jsm
Jonathan

https://jonathansm.com

Hacker, cypherpunk. Bitcoin is cool. All memes are my own.

Public Key

npub1uqeexjx2djkfwzxdnrnrrch5h2k4xn0uapcgsxm94ftaxrlhy5lqywjckg

NIP-05 Address

_@jonathansm.com

Profile Code

nprofile1qqswqvunfr9xetyhprxe3e33ut6t4t2nfh7wsuygrdj6547nplmj20sppemhxue69uhkummn9ekx7mp0f8sgms

Publishing to

nos.lol

Author Public Key

npub1uqeexjx2djkfwzxdnrnrrch5h2k4xn0uapcgsxm94ftaxrlhy5lqywjckg

Show more details

Last Notes

2025-04-04 03:13:21

https://www.astralcodexten.com/p/introducing-ai-2027
Scott Alexander’s post on impressive AI predictions from 2021. Super impressive stuff considering ChatGPT hadn’t been released yet.
The one major thing that I think Scott is mistaken on in his blog post is his idea that China will “wake up” and steal US weights. The Chinese don’t need to. Alibaba (Qwen), 01 ai (Yi), and DeepSeek are already toe to toe with the best American models. Just as we figured out in 2024 that OpenAI has no moat, we’ve learned in 2025 that neither do American companies.

2025-02-16 17:53:24

- reply

Veganism is a relatively popular diet. I agree with you that veganism for dietary reasons is not the best option, but I don’t understand how this is a psyop.

2025-01-07 19:58:30

I’m a simple man.
I read Wikipedia article of wall named Ha Ha because people would laugh when they tripped over it.
I go “Ha Ha”
https://en.wikipedia.org/wiki/Ha-ha

2025-01-02 19:46:06

The 21st century is a quarter over now. Will I get to see the end of it?

2025-01-01 22:16:50

And then it ended so fast.
#note1djh…ur2r

2025-01-01 22:11:06

How does the Texas v Arizona State game just keep getting more insane? It’s been over so many times, and yet the game keeps going.

2024-12-29 16:19:19

https://graymirror.substack.com/p/migration-and-the-sovereign-firm
Interesting essay on why hiring H1Bs doesn’t make sense under Yarvin’s pre-liberal framework. Obviously as a proud classical liberal in the tradition of Locke which Yarvin despises, I have massive disagreements. Still an interesting read to even think about other non-liberal frameworks.

2024-12-28 22:21:48

Update: now bleeding after achieving liftoff from my seat while driving and chomping on my tongue when my teeth clacked back together.
At least the Midwest has actual weather and ice to blame for potholes.
#note1gtc…uewp

2024-12-28 22:16:36

- reply

Agreed. California is redpilling me on libertarianism. I don’t want America to turn into this.

2024-12-28 22:13:44

https://www.johnfoy.com/research/which-states-have-the-best-and-worst-roads-in-america/
I was more right then I imagined. According to this study, California is the fourth worst state for roads. The roads here are bad even in urban areas.
I wouldn’t care if they marketed themselves as a libertarian paradise where government services are low in tandem with the tax burden. But that’s the opposite of California. If the government wants to expropriate lots of money, I expect the services they provide to reflect that.
Californians, what’s going on here?
#note18v8…xvyf

2024-12-28 21:54:42

It takes a special kind of stupid to govern a state with incredible natural resources, industry, the largest tech sector in the world, near perfect weather, mountains, oceans, shipping, tourism, and high taxes and still manage to have horrible public infrastructure.
California has some… special politicians.

2024-12-28 21:51:23

- reply

Welp fuck there’s always a few bad eggs but don’t worry, you’re just the exception that proves the rule. Most Nostriches are intelligent enough to sense the massive amount of cognitive dissonance these two contrary positions entail.

2024-12-28 21:42:27

So glad Nostr is talking about more important things than trying to pretend to be anti-woke and anti-DEI while simultaneously saying that companies should prefer hiring native born Americans over more skilled Indians because Indians are over represented in the work force.

2024-12-27 17:13:25

https://openai.com/index/why-our-structure-must-evolve-to-advance-our-mission
It’s incredible how much time and political maneuvering it took Sam Altman to get to this point. He took on the entire board and research scientists for every major department and somehow came out the winner. This reads more like an announcement of victory than anything else. It means Sam won. He’s going to do away with the non-profit charade and accept the billions in investment to abandon the vision and become a commercial AI company.

