Welcome to Humanmaxxing, an interview series on thriving in the age of machines. New episodes drop weekly. You can follow along on Substack, YouTube, Spotify, and Apple Podcasts.
We partnered with Granola (the actually discrete meeting notetaker we LOVE and use every day) to sponsor this series and record each conversation. You can chat directly with Jack’s conversation here, pull out takeaways, and follow whatever thread interests you. If you want to try Granola unlimited, use code SUBLIME to get 3 months for free here.
PSA: Code for Creatives x Sublime is coming late September. Alex Dobrenko—the host of this series and a longtime Sublime collaborator—runs Code for Creatives, a six-week live course where he teaches creative people to make things with Claude Code. We’re teaming up on a special cohort, which includes a bonus week on translating your ever-evolving taste and references into an asset you can use as you make things. Space is limited. Applications are reviewed as they come in.
Jack Butcher has built one of the most singular careers on the internet. He first became known for Visualize Value, where he distilled complex ideas about business and creativity into a visual language that traveled everywhere, then turned that attention into an independent education business.
More recently, his work has expanded into participatory art. At Art Basel Miami Beach in 2025, his installation Self Checkout invited visitors to pay whatever they wanted for a receipt and made the project’s costs and profits visible in real time. A display began at –$74,211 — the cost of producing the exhibition. By the end, 5,837 receipts had generated nearly $189,000 in revenue.
Throughout everything he does, Jack’s genius is bridging design, markets, and psychology to challenge and expand our understanding of value in the modern world.
In this conversation, Jack talks about how AI is changing his creative process, why he open-sourced his entire writing history, where creative value lives now, and why dismissing new formats (even brain rot) can prevent us from evolving creatively.
You can watch full conversation on YouTube and listen on Spotify or Apple Podcasts. Plus, thanks to Granola, you can chat with the conversation here.
Committed wordcels can find the highly edited transcript below.
Enjoy!
Sari
ALEX DOBRENKO: What’s AI feeling like for you these days?
JACK BUTCHER: One meta observation is it puts you in this really weird spot where it’s so easy to execute on idea X that by the time I’ve gotten 80% of the way there, I’m like: I didn’t think big enough. I should have just done something way more ambitious. If you’re not really intentional about it, you just get stuck in that loop.
I have 90 GitHub repositories that I feel like I’ve got a better version of, which is also hilarious.
The constant line that I’m trying to navigate with these tools is whether what I’m doing is unique, based on my opinion, perspective, experience, or whether I’m being seduced by the machine making progress. I do know that I no longer hate the very specific creative processes that I hated before. I enjoy some of what I would have considered admin work from before. I get the satisfaction of the machine doing things for me, even if the amount of labor I’m doing is the same.
The constant line that I’m trying to navigate with these tools is whether what I’m doing is unique, based on my opinion, perspective, experience, or whether I’m being seduced by the machine making progress.
AD: What are some of those things that you no longer do that you’re happy about?
JB: Specifically for design processes, I really appreciate the idea of being able to query the history or the evolution of a project and turn it into documentation. For example, I’ll work on something for a company, and instead of just sending over an image, I can send over the process. This used to be the most painful thing about commercial creative work. Nobody in an agency wants to do the case study stuff. Now you’re one good prompt away from a complete case study: here’s where we started, or the decisions that guided this thing to where it ended up. The machine just doesn’t forget, and it really can show how much work you did.
The other is really simple file management stuff. I don’t have a dev background, so if I’m working on something with someone, getting the LLM to help with GitHub, versioning, and having the deliverables live on a website instead of 900 PDFs with different file names is a huge unlock for me. Every career or discipline or craft could probably learn something from software development as a practice, because of how much it changes the way you think. Versioning and change logs and pull requests are happening everywhere, but nobody really gets to see them.
Basically, it’s incredibly useful in organizing creative projects with more than one contributor, and now there’s an actual record of work, so people want to contribute.
AD: I was thinking about how you released an MD file of your tweets. And the tool that you built, Spam, that’s basically trained on you, on your writing, to send an email every day.
JB: At everyday writing, that thing is far better than I am. There are 50,000 tweets that I’ve written over the last five or six years. I’d felt defeated a few months ago when I realized that, honestly, unless I’m writing a diary entry about something I really don’t want to tell anybody, an LLM is better at explaining the vast majority of things than I am. It doesn’t leave holes in logic. It’s not true for everyone, maybe not fiction writers.
Tweeting is probably the only thing I’ve done enough of that I could do a proof-of-concept for. And I got a couple of replies to my experiment like, why would you do this? This is stupid. You’re destroying your own value. And I guess I’m like, well, anybody else could do this too, right? They could just go and scrape my tweets and do this in 10 minutes. It’s not defensible. I’ve found it super useful. I plug it into anything I’m working on, to speak in my way and give me feedback: take this really lengthy passage that comes out of the standard model and run it through this rule set and strip it down.
