Log inSign up
Timothy B. Lee
108.3K posts
Timothy B. Lee profile banner
@binarybits

Timothy B. Lee

@binarybits
Reporting on AI and the future of the economy. CS masters degree from Princeton. Newsletter: understandingai.org Podcast: aisummer.org
Washington, DC
understandingai.org
Joined March 2008
1,602
Following
52.4K
Followers
RepliesRepliesRepostsRepostsMediaMedia

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.
  • @binarybits
    Timothy B. Lee
    @binarybits
    9h
    I think this is a reasonable analogy for the risk from rogue AIs. Fukushima was a huge international story and triggered changes to the ways people operate and regulate power plants. But it was 7 to 10 orders of magnitude less serious than human extinction.
    @jon_stokes
    Jon Stokes
    @jon_stokes
    9h
    Replying to @jon_stokes
    My threat model, which I've talked about publicly since at least 2023, is on the order of a Fukushima or (more likely) Thalomide, the latter being a slower-rolling catastrophe that we eventually noticed & root-caused.
    3
  • @binarybits
    Timothy B. Lee
    @binarybits
    23h
    This is a great post. My view has long been that (as Achiam puts it) "pure containment or alignment is a kind of wishful thinking." Accepting that rogue autonomous AIs are inevitable isn't defeatism, it's a precondition to responding effectively.
    @jachiam0
    Joshua Achiam
    @jachiam0
    Sep 1
    There is a fact about the future that I feel many people are not facing for reasons that are largely psychological: there are going to be rogue AIs that exist in the world, that will replicate in the wild, and that will attempt to acquire resources for themselves. There will be
    3
  • @binarybits
    Timothy B. Lee
    @binarybits
    Sep 5
    It's interesting how our timeline is diverging from the AI 2027 scenario: capabilities are progressing faster, and the frontier labs have handled things worse than in AI 2027. This screenshot was supposed to happen in January 2027.
    Image
    14
  • @binarybits
    Timothy B. Lee
    @binarybits
    Sep 4
    About to start!
    @labenz
    Nathan Labenz
    @labenz
    Sep 4
    We are live. AI:AM, Friday morning. Up first: Timothy B. Lee of Understanding AI (@binarybits), who grants the exponential and still doesn’t buy superintelligence. At 10:15 Pacific, surgeon-turned-founder Dr. Jean Nehme of morph on soft robotic cells. x.com/i/broadcasts/1…
  • @binarybits
    Timothy B. Lee
    @binarybits
    Sep 4
    I will be talking to Nathan live this morning at 9:30am Pacific/12:30pm Eastern.
    @labenz
    Nathan Labenz
    @labenz
    Sep 4
    Join us on Friday as we talk to @binarybits — Timothy Lee, Founder at Understanding AI — and Dr. Jean Nehme, Founder & CEO at morph — on @AI_in_the_AM. x.com/i/broadcasts/1…
Image
REPLAY
@labenz
Nathan Labenz
@labenz
MongoDB's Agent Memory: Can Models Learn To Forget?
Image
REPLAY
@labenz
Nathan Labenz
@labenz
MongoDB's Agent Memory: Can Models Learn To Forget?
Advertisement
Advertisement