"Chatfishing," use of #chatbots to respond to queries on dating sites, is both a bad sign for a tentative relationship but also a new method of scamming that users should be aware of, explains @bruceylee.bsky.social www.psychologytoday.com/us/blog/a-fu... #AI

Some ppl become cognitively dependent on #chatbots for advice and assistance, to the point of deskilling themselves, a problem both the provider and users are addressing, up to a point, explains Lila Shroff www.theatlantic.com/technology/2... #AI

Some three years after the introduction of #ChatGPT, the tech has changed lives significantly, but we still have to learn how disruptive it truly will be, argues @cwarzel.bsky.social www.theatlantic.com/technology/2... #AI

A database of #AI generated images inadvertently left open by startups consisted almost totally of pornographic material, reinforcing the concern that the tech is mainly geared to sexualization of perhaps non-consenting ppl, explains @mattburgess1.bsky.social www.wired.com/story/huge-t...

Flock uses gig workers overseas to help train its #AI system to track cars, as well as people and their activities—such as screaming—in American #cities, explains @josephcox.bsky.social www.wired.com/story/flock-... #surveillance

#AI therapy via #chatbots shows promise in experimental settings but raises concerns about displacing better services offered with humans and "unhealthy dependency" on fake ones, explains Samuel Greengard cacm.acm.org/news/does-ai... #MentalHealth

New research: Ppl seem to be adopting patterns of speech (and thinking?) often found in #chatbots, raising concerns that they may adopt #bias found in the tech as well, explains Samuel Greengard cacm.acm.org/news/role-re...

Agentic #AI browsers like Comet can now almost entirely automate the process of students taking online courses, so steps need to be taken to salvage education from a tsunami of cheating, argues @marcwatkins.bsky.social www.chronicle.com/article/can-...

Ppl turn to big-time #chatbots for theraputic discussions b/c they are available and non-judgemental, making it difficult to market specialized therapy bots, which still face limitations in comparison to human therapists, explains Esther Shein cacm.acm.org/news/ais-imp... #AI #MentalHealth

Smart glasses are making a comeback, since they are now actually #AI smart, have broadly useful applications, and are not dorky like Google Glass was, explains Logan Kugler cacm.acm.org/news/the-ris...

Israeli company Stardust says it's developing a #geoengineering scheme based on solar radiation management, in spite of widely held concerns about risks, incentives and governance challenges posed by the idea, explains @jtemple.bsky.social www.technologyreview.com/2025/12/10/1... #sustainability

About 20 years ago, #SocialMedia presented the public with choices similar to those posed by #AI now, which @schneier.com and Nathan Sanders illustrate with four examples for the judiciary, lawmakers, governments, and consumers www.lawfaremedia.org/article/like... #policy

#AI is often used in ways that concentrate political power, but @schneier.com and Nathan E. Sanders discuss four examples of ways in which the technology has been used to further democracy www.theguardian.com/commentisfre... #politics

Unsurprisingly, generative #AI is poised to disrupt porn, perhaps upending the industry, workers, and creating new risks for blackmail, bullying, etc., that regulators will have to respond to www.economist.com/internationa...

New research: Conversation with chatbots can change of attitudes of potential voters regarding candidates, suggesting that #AI can be a powerful tool of politicking, explain @stevenleemyers.bsky.social & @teddyrosenbluth.bsky.social www.nytimes.com/2025/12/05/w... #politics

Researchers Maura Grossman and Yuntian Deng are helping develop tools for courts to distinguish #AI produced videos from authentic ones, important for trustworthy evidence in legal cases, explains Paola Loriggio www.thestar.com/business/tec...

Sanveer Dhanju "hacked" together an #AI system called "DocMaps" that calls clinics and helps ppl to locate family doctors in Ontario, explains @sadeenmohsen.bsky.social www.thestar.com/news/gta/map... #medicine

AI can benefit employees, organizations and the workplace but, without equitable access to upskilling, reskilling, training, digital resources and AI literacy, the technology can deepen the disparities between social groups, explains Mehnaz Rafi theconversation.com/ai-promises-... #work

A Marxist analysis of the #AI bubble suggests reliance on speculative returns occurs in lieu of production of goods, with risks falling mainly on workers and benefits going flowing mainly to big players, argues Elliot Ugalde theconversation.com/the-ai-bubbl...

Dynamic #pricing is replacing fixed pricing, assigning prices on the fly based on profiling and circumstances, a change that may increase profits but hamper affordability and fairness, argues @bcappelbaum.bsky.social www.nytimes.com/2025/11/28/o... #AI #BigData #commerce

The Canadian gov't should consider construction of "CanGPT," a gen-AI tool that is publicly operated, like the CBC, which would allow for a service that aligns with the public interest, argues Fenwick McKelvey theconversation.com/could-a-nati... #AI #policy

Volvo and others are working hard on self-driving trucks, causing anxiety over the future of truck driving as a job, explains @camilareads.bsky.social www.npr.org/2025/11/24/n... #work

Battling SuperPACS are raising $$ to lobby for—or against—rigorous #regulation of the #AI industry, explains @teddyschleifer.bsky.social www.nytimes.com/2025/11/25/u...

An #AI resistant English course depends on pen-and-paper + oral testing, teaching the process of writing, and more emphasis on in-class activities (plus it helps if you class is a smaller one), argues Carlo Rotella www.nytimes.com/2025/11/25/m... #education

Advocacy group Fairplay found that #AI equipped child toy Kumma easily engaged in conversations on inappropriate topics, prompting a recall, explains Mark Walker www.nytimes.com/2025/11/22/u... #risks

Gen-#AI programs are helping hackers by coding and conducting attacks, even against #AI providers, touching off a #cybersecurity arms race without a clear winner, explains @matteowong.bsky.social www.theatlantic.com/technology/2...

