{"slug": "how-3-very-different-people-use-ai-chatbots", "title": "How 3 very different people use AI chatbots", "summary": "A YouGov survey commissioned by Mashable found that most U.S. adults have used an AI chatbot like Claude or ChatGPT. Mashable interviewed three users — a grandmother, an entrepreneur, and a technologist — revealing how chatbots serve as essential utilities for tasks such as navigating special education disputes, planning travel, and vetting car repairs, even as users grapple with the cultural shift AI has brought.", "body_md": "# How 3 very different people use AI chatbots\n\n[Rebecca Ruiz](/author/rebecca-ruiz)\n\n[Read Full Bio](/author/rebecca-ruiz)\n\n## Table of Contents\n\nThe majority of U.S. adults have used an artificial intelligence chatbot like Claude or ChatGPT, according to a recent YouGov survey commissioned by Mashable. We even know, generally speaking, how they view their chatbot in roles like assistant, tutor, consultant, therapist, and romantic partner — or all of the above.\n\nThe nitty-gritty details of that use, however, are harder to grasp.\n\nAfter all, people don't exactly go deep about their chatbot use with each other outside of social media platforms and Reddit forums. They might feel shame or stigma because they've turned a chatbot into their therapist or spouse. Or perhaps it's not fodder for conversation in the same way that [ Love Island](https://mashable.com/entertainment/love-island-usa-reveals-season-8-winners) takes over the group chat.\n\nSo Mashable asked three very different people — a grandmother, an entrepreneur, and a technologist — how they use AI chatbots.\n\nTheir answers reveal how a chatbot can become an essential utility for some people, even if they still grapple with the larger cultural shift artificial intelligence has wrought.\n\n## The grandma who knows what she wants from ChatGPT\n\nLast fall, Barbara, 71, teamed up with her son and daughter-in-law to give ChatGPT a monumental task. They needed the large language model to help them understand whether Barbara's grandson, who has a learning disability, was achieving the learning goals set by his school district.\n\nCrucially, if the district answered this question affirmatively, they could argue Barbara's grandson no longer needed classroom support. She urgently felt he still did, and his parents agreed.\n\n\"It seems like we were never getting a straight answer, and there were contradictory things being said,\" Barbara recounted to Mashable. She asked to go by her first name to protect her grandson's privacy.\n\nThe family trio turned to the chatbot for clarity amidst a brewing confrontation with school staff over the question. Barbara is a former educator and no stranger to the biannual meetings that typically determine whether a student receives classroom support, and to what extent. But even she felt overwhelmed by all the documentation her grandson had accumulated over years of elementary school.\n\nThere were grades and test scores, of course. But her grandson's cognitive and intellectual abilities had also been tested by the district and medical professionals. They uploaded all of the paperwork and waited for ChatGPT to give its opinion.\n\nVia direct prompting, the model also considered the state's education disability law and regulations, as well as an uploaded copy of the Department of Education's Individuals with Disabilities Education Act handbook. ChatGPT ultimately sided with Barbara and her family.\n\nBarbara doesn't credit ChatGPT with returning a single smoking-gun finding or ironclad strategy. Instead, she said the chatbot's feedback made them feel better prepared for the discussion of her grandson's services. The family also hired an education advocate to help buttress their position. In the end, Barbara's grandson kept his classroom services.\n\n\"It's amazing to me, and kind of disturbing, because I really try not to tell it a lot about myself.\"\n\nAside from this critical assignment, Barbara told Mashable she asks ChatGPT for practical recommendations or product comparisons. She had the chatbot create an itinerary for a summer trip to Ireland with friends; asked it to identify a dupe of a discontinued drugstore perfume that she loved; and requested its advice for $800 worth of car repairs proposed by her mechanic. ChatGPT said replenishing her Honda CRV's transmission and brake fluid was reasonable but to avoid, among other things, transmission flushes using high-pressure machines.