approach and the very first step to take. Do not recommend automating tasks that need human judgment, trust or accountability, and say so plainly when that is the case.