In a previous life, I wrote Clojure every day and still look back fondly attending Clojure/Conj and sitting next to Rich Hickey and other Clojure greats at dinner.
My first startup was all Clojure. AWS only had a dozen or two products and I think we must have been the first to compile Clojure to JS and run it on Lambda in production (the only runtime was Node.js 0.10 at the time).
I’m no longer in a professional engineering position and have a family and toddler so less time for learning about or playing with program languages. Clojure is what really made me first start thinking about programming languages and down the rabbit hole of scala and Haskell and idris and type theory books. Which are incredible and I can’t recommend more and also not remotely practical.
I miss it. I’m now more removed and building software as a means to an end.
I miss it just like I miss the program language and type theory group meetups in SF and working through problems in dependently types languages like Idris and being out of my depth.
Yes, sad there's no European presence (unless I missed someone). There's tons of Clojure over there. Metosin, Juxt, Borkent, Gaiwan (Arne), Flexiana, Peter Strömberg (Calva), Dustin Getz (hyperfiddle/electric), Christophe Grand (ClojureDart), Bojidar (CIDER), Renzo, and many, many more.
Clojure had lousy error messages, agents deal with this well.
Clojure is capable of producing some of the most dense code I’ve ever seen, so manual code reviews really start to feel like a bottleneck unless your goal is to level up.
I find it actually the best substrate to write AI tooling. All my custom MCPs are written in Clojure (bb). You hook up the agent to the REPL and let it go wild - it builds something nice. Also, Clojure is one of the most token efficient PLs.
One of the main problems I have with the models coding is the feedback loop is way down the chain from generation, it's out at the commit boundary for python when your hooks are running, maybe at the point where the model wants to push a PR. The REPL lets that happen during generation, and the other safety measures help immensely. Immutable data, STM, all of the features in Clojure that gave devs super powers now do the same for a model.
It definitely is more "mainstream" than others but I just don't see the same level of attention and enthusiasm around it anymore. I'm sure it is still being used in many places but like Elixir, hiring remains on the tough end.
Hiring good talent was always problematic. This has nothing to do with the quality, capacity, and robustness of the language or its relevance. Hiring people who would love to use Clojure but have no prior experience is not that difficult - it's just that every company wants an expert, but they don't want to offer expert salaries. In places where they do, the competition is nuts, on top of that, experienced Clojuristas typically get interviewed and scrutinized with the same level of rigor as architects.
The industry should have optimized for hiring people interested in PLs like Clojure instead of LeetCode drillers. Clojure is rarely the first, second, or even third programming language people choose to learn. It demands a specific vision, dedication, and discipline that fundamentally transforms how people think about computation, data flow, distributed systems, and concurrency. The ROI from hiring an average developer experienced in Clojure has the potential to significantly exceed that of a typical hire. Even when there's zero Clojure in prod.
I've had Clojure on my resume for 10 years, mainly to see if anyone would ask about it. Nobody ever has, until an interview a couple days ago. We'll see if it actually helps in leading to an offer, I guess.
I have the opposite experience - been using Clojure for over a decade and it feels like only that mattered for the last five jobs. Even though it's really only just one of many layers that required to do the job. I honestly would love to find a non-clj team and convince them to use it. There are so many useful scripts we write in babaska alone, it just sounds wasteful not to use that path, fully knowing of its existence.
Clojure is more relevant than ever in post agentic coding because of immutability and the REPL. The two big problems with agentic coding is context growing in unbounded fashion, and agents being able to get quick feedback on what they're doing. Mainstream languages fail on both accounts. I've found Clojure has been a great fit for keeping agents on track.
Incredible: I had not idea NuBank discovered Datomic first and that it's Datomic that led them to Clojure, 100 million+ customers, and eventually acquiring Cognitect.
Good to see David Nolen (aka "swanodette") is in the documentary too.
As a bonus here's a recent talk from David Nolen about Clojure/ClojureScript and using DOM morphing instead of React.
If you don't want to watch it all, just take two minutes to watch from 23m15s to 25m15s. He compares a behemoth slurping all the browser's CPU and RAM resources versus a 13 Kb of JavaScript + Web components and DOM morphing:
I don't know if it's still the case, but at old clojure conferences, or meetups, or places of employment, emacs was a prereq and assumed (and the most enjoyable)
You know there's a saying in Russian, that roughly translates to: "an expert surgeon is capable of helping a bad dancer", which is on itself is a reference to another idiom: "a bad dancer always blames his own balls".
That's quickly becoming befitting for cases like this - so often people rush to blame AI without even trying to use their own reasoning. I don't know what to say, hope you find a good surgeon, because it is obvious - you're shit of a dancer.
