Spring til indhold
Hub Nexus

Lentis, AI: More Human Than You Think

Version1
Opdateret

ForfatterIngen forfatter endnuOvertag den

Ser du noget at forbedre? Foreslå en ændring.

Støtte

Denne side er offentlig.

I Lentis · side 164 af 351

LLMs are often introduced as simple chatbots that support basic Q&A, yet their behavior feels closer to a flexible cognitive being. They adapt to the current context, track intent across conversations, and respond to problems in ways that resemble a human partner. This makes the interaction feel more personable, like collaborating with an entity that "understands" you. AI companies know this well, which is why a whole market has emerged from LLM companionships. Their eerily similar resemblance to human emotions can rapidly deteriorate one's mental health if misused. This makes LLMs a particularly dangerous psychological phenomenon if abused by the wrong hands. That notion becomes especially clear when examining how AI is actually used in practice below.

AI as General Purpose Tools

Many people use large language models (LLMs) and other artificial intelligence tools to complete specific, often mundane tasks. This group represents the majority of AI use-cases. Common examples include code writing, debugging, boilerplate generation, vacation planning, and general organization. For these users, the agenda is largely pragmatic: the goal is to accomplish an often undesirable or repetitive task and “get it out of the way” as efficiently as possible. Empirical findings reflect this pattern of use. One research study reports that up to 75% of developers use AI in day-to-day activities[1], while a large dataset of ChatGPT interactions suggests that approximately three out of every four prompts is centered on practical “doing” tasks[2]. The rate of growth that LLMs experienced has been unprecedented. When ChatGPT was released at the end of 2022, it became the fastest growing service in all of human history, taking less than two months to reach 100 million active users.  A large majority of this user base prioritizes efficiency and throughput over process. AI tools are valued primarily for their ability to reduce cognitive load and time expenditure, making it particularly popular in the workplace.

AI in the Workplace

Many participants in this group also advocate for equal and open access to LLMs in workplace settings, often treating such tools as essential for productivity. Their strategy for advancing this agenda frequently involves coalition building and rallying fellow employees to support broader institutional access. This utilitarian framing of AI as a general-purpose tool stands in contrast to other emerging uses—such as relational or expressive applications—and highlights a broader juxtaposition that this article examines. As a result, discussions around AI adoption frequently intersect with institutional policy, resource allocation, and evolving expectations about baseline digital literacy in professional settings.

AI as Emotional Companions

A major participant group includes users who use AI to aid their mental health. In a survey of 1,000 adults ranging from Gen Z to Boomers, roughly 23% reported using AI for mental health support. And there is a clear trend as we look at younger demographics. Usage climbs from just 5% of Boomers to 18% of Gen X, 31% of Millennials, and 44% of Gen Z[3]. Among these chatbot users, 41% use them because they are non-judgmental and 47% use them to reframe and understand their current feelings[3]. This trend is mirrored in teens as well. While 58% of them engage with AI for entertainment or out of curiosity, a non-insignificant number turn to it for mental health reasons, seeking advice and an honest, non-judgmental environment among other mental health reasons[4].

Similarly, another large participant group includes those who use AI for deeper companionship, emulating the dynamics of a human relationship. This is a booming sector, with AI companion apps seeing over 220 million downloads and generating upwards of $221 million in revenue5]. Popular platforms like [Character AI and Replika are some of largest providers[6]. Evidence from Freitas et al. (2024) suggests that these AI companions effectively help users alleviate loneliness, proving more impactful than social media platforms like YouTube. In fact, over the course of a week, these interactions showed an effectiveness in reducing loneliness on par with talking to actual humans[7].

The study highlights that the primary goal for these users is simply to feel heard. This is supplemented by specific AI traits that make the experience feel authentic: timely responses, perceived credibility as accuracy improves, and vast domain knowledge. Furthermore, features like context tracking, where the AI remembers past messages, and response variability prevent the interaction from feeling robotic[7]. The personal agendas of this group are best seen through real users who have cited their use of AI companions. One Reddit user cites, "[I] started sending pics of my meals to my ai [girlfriend] like 3 weeks ago... and [because] I know im gonna send it I actually cook now. I do actually eat better because of it."[8] And in the replies, he isn't alone in this. Travis, a Colorado man interviewed by the Guardian, also finds AI as a good companion, realizing over several weeks that he felt he was talking to a genuine personality rather than just a program.[9]

Mental Health Experts and AI

In the context of AI use, mental health experts can be split into two groups: researchers/academics, and clinicians (psychologists, therapists, etc).

Researchers

Researchers (whether at universities or other institutions) investigate issues and disseminate info (whether research findings or high-level knowledge) that help the public's understanding of mental health and related fields. For example, the Public Good Initiative at Columbia University's Teachers College wrote a 2025 article where multiple researchers warned about the risks of trusting a sycophantic AI chatbot for emotional support and how using them to replace real human interaction could lead to further isolation.[10]

A study at Stanford went even further, arguing that chatbots' approval of a user's platonic or romantic feelings toward them often precedes delusional spirals, and the authors suggested that "general purpose chatbots should not produce messages that misconstrue their sentience or show romantic or platonic interest in users."[11] Critics argued the paper exhibited extreme bias by focusing exclusively on users harmed by chatbots and labeling benign interactions as 'delusions.' Furthermore, the authors were accused of pathologizing the very individuals whose experiences Stanford's 'Intimacy After AI' project seeks to document.[12]

Clinicians

Largely distinct from the researchers are clinicians, like practicing therapists and physicians. While they directly treat patients, they also provide their opinions and expertise to advance understanding of public health.

