• 3 Posts
  • 8 Comments
Joined 1Y ago
cake
Cake day: Jul 08, 2023

help-circle
rss
Letters from the lawyers: Musk threatens CCDH with brazen attempt to silence honest criticism. — Center for Countering Digital Hate | CCDH
"Last week we got a letter from Elon Musk’s X. Corp threatening CCDH with legal action over our work, exposing the proliferation of hate and lies on Twitter since he became the owner. Elon Musk’s actions represent a brazen attempt to silence honest criticism and independent research in the desperate hope that he can stem the tide of negative stories and rebuild his relationship with advertisers." [With apologies to anyone who dislikes endless Musk/Huffman spam in this community. I put it here because misusing the law to silence independent tech researchers this has wider implications.]
fedilink


Tesla’s Dieselgate
Tesla is a giant shell-game masquerading as a car company. The important thing about Tesla isn't its cars, it's Tesla's business arrangement, the Tesla-Financial Complex:
fedilink

It’s OK. Ordinary people will have no trouble at all making sure they use a different vehicle every time they drive their kid to college or collect an elderly relative for the holidays. This will only inconvenience serious criminals.


You’re agreeing with me but using more words.

I’m more annoyed than upset. This technology is eating resources which are badly needed elsewhere and all we get in return is absolute junk which will infest the literature for decades to come.


In context. And that is exactly how they work. It’s just a statistical prediction model with billions of parameters.


That’s not true! There’s heaps of early-GPT articles pointing out how much bullshit it regurgitates (eg Why does ChatGPT constantly lie?). And no evidence at all that the breathless fanboys have even stopped to check.


It will almost always be detectable if you just read what is written. Especially for academic work. It doesn’t know what a citation is, only what one looks like and where they appear. It can’t summarise a paper accurately. It’s easy to force laughably bad output by just asking the right sort of question.

The simplest approach for setting homework is to give them the LLM output and get them to check it for errors and omissions. LLMs can’t critique their own work and students probably learn more from chasing down errors than filling a blank sheet of paper for the sake of it.


They’re circular. If the text is too predictable it was written by an LLM* but LLMs are designed to regurgitate the next word most commonly used by humans in any given context.

*AI is a complete misnomer for the hi-tech magic 8ball



If you were meeting up somewhere you’d arrange to have someone who was at home (and thus by a phone) to orchestrate any last minute changes of plan or notifications of late arrivals (via payphones, which were a thing, once).

You’d go into town regularly to pick up the new bus timetable.

You’d have a huge pile of maps in the back of the car, or one very big map book, often both. If you drove somewhere once, you’d remember the route the next time.

There was a set of encyclopedias at home to look up facts.

And a calendar on the wall. (That’s probably still a thing?)

There were a lot more newspapers and magazines around.

Everyone had a little notebook with all their important phone numbers in it. Filofax was revolutionary.

And we still remember the most important phone numbers from that little notebook because we had to dial them so very often.

We played eye spy a lot.