i struggle to write about cybernetics i have a blog post about Ashby's Law that i've tried writing like 5 times, just tried again yesterday and i'm probably trashing it again this time too
Information Center
AI semiotics: cool paper suggests that AIs are strongly impacted by visual symbols in ways they do not always recognize in their own chains-of-thought. The way they react to those symbols is very different than how a human might interpret them as well. arxiv.org/pdf/2610.364...
Boston Dynamics rebuilt Stretch around one metric: uptime. Gen 2 was redesigned from 100,000+ operating hours: 81% less cabling, up to 1,000 cases an hour, and a drive swap that now takes an hour, not a day. Source: Boston Dynamics Image: Boston Dynamics #Robotics #WarehouseAutomation #Logistics
Blue Texas is kerrrrrazy
The home robot that hides in plain sight as a lamp. SYNCERE just raised $11M pre-seed led by a16z Speedrun. LUME folds laundry, tidies and makes the bed, starting at $2,699. It is already in homes and learning from them. Source: Business Wire Image: SYNCERE #Robotics #HomeRobotics #PhysicalAI
Needle for Andrews in South Carolina now up to 93 percent
Auspicious
Wow, what an amazing opportunity. I feel perfectly targeted by this outreach effort.
OpenAI says their query volume is now about 20% of Google’s www.theverge.com/ai-artificia...
We are in an era of flux & fighting over how AI should be used to review academic work & by whom. But it is worth proving to yourself that yes, AI used properly can be very insightful at this, by asking GPT-6 Pro/Opus 5.5 High to give helpful, substantive feedback on one of your papers.
300,000 tasks in, Maximo Robotics is past the demo stage. Its US-made robots place heavy solar panels alongside crews at 2.5x manual rates, paid per task done. Next: 3M more tasks for 1.5 GW in 2027. Source: PR Newswire Video: Maximo Robotics #Robotics #PhysicalAI #ConstructionTech
A handful of billionaires are trying to make all the decisions about what AI means for the rest of us, and spending millions to divide the Democratic coalition on this issue. We're not going to let them.
There's a buzzy new paper about GPT-6 Astra being able to directly interpret model weights (prompts full of floating point numbers) - I tried it with Qwen 3.8 27B on my laptop and found that, after 45+ minutes, that model could achieve the same thing simonwillison.net/2026/Nov/3/q...
Binance said on Tuesday that the hackers behind last week’s $6.3 billion theft spent months preparing for the heist, luring an employee with a fake job interview and planting tampered software before using the company’s AI tool to reach its wallet systems. www.nytimes.com/2026/11/03/t...
shit, i’m staring to see value in ontologies and knowledge graphs again. meh, it’ll fade..
Reuters has increased its third-quarter revenue by 12% year on year to $232m, boosted by a surge in licensing revenue from generative AI. Transaction revenue in its news division, excluding subscription revenue, rose by 56% to $44m www.businesstimes.com.sg/companies-ma...
in hindsight, Anthropic dumping all their brainpower into mech interp suddenly makes a ton of sense
a moonshot researcher did this on K3.1 and it works. He insists they didn’t train for it at all (presumably that means no RL for it), evidence that it’s truly emergent behavior pastebin.com/G7VCWZHn
KNAPP just planted its robotics brain in Boston. KNAPPRobotics.ai in Somerville will grow to 30 to 35 people by early 2027, building physical AI for warehouse picking, with MIT next door. Source: KNAPP Video: KNAPP #Robotics #PhysicalAI #WarehouseAutomation
"Opus, great, now turn that into a movie, i know you can make it amazing. figure it out." Yes, that was the only prompt
This is an absolutely crazy thing to discover and not enough fuss is being made of it. There was no reason to believe this was even possible, far less that LLMs knew how to do it. *We* don't know how to do it, so they were not trained in it, they just figured it out.
Here is a mass attempted replication of old findings of LLMs with new models. Much conventional wisdom is outdated. Recent models no longer invent citations and both hallucination rates & stereotyping has dropped. But they are still sycophantic & lean left. zou-group.github.io/llm-claims-s...
If this holds up, it is big: "We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training." rikiyat.github.io/blog/weight-...
Walmart just committed $6B to warehouse robots. None of them are humanoids. Symbotic's deal covers 25 distribution centers and up to $6.92B in added backlog. The catch: Walmart was 85% of its revenue. Source: Symbotic / Walmart Video: ROBOTICS.BIZ BRIEF #042 #Robotics #WarehouseAutomation
Tutor Intelligence just raised $75M to scale its robot workers. The a16z-led Series B brings it to $117M. Sonny and Cassie learn from every deployment, and customers pay by the hour, no upfront buy. Source: Business Wire Video: Tutor Intelligence #Robotics #PhysicalAI #WarehouseAutomation
AI an ‘inevitable friend’ to publishers as deepfakes will put premium on trusted content, claims PR boss pressgazette.co.uk/publishing-s... #Sponsored
When you think about the Labs' weird approach to software releases (sudden changes, bad communication), it is worth remembering that a core belief is that these products are a temporary thing they have to do to raise money for a couple years until "true" AGI makes all products obsolete.
MicroVision's lidar is moving deeper into FieldAI's robots. FieldAI bought MOVIA sensors after testing IRIS on ground vehicles and MOVIA Air Plus drone mapping, and is widening the work across federal and industrial programs. Source: MicroVision Video: FieldAI #Robotics #DefenseTech #Lidar
Loved this post from @marcua.net. It's almost so obvious that it shouldn't have to be written, but I'm glad someone wrote it. A lot of the time I've gotten back thanks to AI has gone to prototyping and performance, benefiting user experience in ways that previously would have meant less features.
Foxconn has #Humanoid robots working in Texas now. More than 10 are deployed at its Houston sites, where six buildings make Nvidia AI servers. Their brains build on Nvidia's GR00T model, and Foxconn plans to go beyond Texas. Source: Reuters Image: Hon Hai USA #Robotics #Humanoid #PhysicalAI
Private to this visit. Text and context are processed by AI.