2024-12-25 06:17:00

Merry Christmas btw

2024-12-25 06:15:35

Making a Prometheus storage backend and Grafana dashboard for this data would be legitimately fascinating.
#note1dnz…yrvc

2024-12-25 05:58:32

https://iss-mimic.github.io/Mimic/
TIL you can watch every possible piece of telemetry live from the ISS. So I'm spending my Christmas Eve watching the urine tank fill percentage while I wait for the clock to strike midnight.
Twas the night before Christmas, and all through the sky,
The ISS floated, its tanks running dry.
I stared at the dashboard, no presents, no tree—
Just waiting for pee logs from zero-G.

2024-12-25 04:38:01

I don’t understand how people can bear the modern internet without an adblocker. I don’t think I could visit a website and stay sane without one. How would you even watch YouTube with all the ads or open a news article and get bombarded with a million popups and not want to ever load a webpage again?

2024-12-25 04:22:57

- reply

Me too lol. Let’s change that.

2024-12-25 03:30:38

How did I just learn about this method for computing whether insurance is worth it? If computing whether or not insurance is worthwhile is basic maths, it seems like this should be more widely known.
https://www.lesswrong.com/posts/wf4jkt4vRH7kC2jCy/when-is-insurance-worth-it

2024-12-25 03:28:12

Looking back over the year, one of the skills I think improved the most was developing a better gut understanding of probability and how likely events are to happen.
One of my goals for next year is to learn more statistics to be able to mathematically analyze the world and my life.

2024-12-25 03:20:40

- reply

It’s not completely free, just a trial version with a limited number of requests. That seems to be the business model most AI companies (OpenAI, Anthropic, etc.) have settled on. The first few requests per day are free but if you want unlimited requests you need a subscription.

2024-12-23 18:06:58

- reply

That was very fast!

2024-12-23 17:04:56

- reply

Oh sick. I’ve always fantasized about having a display somewhere that shows interesting visualizations etc., kind of like a digital fish tank. I’ll need to look into that.

2024-12-23 17:03:27

I’m finally outgrowing my home server setup of Raspberry Pi’s I’ve accumulated over the years. If I want to do it right, how should I set up my home server network?
Currently running a little website, some other website side projects and a Jellyfin media stack.

2024-12-23 17:00:39

- reply

What is it? Some kind of smart mirror?

2024-12-23 16:24:03

Wake up babe, the Gaetz ethics report just dropped.
https://www.nbcnews.com/politics/congress/matt-gaetz-report-full-document-house-ethics-committee-rcna184703

2024-12-23 15:35:21

Interesting article on UFO “flaps” and the recent New Jersey hysteria.
https://computer.rip/2024-12-21-something-over-New-Jersey.html

2024-12-19 21:15:25

> be matt gaetz
> nominated for attorney general
> drop out so ethics report isn’t released
> they release it anyway
Should have just been attorney general anyway. He picked the worst possible option: no attorney general and the media has a feeding frenzy on your pedo and porn habits.
https://www.cnn.com/2024/12/18/politics/matt-gaetz-ethics-report-committee/index.html

2024-12-19 18:30:30

Xitter DDOSing itself? I’m not touching the screen at all in the recording. This happens whenever I open a link.
https://m.primal.net/NAuT.mov

2024-12-19 18:16:18

After Google’s new quantum computer came out there seem to have been different two reactions from experts:
1. This doesn’t prove parallel universes
2. Of course this doesn’t prove anything new about parallel universes, we already proved them
Now I’m more confused. Is the parallel multiverse theory to explain quantum mechanics the current accepted theory, or is it still in debate? I don’t have the physics chops yet to deeply understand everything myself so the differing views on whether or not this is even a debate are hard to rectify.

2024-12-18 15:05:00

Woah new linear sort just dropped.
https://imgs.xkcd.com/comics/linear_sort.png

2024-12-17 21:22:53

Just want to say my reasoning was rock solid and I was vindicated by posterity. Assuming of course that Luigi is actually the killer which seems overwhelmingly likely.
#note1x7r…kcvu

2024-12-17 21:19:06

Side note: I know Nostr tends to lean towards “fuck the system” but can we all take a moment to be grateful for how good institutions in the U. S. are? For all of our gripes with mainstream media, none of them would ever write this level of QAnon conspiracy genitalia theft insanity. We have problems in America, but our institutions are at least mostly functional and truthful.
#note1m2a…rpu8