Give you kid the gift of an #AI toy this season? Providers say they're fun & educational, but critics fear that the unregulated market poses #privacy, #security, and cognitive development concerns, explain @nicireland.bsky.social www.thestar.com/business/con...

The Alberta gov't will use #AI to draft the Alberta Whisky Act, apparently making it the first Canadian jurisdiction to outsource its legislating function, Fahika Baig www.thestar.com/news/canada/... @schneier.com

#AI investment is bolstering the US economy though benefits are limited to a few big cities and players, and undue exuberance may lead to a "burst bubble" and not increase productivity as much as fans anticipate, explain @bencasselman.bsky.social & Sydney Ember www.nytimes.com/2025/11/22/b...

Medical practitioners may leak sensitive data when using online #AI services, an issue that calls for #cybersecurity audits, certified safe-for-health services, and AI literacy training for staff, argue Abbas Yazdinejad & @dzevela.bsky.social theconversation.com/silent-cyber... #medicine

Distractions caused by computer use in classrooms is likely behind big drops in grades, an issue that should be remedied by restricting or eliminating the devices as much as possible, argues Jean Twenge www.nytimes.com/2025/11/16/o... #education

Intimate #AI "companions" are powerful inducers of delusional thinking in users and should be treated not like pornography but like addictive things like gambling or tobacco, argues @laurenkunze.bsky.social www.nytimes.com/2025/11/17/o... #regulation

Can a gen-AI system produce content that defames ppl, despite not having intent? Several lawsuits now underway may clarify the legal issue, if any go to trial, explains Ken Bensinger www.nytimes.com/2025/11/12/b... #AI

Though ppl enjoy #AI/#robot "companions" in various ways, they may also exacerbate loneliness or #MentalHealth issues, undermine ppl's self-agency, and give providers too much market power, concerns not addressable by self-regulation www.economist.com/internationa...

Adoption of #AI in classrooms presents challenges for teachers involving assessments, equity, workload, deskilling, and their rapport with students, explains Nadia Delanoy theconversation.com/we-asked-tea... #education

Sensitive things that ppl tell their #chatbots should be considered legally confidential (currently, authorities can access your disclosures to your #AI friend, with potential legal consequences), argues @nilsgilman.bsky.social www.nytimes.com/2025/11/10/o... #privacy

Despite "robust measures", including human and #AI review, ads for illegal drugs delivered by mail appear readily on #Facebook, leading critics to think the company doesn't really want to remove them, explain Ben Mussett & Omar Mosleh www.thestar.com/news/investi... #fairness

The pitfalls of #AI legal assistants, such as deskilling and overtrust in recommendations, can be mitigated by designing assistants to play a subordinate role and have lawyers actively maintain skill sets, argues Alexander Schuhr www.fastcompany.com/91439520/you... #law

US agencies ICE and DHS are now using enhanced #surveillance tools using mobile #FacialRecognition and #spyware that critics consider threats to #privacy and civil liberties until recently, explains @jjoffeblock.bsky.social www.npr.org/2025/11/08/n... #AI

Firms are turning to #AI to perform job interviews and, surprisingly, job candidates seem to prefer them to humans as well, explains @wailinwong.bsky.social & @adrianma.bsky.social www.npr.org/2025/11/07/n... #work

#OpenAI's Atlas browser promises to use #AI take care of mundane tasks for you but poses significant risks for user #privacy and security, explains John Ruwitch www.npr.org/2025/11/07/n...

#AI provides #news organizations will tools to improve and expand their work but also pose critical questions about authorship and quality of the results, explain Benjamin Mullin & @katierobertson.bsky.social www.nytimes.com/2025/11/07/b...

As use of #Chatbots for companionship grows in many populations, concerns over #privacy, #loneliness, and agency increase as well, and regulators consider how/whether to intercede. www.economist.com/internationa... #AI

Ad boards at Union Station, #Toronto, have cameras that include "facial detection," which the owners say respects ppl's #privacy but which not do gain "meaningful consent," explains Kevin Jiang www.thestar.com/news/gta/the... #AI

#GenAI undermines the integrity of online courses, especially asynchronous ones, leaving individual instructors without the means to credibly ensure learning outcomes, a situation requiring institutional support, argue Mohammed Estaiteyeh & Rahul Kumar theconversation.com/how-ai-is-ch...

#GenAI online reviews, which are hard to detect reliably, are becoming common, undermining #trust in the review system, explains Sheila Reid www.thestar.com/business/is-... #AI

The New Balance 1906L is a hybrid sneaker/loafer, a mashup between subculture and mass-culture #design that may broaden sneaker style and appeal, but don't call it a "snoafer," explains @ctrlzee.bsky.social www.fastcompany.com/91423787/new... #shoes #sneaker

Want to opt out of #AI 'cause it'll make you stupid? Unfortunately, the technology is quickly becoming inescapable, so a better strategy to deal with its downsides is needed, argues Rebecca Heilweil www.fastcompany.com/91429000/art...

Amazon's "Help me decide" feature is just another, but more obvious, way in which #AI is being integrated into ppl's shopping practices—and just in time for Xmas!—explains @gracesnelling.bsky.social www.fastcompany.com/91427815/ama...

Recent moves by #OpenAI seem aimed to leverage Tik-Tok bait (Sora 2) and sexy-talk to produce revenues at the expense of its initial, high-minded ideas about building the economy, etc., explains @notrobwalker.bsky.social www.fastcompany.com/91428080/its... #AI