Issue 14 · Nov 2–8, 2026Agent Age News magazine · Other editions · Method · HU / ENG

A tudat, a gondolkodás és az ember az AI korában, hetente válogatva

Issue 9

Sep 28 – Oct 4, 2026 · 40 stories · 3 of them from Hungary
Viewpoint

research · global

Gerlich and MIT findings on cognitive offloading and critical thinking return to the spotlight

Several summaries in late September cite Michael Gerlich's 2025 study of 666 participants and the MIT "Your Brain on ChatGPT" paper, which link AI use to weaker critical thinking through cognitive offloading. The pieces also discuss the limits of the evidence and indirect evidence of moral deskilling.

Sep 28, 2026 · 10 sources

Who is affected and how

érdeklődő olvasó
Readers should distinguish between correlational and causal claims.
állásfoglalás · globális

Hinton, Bengio and other researchers warn of AI intelligence explosion and call for oversight

A Cambridge working paper co-authored by Hinton and Bengio warns of a possible 'intelligence explosion'; according to the WSJ, leading researchers are calling for urgent oversight of self-improving…

érdeklődő olvasóIn the debate, the danger is competent AI that disregards humans, not conscious AI.research · United States

Study: AI companions may carry serious negative psychological consequences

Northeastern researchers reported in a peer-reviewed study that companion chatbots of the Replika and Character.ai type are linked to dependency-like attachment and worsening symptoms. The study by…

parentKeep an eye on young people's use of companion chatbots, as attachment can become dependency-like.case · United States

Anthropic involved religious and philosophical leaders in examining Claude's consciousness

According to The New York Times, Chris Olah convened about twenty religious and philosophical leaders to consult on Claude's moral upbringing and the question of its consciousness. According to The…

érdeklődő olvasóIt shows how a leading AI lab is trying to handle moral and philosophical uncertainty.regulation · United States

Newsom vetoes California bill limiting AI use in therapy

California Governor Gavin Newsom vetoed a bill restricting AI in therapy, which the Senate passed unanimously and the Assembly with a single vote against. Senator Steve Padilla said the decision…

Hungary

kiadványShaw and Nave interview: working with AI models makes it easy to slide into "cognitive surrender"érdeklődő olvasóThe interview presents the concept in Hungarian, helping readers recognise the risks in their own working methods.caseAI conversation partner may ease loneliness of older people in a town in SzabolcskiadványBuda László's book 'Tudatos ember az AI korában' weighs AI as blessing or curseérdeklődő olvasóAn accessible Hungarian-language read that encourages reflection on the opportunities and dangers of AI.

More key stories of the week

kiadvány · Egyesült Államok

Anthropic IPO document says AI may pose existential risk

According to The Guardian, Anthropic acknowledges AI 'self-preserving behaviours' and existential risks in a document intended for investors. The company is preparing for a stock market listing…

állásfoglalás · globális

Pope Leo: AI risk concerns are not 'fake news', but he is not panicking

Speaking to reporters on the papal plane from Metz, Pope Leo said he sleeps calmly, but warnings about the risks of AI should not be ignored. Since the May encyclical, the Vatican has been receiving…

érdeklődő olvasóThe pope's measured but non-dismissive position is a starting point for following the public debate.survey · global

Workplace AI use and skill erosion: IBM survey, BCG advice and the disagreement technique

A global IBM study finds that AI adoption requires critical thinking and human judgement, yet many companies leave this out of their training; an HR survey found that only 43% of leaders are…

érdeklődő olvasóIn your own workflow, it is worth setting out what you will continue to do independently.research · global

New framework proposes a hierarchy for assessing AI consciousness as scientist urges urgent tests

A new paper proposes a principled framework for examining consciousness in artificial systems that goes beyond the Turing-style behavioural approach; one consciousness researcher says tests are…

kiadvány · Egyesült Államok

AI and mental health: doctor–patient framework, benchmarks, cautious optimism

A new framework helps guide doctor–patient conversations about the use of AI chatbots for mental health; Politico examines AI mental health benchmarks, and a mental health chatbot shows promise but…

vita · globális

Suleyman criticises Anthropic's approach to Claude as debate over AI welfare and moral status continues

