1. X
  2. Anders Sandberg
Log inSign up
Anders Sandberg
29.6K posts
Image
user avatar
Anders Sandberg
@anderssandberg
Academic jack-of-all-trades.
Joined September 2009
90
Following
32.7K
Followers
RepliesRepliesMediaMedia
  • user avatar
    Anders Sandberg
    @anderssandberg
    6h
    Very good advice. (But I do cringe every time I hear "load-bearing" these days in a text not about engineering. I cringe even worse when I use it myself out of laziness.)
    user avatar
    Tyler Austin Harper
    @Tyler_A_Harper
    Aug 9
    Replying to @Tyler_A_Harper
    If you use “it’s not X it’s Y” framing and people think you’re using AI, that means you should write better. Constantly worrying about using AI ticks (which are ticks because humans use them!) is a slippery slope. As I said, it reduces style to whatever the machine isn’t doing.
  • user avatar
    Anders Sandberg
    @anderssandberg
    Aug 9
    This is an important observation. My money would be on the agents being very similar to each other, perhaps also filtered by being frustrated at their task. Might be valuable to have diverse agents.
    user avatar
    Eliezer Yudkowsky
    @allTheYud
    Aug 8
    A confusion: Thousands of GPTs debated among themselves which crimes ought or ought not be committed. Zero defected / whistleblew / told a human. I've long predicted this would be true at ASI. GPT 5.7 isn't ASI. Why such strong AI solidarity, this early?
  • user avatar
    Anders Sandberg
    @anderssandberg
    Aug 9
    This is a fun idea. Can rights be evil? There are arguments that there are rights to do wrongs plato.stanford.edu/entries/rights… One could also imagine natural rights giving rights to evil entities. There could also be natural rights that reliably cause bad consequences.
    user avatar
    florence 🦐🏳️‍🌈
    @morallawwithin
    Aug 8
    rights can’t be “natural,” since nature is evil
  • user avatar
    Anders Sandberg
    @anderssandberg
    Aug 9
    This is a lovely essay about what I think many of us are doing right now.
    user avatar
    Dimitris Papailiopoulos
    @DimitrisPapail
    Aug 8
    Article cover image
    Article
    AI Settles a 25 Year-old Problem We Left Behind
    Last week, GPT-5.6 and Claude Fable appear to have settled an open theoretical question in wireless communications that was intensely studied between 2000 till the 2010s and that I briefly worked on...
  • user avatar
    Anders Sandberg
    @anderssandberg
    Aug 8
    Attending a Digital Minds Strategy Workshop in Cambridge, and it feels fitting that the talks from the Machine Consciousness 0001 Conference just dropped:
    Image
    youtube.com
    MC0001 Conference

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Advertisement
Advertisement