Really excited to share my new article "Dis/similarities in the Design and Development of Legal and Algorithmic Normative Systems: the Case of Perspective API"
The article tries to contribute to the ongoing discussion on ML systems as a regulatory medium.
What I find? A short 🧵:
Last summer @GabeNicholas and me brought together +30 scholars from law and computer science to discuss how to regulate large language models. Today, we publishe our report on the roundtable!:
NEW RELEASE: @CenDemTech + @ILI_NYU release proceedings from a roundtable held last summer w/ 30 computer science & legal scholars about how to approach regulating large language models. Cc: @GabeNicholas + @paulfriedl4 https://t.co/MFFCnUrXXp
🚨 Jobs - Jobs - Jobs 🚨
Exciting news! From October on, I will lead the new #humanet3 project group. We'll work interdisciplinary (behavioral sciences, competition, and European/international law) on the ‘human-centered’ digital transition. All vacancies are full-time positions!
Our Spawning Browser Extension is now available on Chrome, allowing you to discover if the data on the web pages you browse has been included in public AI training sets.
If you discover your work is included, you can choose to opt-out!
https://t.co/ClcckouPKB
The statements from Thierry Breton of the European Commission about shutting down social media during riots are shocking. They vindicate every warning about the DSA that experts from the majority world (aka global south) have been shouting throughout this process. 1/
Appreciate this. In fact, there's something I've been calling the "algorithmic irrelevance" theory, where I suspect that most of what is problematic about online platforms (ie. addiction, misinfo, radicalization) is actually mostly due to design elements outside of the algorithm.
There's straight time (think classical music). There's swing time (think Jazz). Then there's 'Dilla time'.
Great piece in @LRB about how J Dilla used new technology to recreate rhytem in a way which reflects the fragmented life of today
https://t.co/nbeW1tiqk4
Bit embarrassing for a "privacy scholar", but until today I didn't really know what differential privacy was. One youtube search later, I have a good rough understanding and (more importantly) also seem to have found a language extremely helpful for many data protection issues ..
The amount of daily upkeep required to prevent a situation where you have a huge pile of clothes that you move between your chair and bed is insane. It would ultimately amount to a part-time job
It's comforting to think "if you're not paying for the product, you're the product," AKA, "If you can afford to pay for a product, you won't be the product." But it's bullshit. You're not the product because you don't pay - you're the product because you can't stop them. 1/
@willmacaskill This speaks to another huge blindspot in the EA strategy: it doesn't understand that if you want to change the future you have to change the present. Many distrust EA, because this is so obvious, that it seems implausible that anyone with good intentions wouldn't understand it.
@willmacaskill How will those ills *not* continue and manifest in "AGI" if you don't tackle them *now*? Longtermist AI thinking seems to be nothing but hyping up the AI industry and giving it credit which it does not deserve.
With Twitter's change in ownership last week, I'm probably in the clear to talk about the most unethical thing I was asked to build while working at Twitter.
🧵