Pascal Thibeault (@pascalthibeault) 's Twitter Profile
Pascal Thibeault

@pascalthibeault

Liminal Wanderer tinkering with LLMs. Prev. Law @Medtronic & @thefertpartners | Writing: beyondthealgorithms.substack.com

ID: 142828427

calendar_today11-05-2010 21:47:15

3,3K Tweet

975 Followers

5,5K Following

Laz (@lazpieper) 's Twitter Profile Photo

Sitting in on Storm’s trial, I can’t help but feel violated in getting an up close experience of how the government can cherry pick information that’s been collected on you (e.g., transactions, messages, IP address activity) and paint any picture of you that they want. Just

Dan Primack (@danprimack) 's Twitter Profile Photo

Asked Figma CFO earlier today about how he views a stock pop. His answer: Most of the shares were secondary, with the primary to cover RSU tax implications. So even big pop doesn't mean co left much $$ on the table. Both true and also a lie he tells himself.

Pascal Thibeault (@pascalthibeault) 's Twitter Profile Photo

Call me crazy, but I don't think prospective employers should require you to spend $$ for a software subscription just to interviewing with them.

Pascal Thibeault (@pascalthibeault) 's Twitter Profile Photo

Textbook example of the Law of Unintended Consequences where Intended use β‰  Actual use + Human error β†’ Unintended risk

πš‘πšŽπš—πš” πšŸπšŠπš— 𝚎𝚜𝚜 (@henkvaness) 's Twitter Profile Photo

ChatGPT quietly scrubbed today nearly 50,000 shared conversations from Google's index after our investigation. They thought they'd solved the problem. They were wrong. (1/5)

ChatGPT quietly scrubbed today nearly 50,000 shared conversations from Google's index after our investigation. They thought they'd solved the problem. They were wrong. (1/5)
Vinod Khosla (@vkhosla) 's Twitter Profile Photo

And Lina Khan is best equipped to decide this better than founders, employees and investors? Colossal stupidity to not understand retrospective predictability!

Derek Thompson (@dkthomp) 's Twitter Profile Photo

New paper: Since the rise of large language models, there's been a huge shift in academic writing. In 2024, the word "delves" appeared 2,700% more than its historical average, by one account. The analysis suggests that 13.5% of 2024 abstracts were processed with LLMs.

New paper: Since the rise of large language models, there's been a huge shift in academic writing.

In 2024, the word "delves"  appeared 2,700% more than its historical average, by one account.

The analysis suggests that 13.5% of 2024 abstracts were processed with LLMs.
Emmett Shear (@eshear) 's Twitter Profile Photo

The idea that being evil is a personality trait is (a) hilarious (b) terribly mistaken. Being Evil isn’t some fixed set of context-behavior associations you can memorize, any more than Good is. The personality they named Evil is actually Cartoon Villain.

Simon Willison (@simonw) 's Twitter Profile Photo

I wrote about what went wrong with ChatGPT's sharing dialog, and why I think it's reasonable for people to be confused by what looks at first glance like a very clear checkbox description simonwillison.net/2025/Aug/3/pri…