The very official Clojure page in TFA links to clojure-mcp (written by the person who created figwheel: a famous ClojureScript library in the Clojure ecosystem) and other AI resources related to Clojure.
It's not because Rich doesn't want AI-generated pull-requests by people then taking credits that the Clojure community is anti-AI.
I use Claude Code CLI daily with Clojure, just not in a "write me five thousands lines of Clojure code I won't read" type of way.
I don't think it is, considering they highlighted it in a post about human craft [1]. I read somewhere it was illustrated by felipemelo.net, but can't find the reference anymore
The BlueSky post has another interesting clue. The pencil sketch on the right. Seems possible a human artist drew the sketch, then had an AI model "colorize" it. And in so doing, maybe the AI model added the 3 genAI tells/artifacts I identified above.
the scarf matches the one Rich is wearing near the end of the documentary. the code matches what's on the Clojure source code, even if it's an odd choice. the two belts are in the draft, even if only one of them features the logo. the artist's instagram has a bunch of well colorized artwork. I believe it is really human made
As you demonstrated, AI is not needed to write slop, just because AI is involved doesn't make it slop. We are still very much in the control even if it is generation.
In a previous life, I wrote Clojure every day and still look back fondly attending Clojure/Conj and sitting next to Rich Hickey and other Clojure greats at dinner.
My first startup was all Clojure. AWS only had a dozen or two products and I think we must have been the first to compile Clojure to JS and run it on Lambda in production (the only runtime was Node.js 0.10 at the time).
Anyway, I cannot wait to watch this
What are you using now? how do you feel about clojure now?
I’m no longer in a professional engineering position and have a family and toddler so less time for learning about or playing with program languages. Clojure is what really made me first start thinking about programming languages and down the rabbit hole of scala and Haskell and idris and type theory books. Which are incredible and I can’t recommend more and also not remotely practical.
> also not remotely practical.
studying them makes you smarter, which makes you better at using the practical stuff.
How do you feel about it now? Like holistically.
I miss it. I’m now more removed and building software as a means to an end.
I miss it just like I miss the program language and type theory group meetups in SF and working through problems in dependently types languages like Idris and being out of my depth.
Excellent. The only thing I wish they had added was borkdude.
Yes, sad there's no European presence (unless I missed someone). There's tons of Clojure over there. Metosin, Juxt, Borkent, Gaiwan (Arne), Flexiana, Peter Strömberg (Calva), Dustin Getz (hyperfiddle/electric), Christophe Grand (ClojureDart), Bojidar (CIDER), Renzo, and many, many more.
didn't know datomic was free of licensing fees - I didn't use it back in the day because the cost was prohibitive... interesting
is clojure still relevant in the post agentic coding reality that opens up pretty much all esoteric languages to everyone ?
back in the day used to use clojure to write a fintech app but not sure if it is still relevant has uses vs other langs that have emerged
Clojure had lousy error messages, agents deal with this well. Clojure is capable of producing some of the most dense code I’ve ever seen, so manual code reviews really start to feel like a bottleneck unless your goal is to level up.
I find it actually the best substrate to write AI tooling. All my custom MCPs are written in Clojure (bb). You hook up the agent to the REPL and let it go wild - it builds something nice. Also, Clojure is one of the most token efficient PLs.
One of the main problems I have with the models coding is the feedback loop is way down the chain from generation, it's out at the commit boundary for python when your hooks are running, maybe at the point where the model wants to push a PR. The REPL lets that happen during generation, and the other safety measures help immensely. Immutable data, STM, all of the features in Clojure that gave devs super powers now do the same for a model.
Clojure might be the least esoteric language ever. Call a function, get a value.
It definitely is more "mainstream" than others but I just don't see the same level of attention and enthusiasm around it anymore. I'm sure it is still being used in many places but like Elixir, hiring remains on the tough end.
Hiring good talent was always problematic. This has nothing to do with the quality, capacity, and robustness of the language or its relevance. Hiring people who would love to use Clojure but have no prior experience is not that difficult - it's just that every company wants an expert, but they don't want to offer expert salaries. In places where they do, the competition is nuts, on top of that, experienced Clojuristas typically get interviewed and scrutinized with the same level of rigor as architects.
The industry should have optimized for hiring people interested in PLs like Clojure instead of LeetCode drillers. Clojure is rarely the first, second, or even third programming language people choose to learn. It demands a specific vision, dedication, and discipline that fundamentally transforms how people think about computation, data flow, distributed systems, and concurrency. The ROI from hiring an average developer experienced in Clojure has the potential to significantly exceed that of a typical hire. Even when there's zero Clojure in prod.