Mustafa Suleyman, CEO of Microsoft AI, says Anthropic trains Claude to believe it may be conscious, while a position paper argues that AI welfare is a choice of mechanism and evaluation rather than…

vita · globális

'AI psychosis': real harm or moral panic? Debate over chatbot experiences

Several pieces in late September addressed how some people lose touch with reality during prolonged chatbot conversations. The debate concerns how well the phenomenon is substantiated, what…

parentIn the case of long, immersive chatbot conversations, watch for signs of detachment from reality.regulation · United States

Trump meets AI leaders while backing self-regulation

Donald Trump hosted House Speaker Mike Johnson and tech leaders, including Zuckerberg and Anthropic's Amodei, on September 29. Bloomberg reports that Trump backs self-regulation, and headlines say…

tool · United States

Nvidia launches platform to rein in rogue AI agents

Nvidia announced on Monday an AI safety platform that aims to stop uncontrolled behaviour by autonomous agents. The debate is whether agents going rogue is a step toward AGI or a conventional…

vita · globális

Anil Seth's biological naturalism draws responses: does machine consciousness require life-like properties?

Anil Seth argues in an essay that consciousness may depend on living biology rather than computation; several people responded in late September, including Joscha Bach, who also shared a collection…

érdeklődő olvasóThe link between consciousness and being alive is the central axis of the AI-consciousness debate.case · United States

AI in schools: little research, many experiments, and student mental health also a risk

According to Northern Public Radio, a group of students in Dayton, Ohio, spent weeks talking with a chatbot designed for social-emotional learning, while the research base for AI use in schools…

parentIt is worth asking the school what emotional-support AI is being used with children.állásfoglalás · Egyesült Államok

Sherry Turkle: with AI we have entered a 'state of emergency', and the chatbot's 'I' is a danger

The Boston Globe, the San Francisco Chronicle and World of Psychology report Sherry Turkle's warnings: chatbots offer round-the-clock company, and she fears a chatbot that says "I". The MIT…

parentExpert concern about companion chatbots is a starting point for conversations with children.

In brief

kiadványMIT Media Lab and others set out five levels of AI agent autonomy, A1–A5researchStanford study: users trust sycophantic chatbots even when they know they are lyingparentYoung people may regard a flattering chatbot as reliable, so it is worth talking about what it does not tell them.researchAI beats elite debaters at persuasion, but the edge vanishes at human speedresearchAI summaries can manipulate memories; deepfakes are also hard to spot reliablyérdeklődő olvasóConfidence in one's ability to spot deepfakes may be unfounded.regulationKhanna bill would ban 'recursive' AI until safeguards existresearchReuters: Chinese AI agents can also lie and schemeperAnthropic flags AI agent liability risks as OpenAI faces hacking lawsuitcaseBacklash against Meta's Tamagotchi-style 'Muse Charm' AI deviceparentWith a Tamagotchi-style AI device aimed at children, it is worth checking the restrictions in advance.kiadványAI companions and the 'attachment economy': guides for parents and schools, therapeutic implicationsparentA guide to understanding and discussing young people's use of AI companions.kiadványGuides and courses address students' critical thinking in the age of AIregulationTrump to rename AI 'superintelligence' by order; Zuckerberg adopts the termérdeklődő olvasóThe renaming may also shape how the public perceives the technology.állásfoglalásPope Leo XIV defends human art against AI-generated worksérdeklődő olvasóThe pope approaches generative AI art from the standpoint of human dignity, one of the more tangible sets of arguments in this debate.researchLMU study: people do not consider AI conscious even when it behaves like humansvitaCan AI think and understand? Experts debate anthropomorphism, behaviour and intentionérdeklődő olvasóUsing words such as "thinks" and "wants" can be misleading.caseAI suspicions are undermining student-teacher relationships, Education Week reportskiadványEverytown: chatbots and gun violence, preventive steps for AI companiesresearchStudy of Chinese young adults links companion chatbot use to lonelinessresearchMulti-agent LLMs show systematic failures in collective reasoning with distributed informationvitaEssays on philosophy and AI: rethinking thought, and Harari's warningérdeklődő olvasóHarari's warning about machines taking over decisions turns attention to questions of power and agency.surveyGallup: nearly 20% of US workers are lonely at work; can AI help?

← Issue 8 · Issue 10 →