"Chez lui, quelle que soit la tristesse du poète, elle est consolée, surpassée — c’est-à-dire malheureusement vite détruite — par l’allégresse du fabricateur." (Proust, La Prisonnière)
@yudapearl We should also inform people how LLMs distort history related to Jews. LLMs are based off of statistical frequency in training + input. When the input is biased, the output is biased. But the output is treated as gospel by most users. This snowballs over time. Example:
A page called The History Of Music Channel have completely plagiarised one of my Dad's well known cartoons and are claiming it as their work.
What's possibly worse is I have a feeling they used Al to rip him off!
I saw dozens of Palantir applications researching my long read (link below), and looked at equivalent systems built using British technology. Like Manchester's.
The question isn't whether federating and analysing data improves outcomes, we know that it does. It's believing that only once company can do it, and you have to pay them lots of money and given them a fat exclusive contract too.
Key quote:
"It's Databricks + Apache Spark + lobbying"
We don't seem to learn anything from decades of procurement failure.
https://t.co/X7l5kxSYwx
An interesting insight into the viability of Burnham's "pan-Northern" strategy is that the residents of these cities have very little love for one another
The social catchments of Liverpool and Manchester look as if they consciously exclude one another
@bettedavisthizz@typofoto@trevorbaxendale Desperately Seeking Susan is a great film and it her best role. I watched it again recently and thought Madonna was amazingly bad at even acting herself.
Jonathan Rose is v interesting on this in The Intellectual Life of the British Working Classes. Thomas Carlyle "had a huge following among autodidacts" and consistently crops up as a favourite author of early Labour MPs!
This has been bleeding obvious to everyone for 18 months, since DeepSeek’s “Sputnik” moment - maybe marketing professor Galloway was distracted doing his NPR-friendly Andrew Tate for libs grift.
DeepSeek showed just how badly the Americans have wasted their capital on poor engineering and hopium. Overnight, LLMs had become a commodity, and being open source, these models were going to be everywhere. The AI bubble was like giving Winsock a trillion dollars in 1994. There’s no value in something that’s a commodity. You’re not going to get your money back.
And China is not “dumping”, it’s just how technology which has no defensible moat defuses out. And how open source software is supposed to work. They’re playing the Cold War card out of sheer desperation, while trying to make open source illegal.
The USA has been crashed by the dumbest people in history - its VCs.