Video summary

I'm begging you to stop talking to ChatGPT

Main summary

Key takeaways

Technology

Summary

The video argues that habitual reliance on ChatGPT and other large language models (LLMs) can weaken independent thinking, reduce privacy, and displace personal reflection. The creator says the title is intentionally provocative, not a blanket instruction to avoid AI.

Main points

  • Hank Green and AI-assisted work: The video opens with online speculation that a line in one of Hank Green’s scripts came from an AI conversation. Green denied that, but acknowledged using AI to find research resources and said he worried that doing so was diluting his own process. The creator presents this as an example of the pressure to produce more with less time—and the appeal of tools that promise quick help.

  • Cognitive offloading: The creator discusses a small 2025 study in which 54 students wrote short essays using ChatGPT, Google without AI summaries, or no technology. She says the study reported broader brain activity among students working without technology and up to 55% lower activity in the ChatGPT group. She cautions that the study was small, but argues that having an LLM generate or synthesize material can reduce the effort involved in forming and remembering ideas.

  • Why calculators and writing are different comparisons: The video challenges the claim that AI is simply another tool like a calculator. Calculators still require users to understand the underlying math and check whether an answer makes sense; ChatGPT can produce work for someone who lacks knowledge of the subject. The creator also revisits Socrates’ remarks about writing, arguing that his concern was not writing itself, but the risk of treating stored information as a substitute for active discussion and understanding.

  • AI in academia: The creator criticizes using LLMs to summarize papers or rewrite research, saying these practices can bypass skills a PhD is meant to develop: locating sources, analyzing them, and communicating findings. She also raises concerns about submitting one’s own or others’ writing to AI systems, including consent, training-data use, and possible plagiarism risks.

  • Personalization and privacy: Drawing on a paper based on chat histories from 80 ChatGPT users, the creator says researchers identified 2,050 memory entries. The paper reportedly found that 96% were created unilaterally by the chatbot, with many containing personal information or psychological insights. She warns that chatbot “memories” may include assumptions rather than facts, and that intimate conversations are not protected by the same confidentiality rules as therapy. She notes that OpenAI later added an option to turn memory off, while observing that users may keep it enabled for convenience.

  • AI and mental health: The video acknowledges why an always-available, low-cost chatbot can appeal to people who cannot access therapy. The creator does not present herself as a mental-health professional and says the discussion is not professional advice. She argues that reliance on chatbots may not resolve loneliness or distress and can make it easier to seek immediate reassurance rather than develop confidence in one’s own reflection. She also distinguishes these broader concerns from the rarer, more severe cases described as “AI-induced psychosis.”

  • Alternatives: Drawing on Nietzsche, Buddhist thought, and ancient philosophy, the creator presents reflection and learning to live with suffering as skills people can cultivate—not substitutes for professional care when it is needed. Her practical alternative to confiding in a chatbot is writing in a physical notebook: a way to think through experiences, build a record of one’s life, and develop insight over time. Virginia Woolf’s diaries serve as an example of sustained personal writing.

Sponsor mention

The video is sponsored by DeleteMe, a service that seeks removal of personal information from data-broker websites and monitors for reappearances. The creator describes her experience with its removal reports and promotes the service’s privacy-advisor support and discount offer. This is a sponsorship segment, not an independent product review.

Reviews, guides, or tutorials

No product tutorial or formal product review is provided. The video offers commentary and a suggested personal practice—journaling—instead.

Main speaker and sources

  • Main speaker: Lady of the Library, the video’s creator.
  • People and works discussed: Hank Green; Socrates’ Phaedrus and the story of Thoth; Lane Wilkinson’s essay on Socrates and writing; a 2025 study titled Your Brain on ChatGPT; a paper analyzing ChatGPT memory and user chat logs; Zara Abrams’s article on AI-induced psychosis for the American Psychological Association; Friedrich Nietzsche; the Buddha; Seneca; Epictetus; and Virginia Woolf’s diaries.
  • Sponsor: DeleteMe.

Rate this summary

Your feedback will help improve summaries.

Improve this summary

Reprocess with a stronger model when the summary feels incomplete or inaccurate.

Pro

Translate summary in another language

Pro

Ask questions to this video

Chat for follow-up questions, clarifications, and source-backed answers.

Coming soon

Share this summary

Original video