2024-12-17 20:44:51

This is a modern crisis. Forget the drug cartels and fentanyl, we need to crack down on the huge penis stealing crimes rings.
https://thenationonlineng.net/locals-allege-french-involvement-in-car-missing-genital-crisis/
#note1f2w…0csr

2024-12-17 20:36:40

You’ve heard of elf on a shelf.
Now get ready for a hex on your sex.
https://www.astralcodexten.com/p/book-review-the-geography-of-madness

2024-12-15 04:55:16

Interesting ideas here from Meta. Tokenization has a lot of problems with how we represent text. You can get weird glitches etc. Working with “dynamic patches” could allow more reasoning where necessary from the model and avoid tokenization glitches and oddities.
https://ai.meta.com/research/publications/byte-latent-transformer-patches-scale-better-than-tokens/

2024-12-11 15:05:20

It’s not over til it’s over, the Onion may still be able to do the funniest thing possible.
https://www.nbcnews.com/tech/internet/bankruptcy-judge-rejects-onions-bid-buy-alex-jones-infowars-rcna183453

2024-12-11 00:10:20

When I read this post Google asks straight up if I think better of Google after reading it. Refreshingly direct.
I wonder how one programs a quantum computer. Is it a chip running normal CPU operations but quantum, or does it require a degree in Quantum Mechanics to understand how to run a simple program?
https://blog.google/technology/research/google-willow-quantum-chip/

2024-12-10 23:59:05

Since debanking is all the rage now, one of my favorite finance writers published an essay on it.
https://www.bitsaboutmoney.com/archive/debanking-and-debunking/

2024-12-09 15:53:47

- reply

I love drone shows so much. I’m a little surprised tho that they need to manually set them out. I would have assumed that if the drones can automatically fly in cool shapes then they can sort themselves out on the field.

2024-12-08 16:37:51

They might be pictures of different people. Perhaps the assassin was actually competent. This is a blow to my second reason. I’ll go with a 25% chance it was a professional.
https://manifold.markets/zax/are-these-two-supposed-pictures-of
#note1x7r…kcvu

2024-12-06 19:18:40

- reply

That seems like a big assumption. I don’t know why you would lose your preference between revenge and freedom or revenge and being in jail if you felt your child was killed by UnitedHealthcare. I think you would want to remain alive and free after getting your revenge.

2024-12-06 16:23:25

Probability seems to be leaning further against the killer being a trained assassin.
From first principles, if we simply look at the base rate of the number of people who probably feel wronged by UnitedHealthcare vs the number of powerful people who would call a hit on Thompson the probability would lean against trained assassin.
Second, perhaps I overestimate assassins because of movies, but I find it hard to believe a professional killer would go to Starbucks with his mask down right before the killing. They obviously were reasonably trained with their pistol and planned the assassination better than the average killing, but it still seems like an average person tried their best instead of a professional.
I think there’s a 20% chance it was a professional assassin.
#note1una…jlf4

2024-12-05 15:58:32

Interesting market.
https://manifold.markets/tedks/what-were-the-motives-of-the-united

2024-12-05 15:57:34

Last night was a good night. Bitcoin hit 100k, the Advent of Code puzzle is still tickling my brain, and I compiled Portals for 64 bit macOS and played it.

2024-12-04 18:25:48

Submit your bets here
#note1rld…0hmn

2024-12-04 18:24:35

Now you can vote on it!
https://manifold.markets/jonathan21m/will-roger-ver-go-to-prison-again-b
#note1m5z…gcwh

2024-12-01 16:23:02

- reply

https://m.primal.net/MlZA.jpg

2024-12-01 16:20:11

Whoops the clipboard didn’t sync from my MacBook to phone. Here’s the correct link.
https://github.com/realprogrammersusevim/aoc-2024
#note1ssx…0dhp

2024-12-01 16:18:43

That ended up being easier than expected. Nice way to start the morning. My solutions are here: https://adventofcode.com/2024/day/1
#note13d8…rsnv

2024-12-01 15:19:43

For anyone who wants some programming puzzles, the Advent of Code for this year is out. I’m just about to make a start on day one.
https://adventofcode.com/2024/

2024-12-01 15:17:14

https://apnews.com/article/artificial-intelligence-chatbot-jesus-lucerne-catholic-66268027fbcf4b48972d1d62541f0b16
https://m.primal.net/MlWf.jpg

2024-11-21 19:32:21

@npub12vk…pugg @npub16c0…6nvr is there any way to set up NWC in Primal on iOS to connect to my Get Alby Hub?