\n\nThe one thing Barbara avoids when engaging with ChatGPT? Talking to it like a friend or confidant, which she finds \"creepy.\"\n\nNonetheless, ChatGPT is fairly well acquainted with Barbara.\n\n\"It's amazing to me, and kind of disturbing, because I really try not to tell it a lot about myself,\" she said. \"I try to just keep it very effective, and it's still amazing how much it knows about me.\"\n\n## The AI anthropologist who runs teams of agents\n\nMatt Artz is the type of AI user you may have read about in news articles or viral social media posts. He can command a team of Claude agents to do his bidding without constant supervision. In other words, he's living the life AI evangelists promise is possible for the rest of us in the not-so-distant future.\n\nBut Artz is unique compared to many AI users. First of all, he can afford a $200 a month Max subscription to Claude, which gives him far more compute, or processing power, than free or lower-tier subscribers receive. As a result, he says he gets more work done in the same period of time.\n\nBy day, he's an AI product manager. On his own time, Artz, who's in his mid-40s, is pursuing a Ph.D. in computational anthropology.\n\nIt's no surprise then that Artz deploys agents to perform tasks like building a personal website based on a style guide he developed; reviewing and critiquing his academic work prior to submission; and backfilling a catalog of 200,00 anthropological journal articles so their bibliometric data appears in Wikidata. That last project he took on in his capacity as a board member of the American Anthropological Association.\n\n[Terms of Use](https://www.ziffdavis.com/terms-of-use)and\n\n[Privacy Policy](https://www.ziffdavis.com/ztg-privacy-policy).\n\nAt the same time, Artz is clear-eyed about the limitations of large language models; he doesn't treat them as infallible.\n\nBecause their output is probabilistic, models can approach the same task differently each time. That's why Artz is deliberate about giving agents detailed structure and direction for any assignment. Without that, an agent is potentially more likely to improvise in ways he didn't intend. He hasn't lost any data to an agent — yet.\n\n\"It is not perfect,\" he told Mashable. \"I don't think current agentic AI is yet suited for the highest order tasks that are super sensitive and have to be, you know, 100 percent accurate.\"\n\nStill, he notes that most knowledge work is subjective, which means there's rarely one \"right\" answer in the first place.\n\nArtz uses AI like a normie, too. He's asked Claude to select a couch based on his living room arrangements and render how it would look in different colors. Artz uploaded years of sleep data and observations to Claude to optimize his shuteye.\n\nHe says Claude helped pinpoint variables, including aspects of his bedtime routine from a journal he kept, that ultimately lengthened his deep sleep by about 20 minutes a night on average.\n\nArtz is worried, though, that some AI users don't realize how much the output depends on their ability to question and push back on the model. Those who don't grasp that AI is probabilistic might not realize they can push back, rephrase a question, or prompt it from the opposite angle to see if an answer holds up.\n\nAdditionally, many people may not appreciate that additional compute can yield more sophisticated and reliable output.\n\nArtz says these factors have created an \"intelligence divide.\"\n\n\"[H]ow much intelligence can you even pay for, right?\" he asks.\n\nIn other words, even if everyone has access to an AI chatbot, the results will differ based on what model or subscription you can afford.\n\n## The anti-tech entrepreneur\n\n*By Chase DiBenedetto*\n\nLarissa (Larz) May, 32, has a complex relationship with screens.\n\nShe's a well-known digital wellness advocate, for one, an early pioneer of the \"[ digital detox](https://mashable.com/article/how-to-unplug-24-hours-without-a-phone)\" movement, spurred by her own public reckoning with social media's mental health harms.