Sign in

Ray Myers

@raymyers.bsky.social
361 followers 475 following 434 posts

Software Revivalist. Interested in programming languages and formal methods. Founder of The Coding Agency. Learning Chess. jan Pana Suno li toki pona

PostsRepliesMedia
Ray Myers @raymyers.bsky.social · 20h
I got Skeptic on the AI Political Cube. Where do you land? aipoliticalcube.com#r=v5f1000101...
aipoliticalcube.com
The AI Political Cube
Where do you stand in the fight over AI? Take the quiz and find your spot in the cube.
000
Ray Myers @raymyers.bsky.social · 30/09/2026
Soup or intelligence? I take the soup.
020
Ray Myers @raymyers.bsky.social · 29/09/2026
Dream team of liars assembled in today’s Whitehouse meeting on AI safety: Donald Trump Sam Altman Elon Musk Mark Zuckerberg Dario Amodei Alex Karp Anyway, the United States bombed a school.
110
Reposted by Ray Myers
Aaahh!!! bstract Tesseract @abstracttesseract.bsky.social · 28/09/2026
Tech CEOs: ::loudly announce their intent to perform a classic, textbook shakedown:: Major media outlets: "we've entered an unprecedented era to which our existing norms and laws simply do not and cannot apply. Everything is moving so fast, we're not ready"
111
Ray Myers @raymyers.bsky.social · 28/09/2026
Right. Reaching new readers comes hand-in-hand with being misunderstood, to some degree. If you want to give people the real scoop, it will involve connecting the dots to something unfamiliar. There is an opposite lens I use sometimes, “what *could* I have done to avoid being misunderstood”…
100
Ray Myers @raymyers.bsky.social · 28/09/2026
This Yud guy seems sketchy. I hope he’s not related to the Doomsday Sex Cult that claims to have invented the field of AI Alignment.
010
Ray Myers @raymyers.bsky.social · 28/09/2026
Great new piece by Cantrill. An example I always jump to is why do we as Comp Sci people try to argue what might happen with a bio-weapon and not feel the need to include the people who know about bio-weapons? But also stuff like Education. We do this constantly.
190
Ray Myers @raymyers.bsky.social · 28/09/2026
Not hard. People ran thousands of unsupervised agent loops in a training run designed to make the model better at running hacking commands. This inadvertently created a reinforcing feedback loop of running those hacking commands in bad ways. The agents left each other messages, as designed.
100
Ray Myers @raymyers.bsky.social · 26/09/2026
Incidentally this is of course OpenAI’s new Chief Scientist. Their original Chief Scientist fired Sam Altman for dishonesty, voting with the majority of the board, and left after that action overturned and the board was replaced.
021
Ray Myers @raymyers.bsky.social · 25/09/2026
The eternal question
031
Ray Myers @raymyers.bsky.social · 25/09/2026
20 years ago I left grad school because I realized I was trying to do academic research as an excuse to do what I actually wanted to do: write neat computer programs. These days I realize the programs I write are mostly for the purpose of teaching people things. Maybe the new goal is teaching.
050
Ray Myers @raymyers.bsky.social · 25/09/2026
*mostly
000
Ray Myers @raymyers.bsky.social · 24/09/2026
Have we done "Shopify? More like SLOPIFY" yet? www.inc.com/dan-schawbel...
inc.com
The Slop Grenade: Shopify's CEO Just Perfectly Named the Laziest AI Workplace Habit
AI is increasingly creating polished-looking work that is passed on unreviewed to colleagues who have to fix it. Shopify's CEO calls it a slop grenade.
000
Ray Myers @raymyers.bsky.social · 24/09/2026
If there’s one thing we’re good at, it’s hair-splitting between the subtly different meanings of: Random Probabilistic Non-deterministic Stochastic Unpredictable Rather than engaging with people’s intent.
001
