What's the point of humans writing anymore?
*#1: Stop saying/hoping that all AI writing is just slop. Much is excellent; few people write as well.
* #2: AI writing will help make sense of much lousy human blather.
*#3: But something strange happens when I discover that a text was AI-generated: suddenly, I don't care what it says. I call it "the off-switch" of AI writing. But why is this?
*Why, if #1 + #2, is there an "off-switch"? Could it hint at something that could save human writing? I think so.
*The AI economist @alexolegimas predicts that human writing could actually become *more* valuable in the future
*But what's key here is that we can actually tell what's what
*Literary prizes and publishers and dating apps and anyplace that should care about human authorship--they need to get up to date on AI-writing detection apps, and stop incentivizing deception, which is happening way more than most think (h/t @nabeelqu)
*You may not realize it but AI detection apps have gotten excellent, as @max_spero_ explained to me
*Check out my "5 Rules of AI Writing" essay, in which I discuss all of the above:
https://t.co/OZn3jG8XWd
[Many thanks to those whose private comments or public thinking bettered my human writing, including @alexolegimas@sebkrier@bratton@alisonmsnyder@arthur_goemans_@merrierm@econormist@wself@jasminewsun@j_amesmarriott@DKThomp & @davidoks]
I have two obsessions: AI and football (aka soccer). So the @FIFAWorldCup is awkward: how to claim I'm thinking about technology while actually trying to watch every match live?
Thankfully, @daisychristo has the answer, explaining what today's Ugly Game (that is, football under the sway of VAR) says about how societies should apply tech solutions and when to beware.
To somehow mention Maradona, Lee Sedol & deep-learning in a single article is the perfect hat-trick: ⚽️⚽️⚽️...
https://t.co/Ixb5V9CHev
We have made some updates to the Google DeepMind AGI safety course:
- New talk on Security & Control by @MaryPhuong10
- Updated talk on Interpretability by @NeelNanda5
- Updated talk on Robust training & monitoring by @seb_far
Check these out here: https://t.co/9Mts3DPT3Y
@salonium Latest wishlist for improving the quality of scientific writing (in all forms), and peer review, in an AI-enabled world?
Inspired by your previous excellent article on peer review for Works in Progress.
https://t.co/Q7GW61Jx1T
An interview with Vint Cerf, one of the creators of the internet, on AI and agents. Lots of interesting takes, e.g. he argues that while Internet protocols don't need to change fundamentally, we will still need standardization for AI agent interactions. https://t.co/dCa0ynQiFz
🐙 How will institutions evolve as AI agents enter the workforce? I examine how tacit knowledge, workplace dynamics, and organizational structures might transform. From legacy systems to multi-agent startups, from human quirks to principal-agent efficiency. Link below!
AI for science could be more impactful than chatbots. It is already helping win Nobel prizes and accelerating drug development and materials discovery.
Today we published an essay about why it matters, how it’s happening and its implications. Let’s look at it like economists.
Great event launching @TheEIU benchmarking index “Out of the Shadows: Shining light on the response to child sexual abuse and Exploitation" in the EU; joining forces w/ keyplayers on #ChildRights. A multi-stakeholder approach is needed to combat child sexual abuse & exploitation.