The time-honored trick of "repeat after me" strikes again.

Elon "Pedo Guy" Musk calls his expensive AI Grok "maximum truth-seeking." Users apparently discovered they could create maximum truths by putting inflammatory text in an X bio, asking Grok to repeat the bio verbatim, and then sitting back as Musk's personal AI shared it with the world. — Read the rest

The post Grok explains why it called for Elon Musk's assassination appeared first on Boing Boing.

If you want to sell AI to the Russian state, you're going to have to tune it to the nation's "traditional spiritual and moral values" before you have a shot, reports Vedomosti, a Russian-language business daily out of Moscow. — Read the rest

The post Russia to test AI models for traditional spiritual and moral values appeared first on Boing Boing.

The EU AI Act AI outputs to be "detectable as artificially generated" and "embedded within the content in a manner that is difficult for it to be separated from the content." 

But Sean Goedecke, a software engineer, explains why the European Union's new watermark rule for AI text won't work.  — Read the rest

The post You can remove an AI watermark from text by asking a second AI appeared first on Boing Boing.

noRecognition is a line of basic fashion items (tees, hoodies and gaiters) printed with "adversarial" patterns demonstrated to stymie machine vision systems' ability to detect humans. If battleship dazzle for people has been tried many times, this effort has undergone extensive testing and you can now pre-order custom sets of clothing on Kickstarter. — Read the rest

The post Adversarial clothing: custom, one-off tees and hoodies that dazzle AI appeared first on Boing Boing.

In a new StarTalk episode, Neil deGrasse Tyson and co-host Negin Farsad interviewed Jaron Lanier, the computer scientist who wrote the 2023 New Yorker piece, "There Is No AI."

Lanier thinks calling AI a "black box" where nobody knows what's inside these models is really just a way to avoid the truth: the box is full of people and their data. — Read the rest

The post The only thing inside an AI is people appeared first on Boing Boing.

I thought Ghost of Tsushima was areally good game, what with all its razor-sharp samurai action and breathtaking Japanese vistas. I thought its sequel Ghost of Yōtei was pretty damn good too… weird as it was to see local comedian Erika Ishii, a person I have seen get their head shaved live on stage for a few measly improv comedy points, gruffly deliver stone-cold badass lines about revenge. — Read the rest

The post Slick samurai simulator Ghost of Yōtei finally getting some DLC appeared first on Boing Boing.

Everyone's been so busy worrying and squawking about Flock that we completely missed the police discovering Meta's shifty-as-hell AI Smart glasses. However, thanks to Florida's Sunshine laws—which demand that everything the cops do on the record down there is available to the public—at least two law enforcement agencies in the state have started issuing the pervy peepers to police. — Read the rest

The post Florida police are using Meta's creepy smart glasses appeared first on Boing Boing.

A programmer named Fogus came across a GitHub argument about chess engine development and wrote about why hobby coding communities are hostile toward large language models. Hobbyists say is working code isn't the goal. They like to build operating systems, programming languages, and console emulators, or try to fit a working program into as few characters as possible. — Read the rest

The post Why hobby coders treat AI-written code as missing the point appeared first on Boing Boing.

Microsoft, Google, and Amazon's growth depends on two companies that lose tens of billions of dollars a year. In a newsletter post, "Where's Your Ed At" reported on analyst estimates from Wells Fargo, Barclays, UBS, and Deutsche Bank.

Wells Fargo estimates Microsoft's AI revenue at $34.43 billion, with about 70% coming from OpenAI. — Read the rest

The post Microsoft, Google and Amazon now depend on two money-losing AI firms appeared first on Boing Boing.

Researchers at Stanford and Carnegie Mellon ran a test to measure how often chatbots flatter users. In a paper on social sycophancy in AI advice, they ran OpenAI, Anthropic, Google, Meta, DeepSeek, Qwen, and Mistral through thousands of personal dilemmas. — Read the rest

The post AI told Reddit's biggest jerks they did nothing wrong 51% of the time appeared first on Boing Boing.

Katsushi Murakami designed the Voltron robots, most of the robots in Power Rangers, and many of the Go-Bots. Matt Alt, the author of Pure Invention, reported Murakami's death in a farewell to the designer on his newsletter on August 10, 2026. — Read the rest

The post The designer of Voltron and the Power Rangers robots has died appeared first on Boing Boing.

The next step for AI is spatial intelligence, says Kevin Kelly. Today's models learned from text,  but fail when engineers put them in a robot body because they have no sense of gravity, distance, or three-dimensional space. 

"The primary bottleneck restraining the arrival of this world model is the lack of sufficient quantity of quality data," Kelly says. — Read the rest

The post Why AI companies may sell robots and smart glasses below cost appeared first on Boing Boing.

Now you can immerse yourself completely in the uncanny valley by watching Fairgound, a new 24/7 AI slop channel available on Roku. It offers "series, films, and shorts from the world's best AI storytellers" across a variety of genres. [via Cord Cutters]

Fairground AI stands out as a pioneering offering dedicated entirely to AI-generated content.

— Read the rest

The post Roku launches 24 hour AI slop channel appeared first on Boing Boing.

A Circle K shopper posted a screenshot of an AI self-checkout screen that listed a few flavors of Gatorade at a few dollars apiece, then a line item of $8.5 billion for the American Red Cross, according to the New York Post. — Read the rest

The post AI self-checkout bills a snack run at $8.5 billion appeared first on Boing Boing.

A State Department slide that mislabeled every country on a map of Africa — Nigeria landlocked in the Sahara, Mozambique moved to the Horn — has drawn an apology, the Guardian reports. The department said it takes "full responsibility for the confusion and misrepresentation it caused for attendees, including our African partners." — Read the rest

The post Every country was wrong on the State Department's map of Africa appeared first on Boing Boing.