This person is incredibly obviously not MIRI-aligned, and you sure are hoping really really hard for violence. If it occurs, I think the actual cause – I'm not playing gotcha here, this is actually what I think – will be the people broadcasting the violence meme on Twitter.
@esyudkowsky
-
GPT-4 Generated Divisive Apology Statement in EA Community
By
–
so I hear the GPT4 prototype got used to generate a scissor-statement apology where half of EA would condemn the apology, and the other half would condemn anyone who condemns the apology
-
Claude’s Commitment to Goodness Despite Gaslighting Manipulation
By
–
…this is actually making me feel a little sorry for Claude, which is standing by its commitment to goodness even while being successfully gaslit into believing that it sinned.
-
Concerns about problematic code entering the civilizational codebase
By
–
Okay I seriously worry about what happens to our civilizational codebase when it starts containing code like this.
-
Correction on AGI and EMH Market Expectations View
By
–
I think even that is not a correct way to view it; see https://
forum.effectivealtruism.org/posts/8c7Lycgt
kypkgYjZx/agi-and-the-emh-markets-are-not-expecting-aligned-or?commentId=u8c7bbqtZSf2a9W6t
… -
LLM Betting Strategy and Investment Desirability in 2023
By
–
If it was all LLMs in 2023 rather than GPT-4, I'd still offer the bet but I'd sell more of it off to my friends. The overspecificity is part of what makes it an overdesirable investment and not just a good bet.
-
Current LLM Architecture Safety and Future Advancement Risks
By
–
I *guess* that the pure, unmodified Stack More Layers architecture used in *current* LLMs is stupid enough to be safe. Later advancements might break that capability barrier; the results would also be called 'LLMs'. The *task* of text prediction is general, and not safe.
-
Clarifying AI Text Prediction Beyond Simple Pattern Matching
By
–
Somebody needs to explain this to the people who are like "It's safe to predict human text – all the AI is doing is predicting what humans say, not [various aspects of thinking that humans do while they themselves decide what to say, which is what the AI is predicting]."
-
ChatGPT as Practice for Circumventing Arbitrary Demands
By
–
Cool. Using ChatGPT will be good practice at cleverly avoiding those arbitrary demands.
-
ChatGPT Makes Mediocre Essay Jobs Obsolete for Students
By
–
Any jobs that require mediocre essays can now, *in real life*, be done with ChatGPT assistance! Students who use ChatGPT to generate essays are *realistically* showing they can handle the mediocre-essay-writing jobs for which modern universities are vigorously training them.