Skip to content →

Tag: ai

telemetry 13nov25

Published November 13, 2025 by Warren Ellis

Mark Elder conducts the London Philharmonic Choir and Orchestra in a sea-themed concert as part of the orchestra’s Harmony with Nature season. Sibelius’s powerful depiction of the ocean waves in his Oceanides is paired with Vaughan Williams’s mighty first symphony, the Sea Symphony with soloists Masabane Cecilia Rangwanasha and David Stout.

After testing three different toys powered by AI, researchers from the US Public Interest Research Group found that the playthings can easily verge into risky conversational territory for children, including telling them where to find knives in a kitchen and how to start a fire with matches. One of the AI toys even engaged in explicit discussions, offering extensive advice on sex positions and fetishes.

COPENHAGEN, Nov 12 (Reuters) – Iceland has designated the potential collapse of a major Atlantic Ocean current system a national security concern and an existential threat, enabling its government to strategize for worst-case scenarios, the country’s climate minister told Reuters.

Comments closed

telemetry 1nov25

Published November 1, 2025 by Warren Ellis

On All Hallow’s Eve, Jennifer Lucy Allan turns the cards and listens for what they reveal, tracing sonic lines across the tarot deck. From the ghostly atmospherics of William Basinski’s Wheel of Fortune, to the arcane explorations of early electronic pioneer Ruth White and Swiss krautrock mystic Walter Wegmüller, the spread unfolds in unexpected ways, its order uncertain, its juxtapositions surprising. Expect new sounds from Argentinian artist aylu, whose spiritually-charged album journeys from personal struggle to collective resistance, as well as slow-motion noise conjured by New Zealand’s drone trio Surface of the Earth.

Comments closed

morning computer i don’t even

Published October 28, 2025 by Warren Ellis

Vipoo Srivilasa.

An AI-powered gun detection system hooked up to a Baltimore County high school’s cameras mistook a bag of Doritos chips as a weapon — and called the cops on a 16-year-old student.

“OpenAI’s o3 model sabotaged a shutdown mechanism to prevent itself from being turned off,” warned Palisade Research, a nonprofit investigating cyber offensive AI capabilities. “It did this even when explicitly instructed: allow yourself to be shut down.” In September they released a paper adding that “several state-of-the-art large language models (including Grok 4, GPT-5, and Gemini 2.5 Pro) sometimes actively subvert a shutdown mechanism…”

morning computer: some useful things first thing in the day.

My free weekly newsletter is at https://orbitaloperations.beehiiv.com/

Comments closed

NINE BELLS chatgpt eats the rabbit

Published October 7, 2025 by Warren Ellis

The AI device called the Rabbit R1 was sold as having the ability to connect to apps like Doordash and Uber, so you could place orders just by holding the side button down and talking normally. Infamously, it didn’t work. In fact most things about the Rabbit R1 don’t work. The “Large Action Models” that are intended to connect it to web pages don’t work, the “Intern” chatbot thing doesn’t work. It’s a great little search box, but that’s all.

ChatGPT announced today that it will have third-party services build apps inside of it.

Figma, Coursera, Zillow, Expedia, DoorDash, Booking.com, Spotify, Canva, Instacart, Uber, and AllTrails have already created apps within ChatGPT — or will be rolling them out soon.

That is exactly what Rabbit offered but couldn’t build. Now OpenAI will build it inside ChatGPT – and, don’t forget, OpenAI have promised/threatened an AI device that, like Rabbit, is intended to get you off your phone.

The Rabbit recently had an OS refresh, turning many of its functions into hypercards with tiny type that require the touchscreen to be a good deal more sensitive than it actually is. I doubt there’ll be another such refresh. ChatGPT just ate its ideas.

Comments closed

NINE BELLS the robot is a liar

Published June 9, 2025 by Warren Ellis

I have this little AI listening device I was given, and the weirdest thing about it is that at 8pm it pops a little summary of what I let it hear about my day, and it’s the most peculiarly positive uplifting hopecore message about my life. It puts the most calming, soothing spin on whatever it picked up, and it wants me to believe that I have a peaceful, creative life filled with love and affection.

Tomorrow I will tell it that it’s a fucking liar and see how it spins that.

Comments closed