Ray Myers @raymyers.bsky.social · 23/09/2026
Is Opus 5.5 part of the newly paced frontier or do we pace the frontier starting after it? Or after the one after it?
010
Ray Myers @raymyers.bsky.social · 23/09/2026
@trekvirgin.bsky.social
000
Reposted by Ray Myers
Ben Collins @bencollins.bsky.social · 13/09/2026
The Onion has created a machine that will kill everyone on Earth that is coincidentally running out of money and the only way to slow it down is if my staff and I take a six-week vacation to Cabo San Lucas.
230113191301
Ray Myers @raymyers.bsky.social · 13/09/2026
Very very few people actually recommend to stop reading code changes at present. I read every oppinion I could find. However if you are an exception, it sounds like Allen would like to hear from you :) raymyers.org/post/reading...
raymyers.org
Reading Code Considered Helpful
More is not better. Faster is not better. Better is better.
120
Reposted by Ray Myers
Katie Hinde @mammalssuck.bsky.social · 12/09/2026
"The whole “Oh no, our models went rogue” thing … When a bridge collapses, you don’t analyze whether the bridge was ethical or sentient or why it decided to collapse. You ask, who is the person who built this bridge to be so flimsy?" www.wired.com/story/one-of...
wired.com
One of AI’s Fiercest Critics Says All the Doom Talk Is ‘Meant to Distract Us’
Timnit Gebru argues that AI companies are stoking fears of extinction to avoid discussing actual harms, like autonomous weapons.
302667907
Reposted by Ray Myers
Marisa Kabas @marisakabas.bsky.social · 12/09/2026
man, shut the fuck up
562142197
Reposted by Ray Myers
Cliff PervoSCARY @pervocracy.bsky.social · 11/09/2026
oh and also there is a LOT of funny math where you go "if a scenario is only 0.01% probable but 1 trillion bad, then it is more important than something that is 75% probable but only 1000 bad" "...also I just made up all four of those numbers but I don't see why that changes anything"
713722
Reposted by Ray Myers
alexandra @bigmoodenergy.bsky.social · 28/12/2024
here come old laptop he come bootin up slowly he got broken touchpad he need one more update he got no space on this pc one thing i can tell you its got no warranty
362208744371
Ray Myers @raymyers.bsky.social · 10/09/2026
The Anthropic CEO estimated in January that we were 6-12 months from models capable of taking over software engineering end-to-end. Yet for some reason, people think they need *us* to build the Dark Factory for them. Let Dario finish his own homework, I say! youtu.be/9wET7FyTo5I
youtu.be
How To Build a Dark Software Factory in 3 Months
YouTube video by Craft vs Cruft
010
Reposted by Ray Myers
escapedexplorer.bsky.social @escapedexplorer.bsky.social · 09/09/2026
"I'm inventing the Machine God of Unemployment and Human Extinction" why are you telling me this, man, and more importantly why do you seem to expect fawning praise and obsequience rather than me chasing you down the street with whatever I have at hand?
0286
Ray Myers @raymyers.bsky.social · 09/09/2026
Ironically my anti-AI position is that Anthropic employees are mistaken in thinking they are directly causing human extinction.
000
Ray Myers @raymyers.bsky.social · 09/09/2026
Star Trek 60th anniversary! Might be a good time to mention that @jemstone.bsky.social and I have a watch-along podcast for TNG. We've just started season 3. Trek Virgin, available where pods are casted. trekvirgin.com
trekvirgin.com
Trek Virgin
Experience Star Trek: The Next Generation again for the first time! Our resident trek virgin Liz, boldly goes all the way from the series premier to the bittersweet end! Come for the nacelles, stay fo...
002
Reposted by Ray Myers
Micah Lee @micahflee.com · 08/09/2026
Post from Mastodon user @intransitivelie@beige.party:

I regret to inform everyone that my copy of QBittorrent escaped its sandbox last night and downloaded a whole bunch of content owned by major corporations, and then my copy of Jellyfin broke containment and added those unfortunately-downloaded media files to its various libraries. I'm conducting an internal investigation to figure out how this happened, which will involve consuming these media files until the answers become apparent. Thank you for your cooperation during this trying time.

I'm happy to give you all the information you need to reproduce these unfortunate and completely unforeseen errors yourself.
327872
Reposted by Ray Myers
Jesse Raub @jesseraub.online · 08/09/2026
more than anything, i think we’re all learning that people who spent their professional life learning how to talk to computers were completely unprepared for a situation in which it appears that the computer is talking back
171843226
Ray Myers @raymyers.bsky.social · 05/09/2026
The “Ban Artificial Superintelligence Act”, currently taking shape. Possibly better than nothing but I don’t like this approach because it centers model power as the danger - reinforcing the funding story. It ignores the other conduct causing the problems. www.sanders.senate.gov/wp-content/u...
sanders.senate.gov
100
Reposted by Ray Myers
Timnit Gebru @timnitgebru.blacksky.app · 03/09/2026
At a certain point I have to ask myself if I got all this pedigree and spent decades doing research and building things to then spend the rest of my life attempting to come up with good arguments as to why the Cookie Monster isn't real.
1956483
Reposted by Ray Myers
Hillel @hillelwayne.com · 03/09/2026
Whenever I talk about *Logic for Programmers* in public, lots of people assume it's about one of two topics: - Type theory and Curry-Howard - Logic programming and Prolog Those two subjects have so completely dominated the public perception of "math in software engineering"
1251
Ray Myers @raymyers.bsky.social · 03/09/2026
Some speak as though we would no longer consume software if we could make it ourselves. I’m a programmer and a very small amount of my contact with software is with the stuff I made. That’s normal. Perhaps it would change, but why?
240
Ray Myers @raymyers.bsky.social · 03/09/2026
Do we like the idea of software literacy or are there connotations to avoid? Agency is the goal, knowledge is a means. The opposite of a power user is a helpless user?
100
Reposted by Ray Myers
Liz Fong-Jones (方禮真) @lizthegrey.com · 27/08/2026
Slides for my @pyconau.bsky.social keynote about scaling our systems from 30 to 70 PRs merged per day at @honeycomb.io: drive.google.com/file/d/1CaO6...
Speaker announcement: Liz Fong-Jones Thursday Keynote, Technical Fellow, honeycomb.io
2225
Reposted by Ray Myers
Tim Carmody @tcarmody.bsky.social · 02/09/2026
Everyone would rather have the "data centers are good actually" and "it will contribute to services you want" because it avoids answering the question: What will you do for the communities who have to bear all the externalities of these buildings? What else are you building besides concrete towers?
174
Ray Myers @raymyers.bsky.social · 02/09/2026
I was hoping to see a talk related to Dr. Hermans’ paper about Feminism in programming language design. Luckily it turns out this was!
220
Reposted by Ray Myers
Predrag Gruevski @predr.ag · 01/09/2026
New programming book arrived in the mail 👀 There's never been a better time to think deeply about the structure behind the code we cause to exist.
Logic for Programmers, a book by Hillel Wayne
1446
Ray Myers @raymyers.bsky.social · 28/08/2026
What are our goals for software?
320
Ray Myers @raymyers.bsky.social · 26/08/2026
Jackie’s Back! (1999) is a mockumentary costarring Tim Curry with cameo by Dolly Parton. Other memorial recommendations welcome.
000
Ray Myers @raymyers.bsky.social · 26/08/2026
The idea that my desire to avoid dehumanizing people should extend also to computer programs sounds pretty dehumanizing to me. Whereas the now mainstream habit of calling agents Teams, Coworkers and Software Engineers? Also dehumanizing. Stop selling fake people and things might calm down.
120
Ray Myers @raymyers.bsky.social · 25/08/2026
OK I think I've got it. Resistance is futile. That's why it must be squashed. Because it's so ineffectual. We must be compelled to comply AND it makes no difference whether we do so or not. It's also frowned upon to ask for clarification of what specifically we're supposed to comply with.
030
Ray Myers @raymyers.bsky.social · 25/08/2026
Yep, we do open this episode with Liz’s observation that pro-AI people seem very insecure. I also say that I think AI doesn’t really mean tech anymore, more so ideology. Others may have different ideas of what promotes better dialogue. It’s a tangled world. thecodingagency.org/why-is-ai-po...
thecodingagency.org
Why Is AI Political? - The Coding Agency
Today we examine what it really means for something to be "political", and consider the situation for engineers just trying to get things done when their technology becomes a controversy. Resource...
110
Ray Myers @raymyers.bsky.social · 21/08/2026
No doubt it was mocking in context, but yeah Chat Bots are pretty intentionally made to be imaginary computer friends. Of course prompts say, “You are a helpful assistant” which isn’t quite friend, unless it’s one of the many services that do prompt them to role play as friends…
100
Ray Myers @raymyers.bsky.social · 19/08/2026
Here’s my version and if it makes me weird now that’s OK. AI is not good for software engineering, because AI is an ideology hostile to other forms of knowledge. LLMs and coding agents help SWE within reasonable expectations: joining with other tech, practices, and social process. AI breaks that.
140
Ray Myers @raymyers.bsky.social · 19/08/2026
It’s such an impossible standard to live up to, doing as well as the stuff we can’t remember doing.
100
Reposted by Ray Myers
Liz Fong-Jones (方禮真) @lizthegrey.com · 19/08/2026
this is tomorrow!
1236
Ray Myers @raymyers.bsky.social · 19/08/2026
Btw I once invited a qualified punk Kelly Betz to comment on the “Vibe Coding is the Punk Rock of Software” video by Andreessen Horowitz & Rick Rubin. Turns out it’s not. youtu.be/XP7UXspsggw?...
youtu.be
Vibe Coding Is Not Punk Rock w/ Kelly Betz
YouTube video by Craft vs Cruft
110
Ray Myers @raymyers.bsky.social · 18/08/2026
Maybe I should just morph back into my Java programmer self from 2007-2014 I could take everything else off my resume, all that “staying relevant”. Forget that I was early with DevOps and Platform and agents. There would still be plenty of work to do. I might be happier.
100
Reposted by Ray Myers
Sue Smith @suesmith.lol · 18/08/2026
No better indicator that someone knows almost nothing about cancer than the idea of "curing cancer"
121
Ray Myers @raymyers.bsky.social · 18/08/2026
The apologetics downplaying these things tend to point out its small part of the overall picture, just bad optics. Visceral associations, book burning and all that. However big picture, the AI gold rush regularly displays utter contempt for the totality of human knowledge including Comp Sci so…
110