How to think about AI: two hours with Dmitry Merkulov
A philosopher and a practitioner spent almost two hours talking not about how to use AI but about how to think about it. Key ideas, chapters with timestamps and the article series.

On 15 November Dmitry Merkulov from the School of Great Books and I recorded a conversation that runs 1 hour 49 minutes. We agreed from the start that there are plenty of instructions on how to use AI already, so we would talk about how to think about it.
Dmitry and I used to host a show called News, about politics. Then both of us found something more interesting. In half a year my hobby turned into work: I started building a game for myself, and now I build websites, bots and agents for people and earn a living from it. Dmitry watched this from the side for a long time, a little ironically, until he asked me to help him set everything up. A few days in, he said he was walking around like someone in love: you press a button, and you made that.
So the conversation became a practitioner talking to a philosopher. I describe what I see in projects every day. Dmitry, who has an IT degree and years of reading big books behind him, finds the language for it and the places in history where it has happened before.
Video
The recording is out on Dmitry Merkulov's YouTube channel: watch the video. The conversation is in Russian, and I will publish an English version later.
What Dmitry said
News is almost never new, so it is worth looking at it as if from eternity. The same goes for AI: it is more useful to learn to find your way in the future, whatever it turns out to be, than to try to guess it.
The state you are in while you create something should mean something right now, whether or not the project pays off later. Dmitry compares it to writing to someone you have not met yet: however it ends, what you feel today is already yours.
A subscription to an AI tool is better seen as an education. People go to school not only for knowledge but to learn how to live among others. Here too you pay to get used to new rules, and the earlier you start, the better.
His main point about the dark side: talk about machine consciousness and will is convenient because it takes responsibility off people. Responsibility always stays with whoever launched the thing.
What I said
I caught myself giving tasks to the most advanced technology in the world worse than I talk to people: in fragments, irritated, hoping it would figure things out on its own. Voice input became my way of relearning how to speak, especially since I live between three languages.
I run many projects at once and call it the blind archer strategy. I built a website about robots with no goal at all, just in case, and it now gets around 2,500 visits a week.
And I already see employees in companies excusing mistakes with a line like that is what GPT gave me. It is a childish position and a very convenient one, and the temptation is hard to resist.
Chapters
- 0:00 Intro: why we stopped doing News
- 3:52 How a hobby became an income, and the flow state
- 11:07 Why Dmitry wanted to record this
- 19:44 Language: how we talk to a machine
- 28:50 Vocabulary: AI, native, vibe coding, agents, swarm
- 35:14 AI and the word artificial
- 41:32 Native: technologies that become part of life
- 48:46 The barber, the tiler and reels: who will compete with whom
- 55:26 Industrial revolutions, a fly on Mars and why machines were invented
- 1:02:33 A unicorn company of one person
- 1:06:39 What vibe means and how coding became a conversation
- 1:11:13 The answer you got is the question you asked
- 1:14:09 Where to start: with routine or with what you love
- 1:16:57 Music and images made by AI
- 1:22:41 The blind archer strategy
- 1:29:29 Biotech, robots and a low barrier to entry
- 1:32:39 Software for yourself and the end of ecosystem power
- 1:41:30 Responsibility stays with people
The series
The conversation turned into a series of articles that will come out on this site one after another: why I give tasks to a machine worse than I talk to people, the blind archer strategy, who is responsible for work done with AI, how clients will look for contractors, how I found my first clients, and a joint piece with Dmitry on what the stirrup, gunpowder and the printing press tell us about AI.
Two articles grew out of our talks before the recording: AI native: what it means and why the word artificial gets in the way and AI adoption in a company: where to start.
FAQ about the conversation How to think about AI
Who is Dmitry Merkulov?
One of the founders of the School of Great Books, where adults read and discuss classic texts slowly. In the conversation he is the host and the philosopher who finds words for what is happening with technology.
Where can I watch the video?
On Dmitry Merkulov's YouTube channel, here is the video. It is in Russian for now, and an English version will follow.
Is this a ChatGPT tutorial?
No. There are almost no instructions in it. We talk about how to think about AI: what changes in language, in work, in competition and in responsibility.
How long is the recording?
1 hour 49 minutes. The chapters with timestamps are above, so you can jump straight to the part you need.
Will there be a follow-up?
Yes. We agreed to continue in a tighter format: one topic per conversation, plus the news connected to it.
Keep reading
All articlesAI native: what it means and why the word artificial gets in the way
AI native means AI is built into the work rather than living as a separate project. We look at the word itself, at what artificial means in English and in Russian, why vibe coding replaces neither the programmer nor the employee, and where an AI-native company starts.
ReadAI adoption in a company: where to start when nobody knows where to start
AI adoption means bringing AI into the daily work of a company. Step by step: who should build the first agent, which process to pick, what stays manual (registrations, tokens, access), what it costs and which mistakes we see most often.
Readminimir on Zenodo: a preregistered artificial life experiment with a DOI, and what protocol B2 changes
minimir, my artificial life experiment, now has a citable DOI on Zenodo, the open research archive run by CERN. What protocol B showed, how it was run, and how B2 is designed.
Read