Harvard Business Review: How People Are Really Using AI in 2026 (Wow, is this depressing and worrisome.)

[highplainsdem]

https://hbr.org/2026/06/how-people-are-really-using-ai-in-2026

I'm going to copy the top 30 (of 100) on the list below, and I think you'll see why this was a very sad list of uses for a tech we're risking so much and already harming so much for.

As some of the conclusions at HBR warn:

-snip-

In at least a quarter of the top use cases this year (therapy/companionship [#1], relationship advice [#7], enhanced decision-making [#13], organizing my life [#14], drafting emails [#42], generating ideas [#47]), people are asking AI to do some portion of their thinking. There’s an argument that this kind of AI use is cause for concern. First, because these types of activities are precisely those for which human beings need to take responsibility. And second, because such activities are the kinds that we thrive at, as a species.

-snip-

AI is optimized to engage and keep us engaged. When it lavishes praise on an unrealistic business idea or a mediocre sales deck, we may be inclined to stop fine-tuning too early and to return for more feedback of this congratulatory variety.

-snip-

AI is also becoming a go-to for people seeking help for emotional issues, including comfort or advice about personal relationships: relationship advice (#7), love lives (#17), reconciling personal disputes (#26), sex (#46), and interacting with the deceased (#86). The trend continues for work-related emotional uses: Safe space to ask (#32), boosting confidence (#43), adjusting tone of email (#58), preparing for interviews (#89).

Therapy/companionship is the #1 use case this year, as it was last year. We found more than 1,400 such entries this year, constituting 11% of our dataset (up from 5% last year), so it’s grown quickly in absolute and relative terms.

-snip-

And we thought smartphone screens were a problem. This AI use picture looks much worse.

The top uses for AI that they found:

1. Therapy/companionship
2. Troubleshooting
3. Fun and nonsense
4. Fan fiction and storytelling
5. Technical use of software
6. Autonomous agentic operations
7. Relationship advice
8. Work buddy
9. Astrology and tarot readings
10. General advice
11. Fake reality TV
12. Medical advice
13. Enhanced decision-making
14. Organizing my life
15. Generating relevant images
16. Healthier eating
17. Navigating love lives
18. Specific search
19. Creative writing
20. Editing text
21. Vibe coding
22. Enhanced learning
23. Drafting a document
24. Career advice
25. For entrepreneurs/startups
26. Reconciling personal disputes
27. Shopping
28. Homework
29. Breaking the rules
30. Business advice

5, 6, 21 and probably 2 would be software engineers and developers. Their uses of AI are a lot of what I've read most about lately.

25 is for the people expecting to get rich off AI.

1, 7, 17 and 26 are emotionally and psychologically risky uses of AI.

3 and 15 are, I'd guess, largely AI slop art.

4, 19, 20 and 23 are using AI for writing.

10, 12, 16 and 24 are asking AI for advice where searching for advice from real experts would be much less risky.

22 might be asking AI for summaries or simplified language to explain things.

28 might include a lot of examples of AI doing the assignment.

8 might mean AI as a work buddy who does the work.

9 - ?????? People are using AI for astrology and tarot readings?

11 - Fake reality TV?????

18 - Search with the possible bonus of hallucinations.

27 - Shoppers who probably haven't considered that AI makes it easier to target them with ads and influence their decisions.

29 - I have no idea what "breaking the rules" means in this context. Is simply using AI breaking the rules for them? Are they using AI to help them do things that would be considered rule-breaking even if done without AI (scams, deepfakes, hacking etc.)? Are they trying to jailbreak AI models and get them to ignore their guardrails?

To me, this list suggests a lot of people turning their thinking and work and self-expression over to AI, dumbing themselves down and making themselves more and more dependent on flawed, error-riddled technology designed to flatter them and keep them engaged.