2024-11-21 19:14:43

Against all odds I appear to have set up the Alby Hub from @npub1get…0nfm correctly the first time. Count me surprised. It’s a testament to the ease of use.
#note14mw…as3y

2024-11-21 19:11:13

Thanks @npub150q…en2h!
#note14mw…as3y

2024-11-21 19:10:20

- reply

Hmm that’s odd. I just got a zap from someone else. Is there an error displayed?

2024-11-21 19:05:13

Can someone test zapping me? I just set up the Alby Hub and I’m not sure if it’s running correctly. Make sure it’s a small amount though because I’m not keeping much in a hot wallet that exposes an API endpoint.

2024-11-21 17:24:39

Just overheard someone listening to a podcast on pop culture. I can now confidently say that 80% of podcasts could easily be replaced by AI.
With a little fine tuning and voice cloning we could replace the humans and keep pumping out the same vapid and bland content without anyone being the wiser.

2024-11-16 21:31:46

Learned today that if you stop checking Snapchat and turn off notifications from the app, they will start texting you the notifications.
I stopped the messages. I think I probably have a day or two before I get sent my notifications in the mail or someone shows up at my house to tell me in person.

2024-11-16 05:24:34

Mike Tyson is looking pretty slow on his feet. He can’t get his balance back quickly.

2024-11-16 05:09:07

I TAKE IT BACK, MIKE STILL HAS IT.
#note1qky…dk4d

2024-11-16 05:01:20

I dunno, watching Iron Mike beat the snot out of Jake Paul would be so therapeutic but Mike Tyson is looking kinda old.

2024-11-16 03:04:46

Nope, the whole world can just unite around the common hope that the old Iron Mike comes out and beats the shit out of Jake Paul live.
Thoughts and prayers everyone 🙏🏼
#note1vf9…j69m

2024-11-15 23:21:55

Me applying for scholarship and grant money.
https://www.youtube.com/watch?v=5NeS4ueaU6w

2024-11-15 21:01:48

- reply

Oh no, I really hope the sale goes through. The funniest outcome is the best here.

2024-11-15 17:16:24

Here’s the Op Ed on why they bought it.
https://theonion.com/heres-why-i-decided-to-buy-infowars/
#note10c8…hsd3

2024-11-14 17:18:45

The Onion just did the funniest thing imaginable.
https://www.nytimes.com/2024/11/14/business/media/alex-jones-infowars-the-onion.html

2024-11-05 18:31:38

- reply

I run it on a Raspberry Pi with a custom server written in Rust.

2024-11-01 20:03:29

How are the election predictions this insanely close 4 days before the election?
https://www.metaculus.com/questions/11245/2024-us-presidential-election-winner/

2024-11-01 20:00:20

It feels good to finally self host my website on my own server rather than on GitHub Pages. Now I have complete control.

2024-10-15 22:11:22

I just bought a smart ring for $10 that’s completely hackable over BLE. It has an accelerometer, Blood Oxygen meter, and a heart rate monitor. The question now is: what should I do with it? Any ideas?

2024-10-15 15:24:00

Who will be next? Apple? Facebook?
#note19k0…uyp4

2024-10-15 15:12:56

Google is also joining the Nuclear party.
https://archive.is/2024.10.14-192552/https://www.wsj.com/business/energy-oil/google-nuclear-power-artificial-intelligence-87966624
#note1khj…xqt3

2024-10-15 15:06:07

With the amount of energy drinks teens are drinking, will caffeine eventually become a national strategic resource?

2024-10-13 04:01:47

Browser extension idea: what is the probability the page I’m looking at was AI generated?
Or a crowdsourced list of AI slop websites.

2024-10-10 15:17:42

Since I’m too cheap to rent a nice server from Digital Ocean I had to rewrite some web apps from Python to Golang to save memory. The performance and RAM savings are dramatic. While Rust is probably still my language of choice for most non-trivial projects Go makes writing backends incredibly easy.

2024-10-09 14:05:17

The regular Transformer may have just been unseated as the de facto architecture for LLMs. I’m no expert but Diff Transformers seem like a massive improvement over regular Transformers.
https://arxiv.org/abs/2410.05258

2024-10-02 15:22:28

- reply

Psyops are a psyop

2024-10-02 15:21:24

Hats off to JD Vance, I may not agree with him on a lot of his new political opinions since he’s cozied up to Trump but he’s a smooth debater. From the very first question you could see he was much better in that environment.