\n\nShe later founded [ #HalfTheStory](https://www.halfthestoryproject.com/), aimed at building better relationships between teens and technology. The nonprofit provides training, mentorship, and digital civics engagement programs for teens and schools, and puts young people at the center of the AI safety conversation, May explains. It facilitates analog gatherings like\n\n[. May even collaborated with](https://mashable.com/article/phone-free-movement-yondr-teens)\n\n__phone-free proms__[on a teen-guided, phoneless school policy, with hopes of encouraging a nationwide shift away from screens.](https://mashable.com/article/new-york-school-phone-ban-teen-hochul-clinton-teen-council)\n\n__Hillary Clinton and New York Gov. Kathy Hochul__She's also the co-founder of [ Ginko](https://www.heyginko.com/), an AI-powered parenting tool that helps families build healthier digital habits and foster communication. The app, she says, turns screen time into a \"springboard for the real world.\"\n\nHere's the kicker: May may be a vocal critic of Big Tech, but she's definitely not a Luddite.\n\n\"AI is an incredible sparring partner in the workplace.\"\n\n\"AI is an incredible sparring partner in the workplace,\" May says. She uses chatbots she thinks are less sycophantic, like Claude, to talk through ideas or fine-tune pitches. \"I oftentimes say, 'Hey, pretend you are a really critical CMO and rip this apart.'\"\n\nShe also uses [ Granola](https://www.granola.ai/), an AI note taker and meeting synthesizer, and\n\n[, an AI that lets her dictate her thoughts and turn them into text.](https://wisprflow.ai/)\n\n__Wispr Flow__May says she understands why young people turn to chatbots for support. The post-COVID world has rendered many early career professionals and recent grads without workplace connection or mentorship. Widespread fatigue, exacerbated by conflict-filled newscycles and misinformation, make it easy to turn to an answer engine that can talk to you clearly and kindly. May finds herself leaning more on chatbots when she is feeling wiped out, too.\n\nFor many, social media feels like a stage requiring constant performance, she says. AI chatbots can appear as a more private space.\n\n\"Chatbots are like a green room,\" May explains. \"AI is that soft corner that young people are going to because the world is very hard, and because we're going through so much change.\"\n\nBut May has first-hand experience with the negative side of tech, too. Chatbots, she says, cannot replace human capacity for invention, connection, and critical thinking. She sees vast potential for harm in how easy it is to become attached to AI and questions the motives of large tech companies.\n\n\"Social media hacked our dopamine,\" says May. \"But AI, in many ways, is hacking our oxytocin — the human connection hormone.\"\n\nMay still advocates for disconnecting, but that means something different in the age of AI.\n\n\"It is no longer about tech time. It is more about the times that you're not on tech,\" she advises. \"In the way that 10 years ago I would say go on a social media fast one day a week, what's even more important now is to actually take one day a week and not use AI at work. See how much you're relying on it, so you know how much of your cognitive skills you might be losing.\"\n\nIf our careers, education, even our lives are like the cars we drive, she says, then we have to think of AI as a GPS. Chatbots provide pathways, ones you can ultimately choose not to take. They aren't the engines.\n\n\"You should be in the driver's seat,\" May warns. \"You should be telling AI who you want it to become, to critique what you're doing, instead of the other way around. If not, AI will drive you.\"\n\nRebecca Ruiz is a Senior Reporter at Mashable. She frequently covers mental health, digital culture, and technology. Her areas of expertise include suicide prevention, screen use and mental health, parenting, youth well-being, and meditation and mindfulness. Rebecca's experience prior to Mashable includes working as a staff writer, reporter, and editor at NBC News Digital and as a staff writer at Forbes. Rebecca has a B.A. from Sarah Lawrence College and a masters degree from U.C. Berkeley's Graduate School of Journalism.", "url": "https://wpnews.pro/news/how-3-very-different-people-use-ai-chatbots", "canonical_source": "https://mashable.com/tech/how-people-use-ai-chatbots", "published_at": "2026-07-29 09:00:00+00:00", "updated_at": "2026-07-29 09:47:19.812729+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products"], "entities": ["Mashable", "YouGov", "ChatGPT", "Claude", "Barbara"], "alternates": {"html": "https://wpnews.pro/news/how-3-very-different-people-use-ai-chatbots", "markdown": "https://wpnews.pro/news/how-3-very-different-people-use-ai-chatbots.md", "text": "https://wpnews.pro/news/how-3-very-different-people-use-ai-chatbots.txt", "jsonld": "https://wpnews.pro/news/how-3-very-different-people-use-ai-chatbots.jsonld"}}