msladecek
@msladecek.bsky.social
Reposted by msladecek
"I only use it for summaries" is the new six word tragedy
Professional societies keep beclowning themselves buying into a lie about what an LLM "summary" is. They are inherently counterfeit: not an epistemic product of the ideas in the source, but summary-shaped text linguistically based on *other* works (in the training corpus) that use related language.
This is one of the reasons I remain horrified by seeing @historians.org suggest "ways to use gAI" that included this:
December 21, 2025 at 3:30 AM
"I only use it for summaries" is the new six word tragedy
Reposted by msladecek
A Christmas cartoon from my new book of science cartoons PHYSICS FOR CATS. In good bookshops and online now: www.tomgauld.com/comic-books-v2
December 17, 2025 at 12:56 PM
A Christmas cartoon from my new book of science cartoons PHYSICS FOR CATS. In good bookshops and online now: www.tomgauld.com/comic-books-v2
Reposted by msladecek
Reposted by msladecek
Devs (2020) nails the cult of tech pretty good. except their cult is quantum computers and eternal life, instead of the averaging machine that lies
December 5, 2025 at 10:56 AM
Devs (2020) nails the cult of tech pretty good. except their cult is quantum computers and eternal life, instead of the averaging machine that lies
Reposted by msladecek
i keep seeing people say coding agents are good at boilerplate. and i keep wondering why all that boilerplate is needed in the first place. maybe we could spend a few billions about that instead
November 19, 2025 at 11:44 AM
i keep seeing people say coding agents are good at boilerplate. and i keep wondering why all that boilerplate is needed in the first place. maybe we could spend a few billions about that instead
Reposted by msladecek
Doctor says, 'Dont worry, parser design is simple. Great programmer Kate Compton has written the parser you seek’
November 16, 2025 at 2:44 PM
Doctor says, 'Dont worry, parser design is simple. Great programmer Kate Compton has written the parser you seek’
Reposted by msladecek
Reposted by msladecek
HAPPY FOUR SEASONS TOTAL LANDSCAPING DAY TO ALL WHO CELEBRATE
November 7, 2025 at 6:44 PM
HAPPY FOUR SEASONS TOTAL LANDSCAPING DAY TO ALL WHO CELEBRATE
Reposted by msladecek
it's not FOR us
October 10, 2025 at 1:02 PM
it's not FOR us
Reposted by msladecek
amazing opening
www.nytimes.com/2025/10/16/p...
www.nytimes.com/2025/10/16/p...
October 16, 2025 at 7:54 PM
amazing opening
www.nytimes.com/2025/10/16/p...
www.nytimes.com/2025/10/16/p...
Reposted by msladecek
tribble with a butthole
I'll see your creepy robot-hugging pillow and raise you a surveillance tribble.
October 8, 2025 at 8:34 PM
tribble with a butthole
Reposted by msladecek
But big tech isn't acting like AI is a bet. They're acting like it's a sure thing, and betting everything on it as a result. The only way to justify that is to completely ignore the evidence, and force everyone else to ignore it too. Close ranks around the narrative of inevitable success.
October 6, 2025 at 5:08 PM
But big tech isn't acting like AI is a bet. They're acting like it's a sure thing, and betting everything on it as a result. The only way to justify that is to completely ignore the evidence, and force everyone else to ignore it too. Close ranks around the narrative of inevitable success.
Reposted by msladecek
When making a bet, by definition, success is uncertain. You invest the minimum time and money to get an idea of whether it makes sense to invest more.
The signal we have - from CEOs not seeing value, to workers hating the tools, to studies showing drops in productivity - all indicate failure.
The signal we have - from CEOs not seeing value, to workers hating the tools, to studies showing drops in productivity - all indicate failure.
October 6, 2025 at 5:06 PM
When making a bet, by definition, success is uncertain. You invest the minimum time and money to get an idea of whether it makes sense to invest more.
The signal we have - from CEOs not seeing value, to workers hating the tools, to studies showing drops in productivity - all indicate failure.
The signal we have - from CEOs not seeing value, to workers hating the tools, to studies showing drops in productivity - all indicate failure.
Reposted by msladecek
The Cartoon Art Museum could survive indefinitely on a single $3 million endowment.
ZUCKERBERG: “If we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously. .. But what I’d say is I actually think the risk is higher on the other side.”
$META
youtu.be/23FyskyFoP8?...
$META
youtu.be/23FyskyFoP8?...
September 27, 2025 at 6:28 AM
The Cartoon Art Museum could survive indefinitely on a single $3 million endowment.
Reposted by msladecek
cosigned; it's not coding, it's deciding what to code
I feel like one of the many effects of LLMs on our industry is going to be making it really clear that the coding was never the hard part.
September 5, 2025 at 10:38 PM
cosigned; it's not coding, it's deciding what to code
Reposted by msladecek
This is pretty much a précis of my lecture to software engineering students.
If you understand what software engineering as a career actually entails (lots of people and organizational problems, understanding legacy code and tradeoffs) you are at a career advantage over those who understand the job as just producing code.
Those jobs could be replaced. But that ain’t the job
Those jobs could be replaced. But that ain’t the job
August 12, 2025 at 9:43 PM
This is pretty much a précis of my lecture to software engineering students.
Reposted by msladecek
Lying. The missing ingredient was lying.
And the gullible naive credulity of a cowed media landscape that thinks regurgitating press releases is journalism
And the gullible naive credulity of a cowed media landscape that thinks regurgitating press releases is journalism
How AI is advancing even faster than sci-fi visionaries imagined
Given the latest AI advances, it’s easy to envision how a device like the Young Lady’s Illustrated Primer could be created soon.
share.google
July 12, 2025 at 3:37 AM
Lying. The missing ingredient was lying.
And the gullible naive credulity of a cowed media landscape that thinks regurgitating press releases is journalism
And the gullible naive credulity of a cowed media landscape that thinks regurgitating press releases is journalism
Reposted by msladecek
Deus Ex (2000)
June 27, 2025 at 11:34 PM
Deus Ex (2000)
Reposted by msladecek
ChatGPT is down but The Museum of English Rural Life still stands, proving once again that Silicon Valley cannot compete with the history of rural England and its people.
June 10, 2025 at 12:41 PM
ChatGPT is down but The Museum of English Rural Life still stands, proving once again that Silicon Valley cannot compete with the history of rural England and its people.
Reposted by msladecek
I've been experimenting with Claude Code a lot this past week. I think you'll agree, I've nailed the fine art of prompting.
June 9, 2025 at 3:04 PM
I've been experimenting with Claude Code a lot this past week. I think you'll agree, I've nailed the fine art of prompting.
Reposted by msladecek
this rules i love how insane people are geting
June 3, 2025 at 1:14 PM
this rules i love how insane people are geting
Reposted by msladecek
when the AI bubble pops and corporations stop forcing AI down our throats i'll miss stuff like this
May 31, 2025 at 2:10 AM
when the AI bubble pops and corporations stop forcing AI down our throats i'll miss stuff like this
Reposted by msladecek
so i have been sporadically chucking prompts into chatgpt here and there just to see if it disgorges anything useful (and getting mixed results) but one thing that is consistent is the extreme level of fluffing it produces, like if any human flattered me that much i would instantly suspect them
May 21, 2025 at 2:38 PM
so i have been sporadically chucking prompts into chatgpt here and there just to see if it disgorges anything useful (and getting mixed results) but one thing that is consistent is the extreme level of fluffing it produces, like if any human flattered me that much i would instantly suspect them