HUMAN.JROH.
Menu
I need help

HUMAN interactive guide

AI and mental health.

Why do we sometimes find it easier to confide in AI than in a person? When does it help us, and when does it simply agree?

A paper scene of a woman arranging her own thoughts beside a calm digital counterpart.

We will work through it together, step by step. Using examples of conversations, you will explore what happens and what you can do about it.

Start reading

Jakub Roh · Gabriela Rohová

Guide contents
1

Starting and guiding a conversation

  1. 1. Disinhibition: when you can finally ask 5 min
  2. 2. Research: what it means for AI to help 3 min
  3. 3. Open questions: from a label to a specific situation 4 min
  4. 4. Cognitive offloading: when there are too many things in your head 3 min
2

Check and correct

  1. 5. Sycophancy: when an assumption sounds like fact 5 min
  2. 6. Cross-checking: a second pair of eyes in a new chat 4 min
  3. 7. An apology is not a correction: everything that rested on the error 4 min
  4. 8. ADHD and other hypotheses: a suggestion is not a diagnosis 6 min
  5. 9. A medical report: an accurate transcription before interpretation 3 min
  6. 10. Looping: when another answer does not help 4 min
3

Using the result in life

  1. 11. The AI sandwich: taking the result back into your own hands 3 min
  2. 12. Context, memory and permissions: what the assistant carries forward 4 min
  3. 13. Psychosis, suicidal thoughts and further help 7 min
  4. 14. Parents and schools: a conversation first, not a phone inspection 5 min
Practical cards

For classroom work: a plan for one lesson.