Jeremy Kun
banner
jeremykun.com
Jeremy Kun
@jeremykun.com
Portland-based mathematician and software engineer. Building a homomorphic encryption compiler at Google.

https://jeremykun.com
https://pimbook.org
https://pmfpbook.org
https://buttondown.email/j2kun
https://heir.dev
What always bothers me about discussions of "ethics" in software is that the punchline is always "and this cost their employer a lot of money."

I don't think employer expense is the primary ethical problem that software engineers should think about.
"Why do we have to take ethics classes for a computer science degree?"

This is why. And frankly, we should be failing more students for inadequately demonstrating their understanding of the topic.

Let's review some fun historical examples of why "No, actually quality will continue to matter."
December 27, 2025 at 4:21 AM
Reposted by Jeremy Kun
Fuck you people. Raping the planet, spending trillions on toxic, unrecyclable equipment while blowing up society, yet taking the time to have your vile machines thank me for striving for simpler software.

Just fuck you. Fuck you all.

I can't remember the last time I was this angry.
December 25, 2025 at 11:25 PM
Reposted by Jeremy Kun
Reposted by Jeremy Kun
me: how do I think of this complex problem? help please

mathematician, normally: how do you feel about geometry??
December 21, 2025 at 12:14 PM
Reposted by Jeremy Kun
Classical billiards can compute.

With Isaac Ramos, we show that 2D billiard systems are Turing complete, implying the existence of undecidable trajectories in physically natural models from hard-sphere gases to celestial mechanics.
Determinism ≠ predictability. 🎱🧠 @upc.edu @ricardsole.bsky.social
December 19, 2025 at 8:16 PM
I've heard Rad Power bikes has always had troubles, so the bankruptcy is no surprise, but seeing how much of their unsecured claims is attributed to tariffs makes it pretty clear that Trump's policies had a major impact on them.
December 16, 2025 at 8:05 PM
What is the likelihood that my guess of the probability of this event is at least 1/3?

I'd say it's pretty likely.
December 14, 2025 at 3:12 AM
Reposted by Jeremy Kun
I saved this earlier and didn’t know what to do with and now I know what to do with it
October 2, 2025 at 7:28 PM
I handle solitude very well, but the loneliest feeling I know of is to be traveling by myself in a dense metropolis of a foreign country where I don't speak the language.

Something about being solo in never ending crowds of people I can't communicate with... It's a vibe I'm not here for.
December 11, 2025 at 5:05 PM
Is it me or do status pages just not work properly anymore? GitHub status page shows all systems good, but I can't access the web UI.
December 9, 2025 at 10:00 PM
Cursed Cunningham's Law:

The fastest way to get help from an expert is not to ask a question, but to claim an LLM outperformed the experts.
December 9, 2025 at 7:30 PM
Ah yes, the motivational posters in my hotel room, including "You matter" and "sunlight is every color combined"
December 8, 2025 at 2:53 PM
My 5yo just informed me, "I think I have succeeded on my new year's revolution"
December 7, 2025 at 3:16 PM
You know your corporate computing environment is weird if the only way to use python3.11 on linux is to compile it from source.
December 5, 2025 at 9:49 PM
Since the core value proposition of generative AI is to replace labor, I'd say wealthy people are going to become more dependent than anyone on it, because that's how they, in theory, avoid the messy and expensive business of employing people to do things for them.
The first school to market itself as AI free is going to corner the market on people interested in actually learning. And I would not be surprised if rich families and the children of people creating this tech were the first movers.
My employer, Dartmouth College, today boasts it's 1st Ivy "to launch AI at an institutional scale." It is doing this by partnering--"more than a collaboration"--with Anthropic, a company that stole the books of many faculty, me included, which many of us are suing.
December 5, 2025 at 5:00 PM
126 buckeyes made. A bit early for holiday cookies, but we've got a new neighbor and their house ain't gonna warm itself!
November 30, 2025 at 10:51 PM
Oh dear, I just saw an academic paper in which the baseline, described as "naive search," was just querying an LLM...
November 30, 2025 at 8:33 PM
Woke up this morning to some fresh identity theft! Sucks but glad I caught it early.
November 29, 2025 at 7:04 PM
FWIW, I don't think my writing from 15 years ago is good either 🫠
November 27, 2025 at 8:42 PM
"Freedom Foundation" sent a mailer to my teacher wife to ask her to opt out of union dues, implying that the right thing to spend that on is to donate it.

Fuuuuuuck them
November 27, 2025 at 3:52 AM
Reposted by Jeremy Kun
(As it turns out, the Torment Nexus had already been considerably improved in [Schmidhuber, 1993])
November 24, 2025 at 8:40 AM
Reposted by Jeremy Kun
Trying to think of something serious to say about the “cryptographers lose the key for the cryptographer election” story and, mostly, hey: I just love that cryptographers are actually using the weird cryptography! www.nytimes.com/2025/11/21/w...
Cryptographers Held an Election. They Can’t Decrypt the Results.
www.nytimes.com
November 22, 2025 at 2:47 PM
LLMs will end all disease and double the human lifespan! With all that extra time, and no useful activities left for humans to do, you'll have plenty of time to watch seasons 430-457 of "The Deepfaked Housewives of Datacenter 3f56ea9" on Amazon Prime!
November 21, 2025 at 4:21 PM
Reposted by Jeremy Kun
I've published a blog post here, with implementation details and metrics:
recursion.wtf/posts/detect/

this may not come as a surprise, but it not spawning a new process via `-exec` for every file or running multiple `xargs` passes makes everything significantly faster
November 18, 2025 at 6:57 PM