Dr Prompt
General practitioner in agentics
Open at all hours, mostly at night
Written in us-east-1
In-network with every GPU
Billed per token
Tokens: 35 billion
1 prompt every 23 min, nights included
For 52 days
Refills: unlimited
Parody prescription, no medical value. Do not take to a pharmacy.
Tokens Anonymous. Hello, doctor. It’s been since my last prompt.
The fellowship for people who have said “one more prompt and I’ll go to bed”… at 3 a.m.
Dr Prompt
General practitioner in agentics
Confidential
Patient #001’s chart
| Mar–Apr 2026 | Aug–Sep 2026 | |
|---|---|---|
| Projects per day | 2 | 4.5 |
| Project switches per day | 1 | 14.5 |
| Time on one project in a row | 1 h 57 min | 27 min |
| Prompts between 10 p.m. and 6 a.m. | not measured yet | 26% |
Read Member #001’s share
Hi, my name is Member #001, and I’m addicted to omnipotence.
I work in tech. I’ve always had a hundred ideas a day. I used to build one a year. Now I build all of them. The same night.
Since September 2025, my agents have burned through 35.2 billion tokens. On October 5, I used 4.3 billion in a single day.
In 52 days, I’ve typed 3,287 prompts by hand*. Never between 5 and 7 a.m.: I think that’s when I sleep.
My longest agent turn ran for 10 h 50 min without stopping. At least it doesn’t brag about it on LinkedIn.
I’ve hit Escape 329 times to interrupt it. That’s the only measurable trace of my willpower.
| 0 h | 164 |
|---|---|
| 1 h | 148 |
| 2 h | 94 |
| 3 h | 59 |
| 4 h | 25 |
| 5 h | 0 |
| 6 h | 0 |
| 7 h | 3 |
| 8 h | 24 |
| 9 h | 33 |
| 10 h | 121 |
| 11 h | 141 |
| 12 h | 149 |
| 13 h | 305 |
| 14 h | 255 |
| 15 h | 191 |
| 16 h | 195 |
| 17 h | 287 |
| 18 h | 241 |
| 19 h | 157 |
| 20 h | 166 |
| 21 h | 178 |
| 22 h | 181 |
| 23 h | 170 |
A typical day per period, computed on October 8, 2026 from Member #001’s usage log, without a word of the conversations. Night prompts: tracked since August 17.
Overheard in the waiting room
HIPAA violation? Not even: they posted it themselves. Just like our prompts, which all go to Anthropic, OpenAI or Google.
It’s 2:47 AM on a Tuesday. I’m not debugging a production outage. There’s no deadline. I’m just watching Claude Code refactor a module — and I can’t stop.
So addicted to Claude Code, I stayed up 19 hours yesterday and didn’t sleep til 5AM.
Agentic coding is addictive. You will hear it more and more often.
So, what brings you in?
Tick what sounds like you. Nothing is sent: it all stays in your browser.
Diagnosis
0 of 9
Recreational use
Or denial. Either way, your chair is still here.
Grey zone
You’re not addicted. You’re just very, very productive between 11 p.m. and 3 a.m. First goal: the 24-hour chip.
Welcome
Grab a coffee (drip, lukewarm, styrofoam cup). Your first chip: 24 hours without a prompt. It starts tonight.
Your next chip24 hours without a prompt
Fewer tokens. More chips.
This is not a diagnosis. The 9 questions follow the structure of the 9 criteria for gaming disorder, which research scales on generative AI also build on4.
Dr Prompt
General practitioner in agentics
Patient: same one
Tapering plan
One chip per milestone, like at meetings:
- 24 hours without a prompt
- 7 days without an agent running overnight
- 30 days without --dangerously-skip-permissions
- 1 year without asking the AI if you’re addicted
Do not refill indefinitely.
Parody plan, no medical value. For real support, see the resources further down.
This is not a prescription.
End of the parody
Here’s the real questionnaire.
Anonymous, two minutes, nothing that could identify you. We’ll publish the results here once we have enough answers.
- Your answers become counters as soon as they arrive: no IP, no cookie, no timestamp.
- Fatigue answers are never kept together, only in pairs.
- Nothing will be published for any group with fewer than 30 answers.
For real
A parody site with serious sources. Here’s what holds up, and what doesn’t yet.
What we know
- The slot machine
- An agent sometimes nails it and sometimes fails, with no warning. That’s variable-ratio reinforcement, the same mechanism that makes slot machines so hard to walk away from. Well-known developers describe exactly this1.
- Agent brain fog
- BCG gave a name to the exhaustion of people supervising AI agents: “AI brain fry”. Mental fog, slower decisions, headaches, and in the end more errors and more intent to quit7.
- AI brain fry and AI fatigue: the signs, the study, what helps
- The ADHD brain
- In young adults with ADHD, it’s reward processing, not attention deficits, that predicts internet addiction. The study followed 682 people over time2.
- The known profiles
- One study describes three ways to get hooked on a chatbot: escapist roleplay, the pseudo-social companion, and the epistemic rabbit hole3. One is missing: the builder, who starts a new project every night.
- The order of magnitude
- A scale tested on 1,500 people in China and the US puts problematic generative AI use at 5 to 10%4.
- AI helps, too
- People with ADHD who talk about their chatbots online are mostly positive about them6. That’s exactly the problem: nobody gets hooked on something that doesn’t work.
What we don’t know
- “AI addiction” is not a recognized diagnosis. Some researchers question the concept entirely5.
- The link between ADHD and generative AI has barely been studied: in a first, preliminary study, only the group with autistic traits stood out8. We’re mostly extrapolating from internet and gaming research.
- A token counter doesn’t measure suffering. 35 billion tokens can also be a very good month of work.
If it’s not funny anymore
Sleeping badly, pulling away from people, unable to stop even when you want to? Talk to someone. A human.
- Internet and Technology Addicts Anonymous
A real 12-step fellowship, free and online, with a page dedicated to AI.
- SAMHSA National Helpline (US)
Free, confidential, 24/7 treatment referral and information, in English and Spanish.
- Your doctor
Outside the US, a GP can refer you to an addiction or ADHD specialist.