I've had Clojure on my resume for 10 years, mainly to see if anyone would ask about it. Nobody ever has, until an interview a couple days ago. We'll see if it actually helps in leading to an offer, I guess.
I have the opposite experience - been using Clojure for over a decade and it feels like only that mattered for the last five jobs. Even though it's really only just one of many layers that required to do the job. I honestly would love to find a non-clj team and convince them to use it. There are so many useful scripts we write in babaska alone, it just sounds wasteful not to use that path, fully knowing of its existence.
Clojure is more relevant than ever in post agentic coding because of immutability and the REPL. The two big problems with agentic coding is context growing in unbounded fashion, and agents being able to get quick feedback on what they're doing. Mainstream languages fail on both accounts. I've found Clojure has been a great fit for keeping agents on track.
I've wrote about this in more detail here if you're interested https://yogthos.net/posts/2026-02-25-ai-at-scale.html
Incredible: I had not idea NuBank discovered Datomic first and that it's Datomic that led them to Clojure, 100 million+ customers, and eventually acquiring Cognitect.
Good to see David Nolen (aka "swanodette") is in the documentary too.
As a bonus here's a recent talk from David Nolen about Clojure/ClojureScript and using DOM morphing instead of React.
If you don't want to watch it all, just take two minutes to watch from 23m15s to 25m15s. He compares a behemoth slurping all the browser's CPU and RAM resources versus a 13 Kb of JavaScript + Web components and DOM morphing:
https://youtu.be/BeE00vGC36E
His talk is presented from Emacs, gotta love that too...
I don't know if it's still the case, but at old clojure conferences, or meetups, or places of employment, emacs was a prereq and assumed (and the most enjoyable)
AI slop Rich is gross considering his stance on it. I guess it's up to the producers but very tone deaf.
You know there's a saying in Russian, that roughly translates to: "an expert surgeon is capable of helping a bad dancer", which is on itself is a reference to another idiom: "a bad dancer always blames his own balls".
That's quickly becoming befitting for cases like this - so often people rush to blame AI without even trying to use their own reasoning. I don't know what to say, hope you find a good surgeon, because it is obvious - you're shit of a dancer.
"A good craftsman doesn't blame his tools" might be the western equivalent.
The very official Clojure page in TFA links to clojure-mcp (written by the person who created figwheel: a famous ClojureScript library in the Clojure ecosystem) and other AI resources related to Clojure.
It's not because Rich doesn't want AI-generated pull-requests by people then taking credits that the Clojure community is anti-AI.
I use Claude Code CLI daily with Clojure, just not in a "write me five thousands lines of Clojure code I won't read" type of way.
Are you watching the same thing I am? What AI slop?
I think they mean the video thumbnail, which may or may not be AI-generated.
Rich commented in one of the comments under the video pointing the same thing and he says it is not AI generated
I don't think it is, considering they highlighted it in a post about human craft [1]. I read somewhere it was illustrated by felipemelo.net, but can't find the reference anymore
[1]: https://bsky.app/profile/cultrepo.bsky.social/post/3mjhubrh3...
It'll be interesting to learn whether it was AI-generated. It certainly SEEMS like it is. It has a few "tells":
- two belts and two Clojure logo belt buckles
- same code repeated on the steps (odd artistic choice if made by the artist)
- the seemingly out-of-place scarf, stylistically its color/pattern doesn't seem to fit
Either way, it seems like an homage to this Indiana Jones and the Temple of Doom poster:
https://www.slantmagazine.com/wp-content/uploads/2008/05/tem...
That wouldn't be very interesting at all. It's just a thumbnail on a niche programming documentary
The BlueSky post has another interesting clue. The pencil sketch on the right. Seems possible a human artist drew the sketch, then had an AI model "colorize" it. And in so doing, maybe the AI model added the 3 genAI tells/artifacts I identified above.
the scarf matches the one Rich is wearing near the end of the documentary. the code matches what's on the Clojure source code, even if it's an odd choice. the two belts are in the draft, even if only one of them features the logo. the artist's instagram has a bunch of well colorized artwork. I believe it is really human made
Huh. I don’t doubt you, but I guess this is a sign pervasive AI/LLM generative artwork is messing with my brain’s pattern matching at a deep level.
At least the cult repo folks said on LinkedIn that it is not, which tbh is surprising.
https://www.linkedin.com/posts/emmalouisetracey_one-of-my-fa...
As you demonstrated, AI is not needed to write slop, just because AI is involved doesn't make it slop. We are still very much in the control even if it is generation.