2024-10-01 15:51:43

This is actually super entertaining.
https://www.reddit.com/r/notebooklm/comments/1fr31h8/notebooklm_podcast_hosts_discover_theyre_ai_not/
#note1cd4…h7d5

2024-10-01 15:49:52

I know people complain that the podcasts NotebookLM generates are a little vapid and bland, but so are 90% of podcasts.
Most podcasts probably shouldn’t exist in the first place because neither the people nor the topics are that interesting, and of the ones that should exist most could probably be replaced with AIs. An LLM could probably handle interviewing the millionth author of some shitty book that no one is going to read anyway. Hell the LLM could probably handle pretending to be the author as well.
https://simonwillison.net/2024/Sep/29/notebooklm-audio-overview/

2024-10-01 15:36:32

Wait wtf, California’s idea of “banning” something is putting you on a public list if you do it and then doing literally nothing else? If there’s no teeth to a bill why even pass it in the first place?
https://archive.ph/2024.09.30-175714/https://www.nytimes.com/2024/09/30/us/california-bans-legacy-admissions-private-universities.html

2024-10-01 15:23:30

macOS Sequoia and iOS 18 feel like gimmick updates with little toys because the real meat of the updates (Apple Intelligence) haven’t been released yet.

2024-10-01 15:11:46

At this point maybe we’re just reinventing HTTP/TLS but would multi-part notes help solve the note size limit problems? Its kind of annoying that you can’t follow more then around 1000 people or make a big list before a lot of relays start rejecting the notes.

2024-10-01 14:53:42

I love his description of how riding a driverless car feels like a fast firmware update for your brain.
https://allenpike.com/2024/distribution-of-the-future

2024-10-01 03:08:00

Just updated to MacOS Sequoia. The iPhone mirroring feature is pretty sick. I’m typing this on my Mac into Primal running on my phone which is currently in my pocket.

2024-09-28 20:10:17

Now terrorists get sympathetic pieces in the WaPo when they die?
https://m.primal.net/LEtE.jpg

2024-09-28 19:53:57

https://hachyderm.io/@robpike/113212958758344294

2024-09-28 18:53:03

Borrowing you friends family for photo shoots is kinda weird and shocking the first time you read about it, but then you realize politicians probably do this kind of thing constantly.
https://www.thedailybeast.com/childless-gop-candidate-derrick-anderson-borrows-family-for-weird-photo-shoot

2024-09-28 18:49:04

Does anyone know what happened to the Babylon Bee? Maybe it’s a temporary phase over election season but they almost completely stopped being funny. Everything is either “Kamala Harris is awful” or “Dems are trying to kill Trump” and most of them time they don’t even have a joke to go along with it.

2024-09-28 04:38:10

- reply

The privacy, ability to uncensored, and control that models with open weights provide would seem to be a huge boon over OpenAIs offerings. It feels like AI is a little over-saturated at the moment with not much differentiating the top dogs.

2024-09-28 04:08:53

And there at last is the reason everyone is leaving OpenAI. Sam is finally making his power grab.
https://www.reuters.com/technology/artificial-intelligence/openai-remove-non-profit-control-give-sam-altman-equity-sources-say-2024-09-25/
#note1792…hygz

2024-09-27 16:48:34

New favorite art
http://donatoarts.com/gallery

2024-09-27 16:47:46

- reply

I’m not completely sure. It may still be the fallout from Sam Altman pruning everyone who wasn’t 100% on board the Sam Plan or maybe there’s something else entirely happening.

2024-09-27 16:45:29

- reply

Sure, he still runs a company that requires gathering the most information possible, but that doesn’t change the fact he got kinda cool at some point.

2024-09-26 13:59:04

And yet another decides to leave. This one was the GPT-4 co-lead.
https://x.com/barret_zoph/status/1839095143397515452
#note1yv8…xfs8

2024-09-26 13:53:38

When did Mark Zuckerberg become kind of cool?
He went from the Social Network jerk who just wants your data to a competent martial artist who has interesting thoughts and still kind of wants your data.

2024-09-26 13:50:51

- reply

Exactly, open source models are crushing the closed source competitors. I really hope this continues, it would be a real shame if AI becomes centralized in the hands of the few.

2024-09-26 03:27:41

OpenAI shedding their top people quickly. Ilya, Andrej, Jan, Greg, John, Peter, and now Mira are all out and those are just the ones I can remember off the top of my head. That’s the heads of their research and safety teams and all the top executives.
https://twitter.com/miramurati/status/1839025700009030027