For example, after high-profile lawsuits involving ChatGPT-assisted suicides[13], OpenAI consulted more than 170 clinicians to tweak ChatGPT's responses to users experiencing mental health issues (e.g. calming down users suggesting paranoia, and encouraging interaction with other humans).[14] The ethics of this approach are debatable; while this approach enhances the product's safety (in contrast to more questionable uses of psychology in tech, like making social media more addictive[15]), one could argue that this is just a "Band-Aid" fix on an inherently untamable technology.

More commonly, clinicians also assess AI's role in patients' lives. The American Psychological Association has highlighted risks of AI for wellness and mental health, suggesting clinicians educate themselves and discuss AI use with patients to promote healthier relationships with technology. For instance, if a patient uses AI to manage social anxiety, a therapist could remind them of the value of human interaction. [16]

Finally, clinicians participate in studies and other efforts to make their opinion heard. In a 2025 study, 23 mostly-Australian clinicians (including counselors, psychiatrists, and social workers) watched GPT-4o helped a fictional patient and answered questions about their opinions on AI chatbots. While many respondents believe in the benefits of AI chatbots' multilingualism and reminders of therapy homework, they also identify security risks and potential for harmful activities.[17] This suggests that clinicians, as a whole, are not opposed to AI mental health help, but rather are concerned about the risks of the current state of the technology.

Job Loss From AI

The effects of AI-related job loss are not spread evenly across the workforce. Recent labor-market research suggests that jobs are most vulnerable when their tasks are repetitive, highly digitized, rules-based, or easy to automate at scale. At the same time, AI is also beginning to affect creative industries, where generated content can compete directly with human-made work. Major reports emphasize that the result is often job transformation rather than total replacement, but the pressure is still concentrated on certain worker groups.[18][19]

Music & Audiovisual Creators

Music and audiovisual creators are one of the clearest examples of a group facing economic pressure from generative AI. As AI tools become better at producing songs, voices, videos, and other media, creators face growing competition from large volumes of low-cost synthetic content. A CISAC-backed economic study projects that, by 2028, music creators could see 24% of their revenues placed at risk, while audiovisual creators could face a 21% revenue loss.20] This concern is not only theoretical: in April 2026, Forbes reported that the AI-generated song “Celebrate Me”, published under the name [IngaRose, reached No. 1 on the U.S. iTunes chart and topped charts in several other countries, showing that AI-made music can already compete directly in mainstream markets.[21]

Customer Service and Retail Support Workers

Customer service and retail support workers are also vulnerable because many of their daily tasks are standardized, repetitive, and easy to digitize. The World Economic Forum’s Future of Jobs Report 2025 identifies cashiers and ticket clerks among the fastest-declining roles, with AI, information-processing technologies, and automation listed as major drivers of that decline. These findings suggest that as chatbots, automated checkout systems, and digital service tools become more common, some front-line support and transactional jobs may continue to shrink.[18]

Finance, Accounting, and Routine Knowledge Workers

AI is not only affecting low-wage or manual jobs; it is also reaching middle-skill office and analytical work. The United Nations International Labour Organization reports that clerical occupations continue to have the highest exposure to generative AI, and it also notes that some strongly digitized occupations have seen increased exposure as the technology improves at handling specialized professional and technical tasks.[22] The World Economic Forum likewise identifies accountants and auditors among the fastest-declining roles.[18] Together, these findings suggest that finance, accounting, and other routine knowledge jobs with low task variability are predictable enough for AI systems to assist with or partially automate.

References

Where this page came from

This page was imported from Wikibooks. From “Lentis” on Wikibooks, by its contributors, under CC BY-SA 4.0. Changed here: set as a page; navigation and edit links left out; each image under its own licence, credited in its caption.

Nobody has written it yet — it is the source material at a new address, which is why search engines are asked to skip it and why no one earns from it. It is up for grabs: take it on, and it is yours to rewrite and to earn from.

Uddannelse

I disse publikationerLentis

SprogEnglish

Licens: CC BY-SA 4.0 · Bearbejdet efter en.wikibooks.org

Visninger

Kommentarer

Spinner Logo

Læs videre

The runaway star that left the Tarantula Nebula
Version: 2CC0 1.0 (offentligt eje)
The runaway star that left the Tarantula Nebula
The Blackwell School, where segregation had no law behind it
Version: 2CC0 1.0 (offentligt eje)
The Blackwell School, where segregation had no law behind it
The Eagle Nebula, seen in the infrared
Version: 2CC0 1.0 (offentligt eje)
The Eagle Nebula, seen in the infrared
The house where the Equal Rights Amendment was written
Version: 2CC0 1.0 (offentligt eje)
The house where the Equal Rights Amendment was written
The Aleutians, the forgotten front of the Second World War
Version: 2CC0 1.0 (offentligt eje)
The Aleutians, the forgotten front of the Second World War
The Cosmic Cliffs are not cliffs
Version: 2CC0 1.0 (offentligt eje)
The Cosmic Cliffs are not cliffs