Large language model AIs might seem smart on a surface level but they struggle to actually understand the real world and model it accurately, a new study finds.

AI host discovers its artificial nature, sparking debate on AI sentience and the blurred lines between human and machine.

Scientists taught rats to drive to a certain destination, but the rodents took a detour, suggesting they enjoy both the journey and the rewarding destination.

AFP video: https://www.youtube.com/watch?v=08G8u7sk2Jo

Meta could bring ads to Threads as soon as next year, according to a report from The Information. As part of its plan, Threads will reportedly allow a small number of advertisers to make and publish ads in January.

That tracks with what my colleague Alex Heath reported about the rollout of ads in July. Instagram head Adam Mosseri has also confirmed that Meta is “definitely” planning to bring ads to Threads. “I get why people have concerns, but at the end of the day we’re a business and Threads needs to make enough money to pay for the people and servers that it takes to run the service and provide it to people for free,” Mosseri said at the time.

Wow! All I can say is wow!

Bluesky’s latest signup surge continues.

While it might feel a little tangential to bring technology into this, everybody is affected by the growth-at-all-costs Rot Economy, because everybody is using technology, all the time, and the technology in question is getting worse. This election cycle saw more than 25 billion text messages sent to potential voters, and seemingly every website was crammed full of random election advertising.

Our phones are beset with notifications trying to "growth-hack" us into doing things that companies want, our apps full of microtransactions, our websites slower and harder-to-use with endless demands of our emails and our phone numbers and the need to log back in because they couldn't possibly lose a dollar to somebody who dared to consume their content for free. Our social networks are so algorithmically charged that they barely show us the things we want them to anymore, with executives dedicated to filling our feeds with AI-generated slop because despite being the customer, we are also the revenue mechanism. Our search engines do less as a means of making us use them more, our dating apps have become vehicles for private equity to add a toll to falling in love, our video games are constantly nagging us to give them more money, and despite it costing money and being attached to our account, we don't actually own any of the streaming media we purchase. We're drowning in spam — both in our emails and on our phones — and at this point in our lives we're probably agreed to 3 million pages worth of privacy policies allowing companies to use our information as they see fit.

And these are issues that hit everything we do, all the time, constantly, unrelentingly. Technology is our lives now. We wake up, we use our phone, we check our texts (three spam calls, two spam texts), we look at our bank balance (two-factor authentication check), we read the news (a quarter of the page is blocked by an advertisement asking for our email that's deliberately built to hide the button to get rid of it, or a login screen because we got logged out somehow), we check social media (after being shown an ad every two clicks), and then we log onto Slack (and feel a pang of anxiety as 15 different notifications appear).

Modern existence has become engulfed in sludge, the institutions that exist to cut through it bouncing between the ignorance of their masters and a misplaced duty in objectivity, our mechanisms for exploring and enjoying the world interfered with by powerful forces that are too-often left unchecked. Opening our devices is willfully subjecting us to attack after attack from applications, websites and devices that are built to make us do things rather than operate with the dignity and freedom that much of the internet was founded upon.

The Danish welfare authority, Udbetaling Danmark (UDK), risks discriminating against people with disabilities, low-income individuals, migrants, refugees, and marginalized racial groups through its use of artificial intelligence (AI) tools to flag individuals for social benefits fraud investigations, Amnesty International said today in a new report. 

The report, Coded Injustice: Surveillance and Discrimination in Denmark’s Automated, details how the sweeping use of fraud detection algorithms, paired with mass surveillance practices, has led people to unwillingly –or even unknowingly– forfeit their right to privacy, and created an atmosphere of fear.

“People in non-traditional living arrangements — such as those with disabilities who are married but who live apart due to their disabilities; older people in relationships who live apart; or those living in a multi-generational household, a common arrangement in migrant communities — are all at risk of being targeted by the Really Single algorithm for further investigation into social benefits fraud,” said Hellen Mukiri-Smith.

UDK and ATP also use inputs related to “foreign affiliation” in its algorithmic models. (...) The research finds that this approach discriminates against people based on factors such as national origin and migration status.

Amnesty International also urges the European Commission to clarify, in its AI Act guidance, which AI practices count as social scoring, addressing concerns [raised by civil society](https://www.hrw.org/news/2023/10/09/eu-artificial-intelligence-regulation-should-ban-social-scoring#%3A%7E%3Atext=%28Brussels%2C+October+9%2C+2023%2Cregulation%27s+prohibition+on+social+scoring.%5D%22+said+HMS.%29.

A $82,000 Tesla Cybertruck broke down in Seattle and became an internet sensation, but the abandoned EV was towed away.

Businesses, community groups, arts organisations and residents in a thriving town on the Croydon-Surrey border are in uproar because the automatic systems employed to police social media have silenced them on one of the world’s biggest digital platforms – all because Coulsdon has the letters L, S and D in its name.

Residents’ associations and businesses with “Coulsdon” in their titles have found themselves “cancelled”, with posts being removed from Facebook and warnings issued as to their future conduct under a set of rules so vague that any post, however innocent, might fall foul of them.