Matt
@flyinghyrax.bsky.social
$work uses Google Workspace and I started a shared collection of screenshots for “you should use gemini please we’re begging you” prompt UIs. Easily 20+ unique examples across products in the last year. I laugh so I don’t cry.
January 7, 2026 at 11:14 PM
$work uses Google Workspace and I started a shared collection of screenshots for “you should use gemini please we’re begging you” prompt UIs. Easily 20+ unique examples across products in the last year. I laugh so I don’t cry.
Good for them! And all it cost the rest of us was irreversibly poisoning the entire information ecosystem.
November 30, 2025 at 7:06 PM
Good for them! And all it cost the rest of us was irreversibly poisoning the entire information ecosystem.
Thread on the theme: bsky.app/profile/tres...
This is one of the absolute toughest things about predatory inclusion — the predation is often fun or a little useful. Criticism makes a reasonable person sound like an asshole. LowerEd weathered me for this kind of social death.
So I’ll say it: your enjoyment does not matter & should not matter.
So I’ll say it: your enjoyment does not matter & should not matter.
I don't fundamentally disagree with a lot of this, especially the environmental cost/ benefit. I'll just offer that as a maker,tinkerer,experimenter, having access to a LLM has massively decreased the time required to design build and debug relatively sophisticated hw/sw prototypes.
November 27, 2025 at 2:04 PM
Thread on the theme: bsky.app/profile/tres...
Apart from everything by else, the circular citations are so obvious in hindsight. I introduced people to his newsletter and now feel very silly.
Thank you for being a better voice; it must get frustrating to have to repeat the same things so often.
Thank you for being a better voice; it must get frustrating to have to repeat the same things so often.
October 27, 2025 at 11:25 PM
Apart from everything by else, the circular citations are so obvious in hindsight. I introduced people to his newsletter and now feel very silly.
Thank you for being a better voice; it must get frustrating to have to repeat the same things so often.
Thank you for being a better voice; it must get frustrating to have to repeat the same things so often.
The “One Book“ and “One Pundit“ theories continue to be validated! 🙂
October 7, 2025 at 12:13 PM
The “One Book“ and “One Pundit“ theories continue to be validated! 🙂
It really is insane. I charged ours day before yesterday for what I’m pretty sure is the second time the entire time we’ve had it. Could be off by one but still.
October 3, 2025 at 5:28 PM
It really is insane. I charged ours day before yesterday for what I’m pretty sure is the second time the entire time we’ve had it. Could be off by one but still.
It is interesting how differently we think - to me, what you’ve quoted from Bjarnson is self-evident! And while I appreciate the “Worse Is Better“ essay (ty for reminding me of it!) I’m not sure I take the same point from it. Ty for the food for thought. Ultimately we may just have different values.
September 28, 2025 at 10:54 PM
It is interesting how differently we think - to me, what you’ve quoted from Bjarnson is self-evident! And while I appreciate the “Worse Is Better“ essay (ty for reminding me of it!) I’m not sure I take the same point from it. Ty for the food for thought. Ultimately we may just have different values.
Personally, I think this all sounds very exciting for whoever is on call when it breaks in production.
September 28, 2025 at 10:41 PM
Personally, I think this all sounds very exciting for whoever is on call when it breaks in production.
Reasonable enough, thanks for elaborating. There’s never been a consensus on what makes software teams productive, and LLMs have only made it less clear. I’m glad your personal connections have given you insight, I just don’t have that clarity myself. I suppose we’ll all see what is built!
September 28, 2025 at 10:39 PM
Reasonable enough, thanks for elaborating. There’s never been a consensus on what makes software teams productive, and LLMs have only made it less clear. I’m glad your personal connections have given you insight, I just don’t have that clarity myself. I suppose we’ll all see what is built!
Anecdotally from a colleague, apparently the hope is that the code doesn’t have to be maintainable, since when it breaks you’ll use the Text Emitter to debug it too. And when requirements change you just regenerate the entire thing from scratch instead of trying to modify anything. 🙃
September 28, 2025 at 10:24 PM
Anecdotally from a colleague, apparently the hope is that the code doesn’t have to be maintainable, since when it breaks you’ll use the Text Emitter to debug it too. And when requirements change you just regenerate the entire thing from scratch instead of trying to modify anything. 🙃
As far as what you’ve personally observed, fair is fair, and if you feel like it helps you then fine. But regarding the post you linked: www.baldurbjarnason.com/2025/trustin... from @baldurbjarnason.com is one of several thorough responses.
Trusting your own judgement on ‘AI’ is a huge risk
Web dev at the end of the world, from Hveragerði, Iceland
www.baldurbjarnason.com
September 28, 2025 at 10:09 PM
As far as what you’ve personally observed, fair is fair, and if you feel like it helps you then fine. But regarding the post you linked: www.baldurbjarnason.com/2025/trustin... from @baldurbjarnason.com is one of several thorough responses.