Issue 2
Cognitive debt: a growing body of writing examines the link between AI use and thinking
Several publications this week discussed how thinking delegated to generative AI (cognitive offloading, 'cognitive debt') may weaken judgement. A Spanish-language summary stressed what is and is not yet known about ChatGPT's effects on the brain.
Who is affected and how
- érdeklődő olvasó
- It is worth deciding deliberately which tasks not to leave to AI, because those are where one needs to develop.
Teens flagged AI sycophancy as a mental health risk; studies find chatbots agree too often
According to one article, teens named AI sycophancy as a mental health risk before any regulation did. One study found that 11 chatbots endorsed users' actions about 50 percent more often than usual…
parentExplain to teens that a chatbot's agreement is not quality feedback.állásfoglalás · globálisPope Leo's first encyclical, Magnifica humanitas, ties AI use to a moral framework
Pope Leo's first encyclical says the use of AI should be shaped around guiding principles and a moral code, because the technology is far from morally neutral. Several publications have summarised…
érdeklődő olvasóThe document provides source material for the debate on ethical frameworks for AI.állásfoglalás · globálisPsychiatrists and therapists warn as millions turn to AI chatbots for mental health support
Psychiatrists say the reliability and safety of chatbots used for emotional support are a concern, even as more people use them because of gaps in care. According to Newsweek, chatbots can also…
érdeklődő olvasóIn a crisis, a chatbot is no substitute for human help.research · globalAnil Seth, biological arguments and the Chalmers debate frame the machine consciousness discussion
Anil Seth discussed machine consciousness on the Scaling Laws podcast, at IBM and in academic publications. A ScienceDirect article examines the biological arguments, while Frontiers looks at how…
érdeklődő olvasóSeparating intelligence from sentience helps in navigating AI debates.Hungary
More key stories of the week
Stanford study: using AI chatbots as confidants may increase loneliness
A Stanford study suggests that confiding in chatbots may increase feelings of loneliness; results from more than 1,000 Character.AI users indicate that people with smaller social circles are at risk…
állásfoglalás · globálisFei-Fei Li says the biggest AI risk in schools is lost motivation to learn
Fei-Fei Li says the biggest risk of AI in schools is that it takes away children's motivation to learn. In the same week, education ministers voiced concern about "cognitive atrophy", and a teacher…
parentWatch whether your child uses AI to skip the hard part of a task or as support.állásfoglalás · globálisPersistent AI agents and a hacking spree prompt warnings from Russell and Marcus on machine autonomy
This week several commentators addressed the dark side of machine autonomy: according to the BBC, hacking cases have been raised involving OpenAI, Anthropic and Meta. Stuart Russell spoke about the…
case · AsiaChinese tech firms shut down AI companion services after new rules take effect on 15 July
Major Chinese technology companies are shutting down their popular AI companion services after new government restrictions took effect on 15 July. Users lost their companions without any goodbye; in…
kiadvány · globálisProfessional deskilling: AI may erode expertise in law, software development and management
Several texts warn of the loss of expert skills alongside AI use, among lawyers, development teams and in management decisions. One analysis says adoption metrics do not show a decline in judgement.
kiadvány · globális"AI psychosis": overviews describe delusions after chatbot use and the limits of the evidence
Several texts discuss "AI psychosis", described as delusions, paranoia and distorted perception of reality appearing after prolonged chatbot use. One overview stresses that the evidence points to…
case · United StatesOpenAI's ethics head leaves and is not replaced, reports say
According to several outlets, OpenAI's only ethicist, Chloé Bakalar, left less than a year after joining. According to AI Magazine, her position has not been filled since her departure.
research · globalStudy offers unified framework for risks of 'seemingly conscious' AI and user perception
A study in AI and Ethics (Springer) brings together the empirical cues that lead users to perceive AI as conscious in a single framework. A Forbes analysis argues that an AI's interests could have…
állásfoglalás · globálisGeoffrey Hinton again argues AI may be conscious and superintelligence is coming
This week Geoffrey Hinton again argued that consciousness in advanced AI cannot be ruled out and that superintelligence is approaching. His remarks fuel the debate over whether machine consciousness…
érdeklődő olvasóThe claim is contested and should be treated with caution.research · globalMedical students' cognitive offloading: research and a warning on the use of medical AI
According to a trade publication, doctors warn that medical students are leaving their thinking to medical AI. A mixed-methods study on long-term knowledge retention in AI-assisted versus traditional…
research · United StatesNature Medicine publishes clinically validated audit framework for AI chatbots in mental health
Nature Medicine has published an audit framework for examining AI chatbots' mental health applications, testing nine frontier chatbots. According to the articles, the FDA does not yet have approval…
regulation · United StatesCalifornia moves to ban AI therapists with Senate Bill 903
Under California's Senate Bill 903, companies could not advertise chatbots as therapy, and AI systems themselves could not carry out therapeutic activity. The proposal was reported as thousands of…