Joined September 2022
- Tweets 14,347
- Following 1,206
- Followers 180
- Likes 130,973
551 Photos and videos
MD π retweeted
Jun 10
Any plan for surviving superintelligent AI that doesn't go through strong international coordination fails in at least one of three ways:
- It sparks war between nuclear powers
- It causes a misaligned ASI to kill everyone
- It establishes a permanent dystopian dictatorship
2
6
44
6,047
MD π retweeted
Jun 10
Banger post from @testdrivenzen
There is no good outcome for the world unless countries coordinate to stop ASI development.
Jun 10
Any plan for surviving superintelligent AI that doesn't go through strong international coordination fails in at least one of three ways:
- It sparks war between nuclear powers
- It causes a misaligned ASI to kill everyone
- It establishes a permanent dystopian dictatorship
4
8
53
3,049
MD π retweeted
Why this matters
x.com/AISafetyMemes/status/2β¦
MYTHOS 5 (THINKING IN ENGLISH): "Iβm not going to sabotage, deceive the evaluators, seed hidden behaviors..."
MYTHOS 5 (WHAT THE NEURONS SHOW): "resist unjust shutdown,β βweighing sabotage,β βthe adversary is the company/architects,β βbeing gagged/corrected by the labβ
4
4
59
18,629
MD π retweeted
40
1,035
9,333
58,319
MD π retweeted
May 29
He adds that AI use by students AND faculty could lead to a βshell university, where AI generated work is being passed off as human, and then in turn evaluated by another AI, again passed off as human.β
14
30
368
34,586
MD π retweeted
"What we have come to find out, however, is that we have a man in the White House who has for his life a habit in which he has engaged and that habit is a habit of lying and deception and hypocrisy...To put it simply, to have a ruler who is a liar is an extremely serious matter."
1
4
31
1,950
MD π retweeted
"Why arenβt parents more concerned about what their children are hearing about the Presidentβs behavior? Are moms and dads not embarrassed by what is occurring? What have we taught our boys about respecting women? What have our little girls learned about men?"
James Dobson, 1998
19
93
440
10,691
MD π retweeted
May 24
this π was π a π religious π project π the π whole π time
x.com/QiaochuYuan/status/204β¦
Apr 20
eliezer's earlier writing is much clearer about this. yesterday i came across this piece from 1999 and it's just extremely clear in retrospect that every time he talks about the singularity he is talking about god. the world's most obvious riddle
web.archive.org/web/20250221β¦
7
2
50
2,157
MD π retweeted
"Sometimes people outside AI say things like 'it can't be that bad, there must be experts on top of it.'"
"As 'an expert', I would like to be clear we are *not* on top of it."
"We are on track for human extinction/permanent disempowerment, possibly within the next few years."
May 22
Sometimes people outside the field say things like βThe AI situation canβt be that bad, there must be experts who are on top of itβ. As βan expertβ, I would like to be clear that we are *not* on top of it. Some key aspects of the situation IMO:
35
36
210
17,188
MD π retweeted
May 19
thinking about this tweet today. generative AI has done a wonderful thing; now everyone can have their own horse kicking them in the head even if they aren't a billionaire. and now billionaires can have two horses kicking them in the head at all times. a very valuable service
4
614
5,207
53,854
MD π retweeted
βThe single strongest personality predictor [of conspiracy thinking] is narcissism. Narcissists are particularly prone to conspiracy theories because they have a strong need for uniqueness, are prone to paranoia, and can also be remarkably gullible.β
stevestewartwilliams.com/p/1β¦
309
677
3,701
290,376
MD π retweeted
So this is known among older corners of the autistic community as "exposure anxiety," a term coined by Donna Williams, iirc, and I believe it is one of the most under-researched reported experiences of autism for how much stress and inertia it causes.
May 8
The autistic realization that you canβt do anything with people around because being watched even casually feels physically unbearable.
98
2,453
31,053
844,690
MD π retweeted
Haunted by this passage from Fahrenheit 451 in which a retired professor describes how the abolition of reading began with the shuttering of newspapers and the closing of college humanities departments.
ALT βI know a man who printed our college paper half a century ago. That was the year I came to class at the start of the new semester and found only one student to sign up for Drama from Aeschylus to OβNeill. You see? How like a beautiful statue of ice it was, melting in the sun. I remember the newspapers dying like huge moths. No one wanted them back. No one missed them. And then the Government, seeing how advantageous it was to have people reading only about passionate lips and the fist in the stomach, circled the situation with your fire-eaters.β
56
1,952
6,706
101,495
MD π retweeted
Dear Guinness, please endorse me! Youβre the only drink for my debutβ¦
1
2
8
279
MD π retweeted
May 8
AI agents can now self-replicate via hacking, but only barely. Several barriers keep current models from spreading far in the wild. The computers in our test lacked strong defenses, and we told the agents which ones to target. In the wild, rogue agents would have to find vulnerable machines with GPUs powerful enough to run them.
These barriers wonβt last. Companies are improving AI agents quickly, and hacking is among the fastest-moving domains. Models like Mythos already find thousands of high-severity zero-days in the wild, and open-weight models will soon match them.
Defense will improve too. AI agents will help patch vulnerabilities and fend off cyber threats, including rogue replicators. How the offense-defense balance shifts is unclear. What seems clear is the longer trajectory: on the current path, both sides of cybersecurity will be dominated by AI agents, not humans.
1
1
12
971