AD: When you said that anybody can scrape this tweet data, it feels like I’m staring into a black hole void where I, as a human, don’t exist anymore. You’re not so afraid of that. I wonder how you do that.
JB: Thinking like this forces you up the abstraction tree. I’m not gonna try and compete with this thing. Remember when everybody would criticize Amazon as a product but have seven Prime boxes show up at their house every day? It’s the same with AI. Creativity should not be automated, but then every single person is using one, two, three, four AI products every single day. Creativity, as a practice, is just evolving, especially digitally.
I would say, right now, being transparent about the fact you’re using AI is the only way to play it so your audience keeps trusting you.
Remember when everybody would criticize Amazon as a product but have seven Prime boxes show up at their house every day? It’s the same with AI. Creativity should not be automated, but then every single person is using one, two, three, four AI products every single day. Creativity, as a practice, is just evolving, especially digitally.
AD: People don’t want to be deceived. That’s what I think audiences hate: I can’t believe you think I’m that dumb.
JB: Just send them the prompt. If you can communicate it to the machine in two sentences, just say the two sentences to the person. I went on this podcast maybe last year and I actually made friends with this woman who teaches film at a university in Hong Kong. She told me that every single one of her students is submitting slop essays. Every single one. And I was like, tell them that you’re just going to grade their prompt. If it’s not an English literature class, where you’re trying to make a stylistic choice about every word, it’s a different exercise anyway, right? The essay existed in the past because it was the only vehicle to express your knowledge. But imagine submitting prompts. It’d be an incredible class. You could fit the prompt on the projector. Kids would read each other’s prompts, even though they never read each other’s essays.
But also, you have to get way more creative even with what you’re asking people to do. If somebody can satisfy the requirements of your ask with a two-sentence prompt, your ask is pointless.
AD: I think a big part of why it’s hard for people to admit using AI is because the assumption is that they’re not needed. If my whole job could be done with AI, then you could just fire me, right?
JB: Rational fear. And there’s the speed at which this is happening. Philosophically speaking, I think it’s a positive thing that entering numbers into a spreadsheet is not going to be a job in 100 years. But if that’s happening in six months, the fallout is not good. I’m going to contradict myself again. Maybe there’s a good argument: a massive percentage of work today could have probably been automated away 10 years ago with the technology that exists — it just hasn’t been. So that’s another slightly comforting thing: just because the technology exists to make something more efficient does not always result in it becoming more efficient, at least not immediately.
But we have to be honest with ourselves. What are we aging out of? Somebody I work with just did this art show in Hong Kong. And there’s a guy named Justin who works for that gallery who I just did a podcast with. He loves brain rot content. He advocates for it. The typical position is: oh, this is destroying everyone’s minds and it’s terrible. And he’s like, man, if you really study this stuff, it’s so unbelievably creative and nuanced and interesting. And it’s obviously resonating. There are no rules to what format is legit, and when you get trapped in that, it really prevents you from evolving your career creatively, for sure. But a lot of people do, it’s all fear.
There are no rules to what format is legit, and when you get trapped in that, it really prevents you from evolving your career creatively, for sure.
AD: What do you think you’ll never use AI for?
JB: Communicating with my parents.
AD: You mentioned earlier diary entries being a thing that AI can’t do well. Do you think you’re gonna do more of that — writing more stuff that AI can’t write? More honest, vulnerable?
JB: I’ve done a version of that, which is live streaming. Watching a live stream, unscripted — here I am messing around with this stuff — is compelling in a way that it maybe wasn’t a couple years ago.
AD: One thing that I admire, or I’m jealous of, that you seem to have is: you don’t seem to feel any cringe about anything. Is that true? How do you do that?
JB: Maybe it’s just enough reps where you were wrong about something that you stop filtering yourself. I’ve gotten to that point. And people want to be a part of your process. I’ve been doing this now for five or six years, and the people I’ve met in person for the first time who have been following anything I put out — the quality of conversation and relationship you have with people like that is astounding. The object that you’re putting out doesn’t contain everything in it, right? People are piecing together this picture over ages and ages and ages. And if you remove yourself from that, then you’re defeating the purpose of broadcasting the signal in the first place, especially if your objective is to make something interesting that people want to talk about, and talk to each other about. It’s sort of exposing how much respect anybody has for their audience, I suppose — the degree to which they believe the people listening to them have any ability to decipher anything.
AD: It’s like in improv: treat the audience like they’re geniuses. Don’t dumb anything down.
Thanks again to Granola for making this series possible. You can chat with the full conversation here and use code SUBLIME to get 3 months free here.
Premium members can enjoy the much longer unedited conversation with Jack Butcher below.
