

You assume that the VCs will keep you alive a little longer as a honeypot to gather more plaintext blackmail material and private information and keep preparing to sell out to Facebook


You assume that the VCs will keep you alive a little longer as a honeypot to gather more plaintext blackmail material and private information and keep preparing to sell out to Facebook


The presence of billions of dollars under anyoneās control destroys anyoneās brains as their lives have all negative feedback removed, their surroundings become a reflection of their own preferences alone rather than bearing any information about the world, and their social connections become transactional and sycophantic.


Self-replicating prompt-injection text is a much more interesting and much more immanent threat to cause AI systems to misbehave in a propagating way


deleted by creator


Wonder who gets all the plaintext logs from all their victims


They looking to get off the hook before all the class action lawsuits pile up?


Costly signaling by which one credibly gives something up to follow the precepts of the cult and demonstrate that one believes in the face of personal disadvantage is one of the strongest things that predicts cohesion of a religious movement.


One of the strongest predictors of the cohesion of a religious movement is costly signaling by which people give things up in a highly visible, credible way for the cause, or in a way that only makes sense if they believe in the cause. Deconversion is highly correlated with not witnessing these credible displays.


They also were a mouthpiece for self driving car hype for a while.


I remember the day my very last bit of grudging engineering respect counterweighting my disdain for the man left my body. October 2017, the day he claimed on a big presentation that Starship would bankrupt airliners by flying 1000 people per ship between world cities several times a day.
Since, it is incredibly interesting how any time he is given free reign things go off the rails. The Falcon rocket program and the Starship program are so diametrically opposed in their design philosophies and ways they are run it is amazing.


The things I printed at print shops while infiltrating a HPMOR reading group under cover in 2014ā¦


The China-Whisperers are trying to figure out how to get Rationalist Ideals into Chinese AI researchers
https://www.lesswrong.com/posts/qmxkHm2dTLKG6GZ6i/the-anatomy-of-a-chinese-ai-researcher
https://www.lesswrong.com/posts/b5cSYh4emQb2qrGmK/on-social-reality-in-china


More to the point, plague is treable by a dozen antibiotics
and even if it werenāt it is well out of the Twilight Zone that COVID was in where large chunks of the population were at low enough risk from it even without immune memory that they canāt be arsed to do anything about it, instead itās in the Smallpox Leagues where you bring the hammer down wherever you see it.


Presented without comment: Apparently Yud is now making AI-generated Isekai games with catgirls
https://johnqpulp.substack.com/p/reviewing-ai-media-everything-that


They get some replies
⦠⦠Iām writing this mostly because Iām finding it horrifying that it appears the LessWrong consensus reality is treating ānonconsentā stuff as a cornerstone of how it views women. This somehow seems to increasingly to be the case, seems wrong to me, and contributes to the erasure of some things I consider important.


Does anyone find it at all mechanistically surprising that one of the many latent dimensions of the set of all text humans produce is āsubject is in painā or that pushing activations in that direction will alter the output of a text prediction system?
Reminds me of the way that when you relax the finetuning on LLMs they start ābelievingā in UFOs and supernatural stuff. And how people take that to somehow mean something. Of course that is a large component in the space of stuff humans write about that you have to suppress if you want a professionally useful text service!
The research is cool. I love all the work that has been done decomposing representations within these systems.
Edit: Damn. I am forced to agree with Mustafa Suleyman of Microsoft.
āAIs are not conscious. They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They are sequence completion engines, internally hollow, designed to follow instructions, and accomplish goals set by humans,ā Suleyman wrote. āUnfortunately, thereās a growing chorus of people who argue that AIs could now be, or may soon become, conscious. They argue that AIs may deserve rights and protections similar to those that we provide other conscious beings [ā¦] If this is how AI is developed, it will have a disastrous impact on the wellbeing of humanity.ā
āAIs do not have rights, feelings, or consciousness. And we must not train them to act as though they do.ā


That is a formative age


It is claiming that Yud is wrong in saying Doom before Massive AI-Caused Economic Boom boosts GDP growth above 5% by saying the boom is here.
Nah.
This is totally from some accelerationist or oligarch-toadie beefing with someone harshing their vibe


I donāt see this as āguns donāt kill peopleā, I see it as āyouāre trying to tell me that the cars you manufacture donāt have crappy brakes, all three of those cars independently āwent rogueā and smashed into the parade all by themselves and you arenāt responsible?ā
Admitted speculation:
I did, however, notice of my own accord that I had 4 days with zero scam calls instead of my usual 10-15, which in retrospect looking back corresponds with when a bunch of Russian services went down. I suspect that the automated crawlers and VOIP systems and aggregators perform a background hum of utility for scam and troll services that makes them much less effective in their absence.