Natalie
@natkr.com
Reposted by Natalie
If something is tedious or clunky, your reaction shouldn't be "let's have the machine write it," it should be "how can I redesign this to be less tedious or clunky."
November 28, 2025 at 11:21 AM
If something is tedious or clunky, your reaction shouldn't be "let's have the machine write it," it should be "how can I redesign this to be less tedious or clunky."
Reposted by Natalie
“Schemaless” and “self-describing” are the dynamic typing of data formats: you put more crap into the runtime representation, making it perform worse and in exchange you get worse reliability because errors are identified less reliably. Lose/lose
November 23, 2025 at 11:05 AM
“Schemaless” and “self-describing” are the dynamic typing of data formats: you put more crap into the runtime representation, making it perform worse and in exchange you get worse reliability because errors are identified less reliably. Lose/lose
PCI Express (the thing internal expansion cards use) on a wire. Goes vroom, if you have anything to plug it into.
These days it's also the same as USB4 (though over USB-C cables, like Thunderbolt 3 and up). Which matters, because USB is an open…ish standard, while TB is an Intel/Apple specialty.
These days it's also the same as USB4 (though over USB-C cables, like Thunderbolt 3 and up). Which matters, because USB is an open…ish standard, while TB is an Intel/Apple specialty.
October 8, 2025 at 11:23 PM
PCI Express (the thing internal expansion cards use) on a wire. Goes vroom, if you have anything to plug it into.
These days it's also the same as USB4 (though over USB-C cables, like Thunderbolt 3 and up). Which matters, because USB is an open…ish standard, while TB is an Intel/Apple specialty.
These days it's also the same as USB4 (though over USB-C cables, like Thunderbolt 3 and up). Which matters, because USB is an open…ish standard, while TB is an Intel/Apple specialty.
Reposted by Natalie
The idea that we have to "encourage responsible AI usage" is a core part of the inevitability narrative driving the institutional adoption that makes people think it's worth trying out.
it contributes to the very problem it supposedly tries to solve. we can't let it go unchallenged.
it contributes to the very problem it supposedly tries to solve. we can't let it go unchallenged.
September 8, 2025 at 3:48 AM
The idea that we have to "encourage responsible AI usage" is a core part of the inevitability narrative driving the institutional adoption that makes people think it's worth trying out.
it contributes to the very problem it supposedly tries to solve. we can't let it go unchallenged.
it contributes to the very problem it supposedly tries to solve. we can't let it go unchallenged.
Hah! That's a fair catch.
September 14, 2025 at 12:01 AM
Hah! That's a fair catch.
I've never had a headphone jack fail to pair.
September 13, 2025 at 11:32 PM
I've never had a headphone jack fail to pair.
Reposted by Natalie
Feeling bad for all the people who, long after all the customer-facing traces have been erased, will spend the next several years dodging the AI components lodged in their systems.
"Whatever you do, don't click the AI button," will be the "you gotta jiggle the handle" of enterprise software
"Whatever you do, don't click the AI button," will be the "you gotta jiggle the handle" of enterprise software
September 5, 2025 at 6:23 AM
Feeling bad for all the people who, long after all the customer-facing traces have been erased, will spend the next several years dodging the AI components lodged in their systems.
"Whatever you do, don't click the AI button," will be the "you gotta jiggle the handle" of enterprise software
"Whatever you do, don't click the AI button," will be the "you gotta jiggle the handle" of enterprise software
Nevermind, spoke too soon. Turns out, today's positive sides are cancelled.
![google docs trying to correct "We do a little [lite] embezzlement." to "[bit of]"](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:r6ivhyag7ece4ej5nphdvao4/bafkreiferqhc5tfsh4ct7qjdwb5gc77rxspnz2whg2hor2uhnw42qrrjw4@jpeg)
August 18, 2025 at 6:31 PM
Nevermind, spoke too soon. Turns out, today's positive sides are cancelled.
On the negative side, today I found out that google docs apparently now asks about whether to apply the template for.. email drafts?! Who is this for? Why?
August 18, 2025 at 6:29 PM
On the negative side, today I found out that google docs apparently now asks about whether to apply the template for.. email drafts?! Who is this for? Why?
I had to go check this for myself, and... On the positive side, I can't reproduce /this/ for my account. Maybe having disabled spellcheck also disabled the tone police bullshit? Or maybe it's just another case of "machine learning garbage is inherently unreliable, more news at 11".
August 18, 2025 at 6:29 PM
I had to go check this for myself, and... On the positive side, I can't reproduce /this/ for my account. Maybe having disabled spellcheck also disabled the tone police bullshit? Or maybe it's just another case of "machine learning garbage is inherently unreliable, more news at 11".
Reposted by Natalie
Why would anyone do this? It's so bad from all angles. If you don't have the energy or interest to reply to a comment, just don't. It's fine. It's literally so weird that they're pushing an elevation of the already really bad automatic responses. If I learned a youtuber used these, I'd shame them
August 5, 2025 at 12:02 PM
Why would anyone do this? It's so bad from all angles. If you don't have the energy or interest to reply to a comment, just don't. It's fine. It's literally so weird that they're pushing an elevation of the already really bad automatic responses. If I learned a youtuber used these, I'd shame them