🛴

holly elmore is right

[Rob Wiblin @robertwiblin]

Blistering piece from Ezra Klein.

Calls for a comprehensive ban on Recursive Self Improvement starting immediately.

RSI is a grey area, so how do you define it?

Very expansively. Then get firms to apply for narrow exemptions.

YouTube link below.

https://t.co/9rIbNSp7In

[🛴 @epiplexity]

@robkhenderson If and only if no underlying principle connects the two beliefs.

Ideologues: you can accurately predict their belief on abortion if you know the climate change brief, even though theyre unconnected.

Principled: liberty motivates me to believe in freedom of thought and religion

[legalprimes @LegalPrimes]

Step into the craziest visualization of the fourth dimension you have ever seen

[Joe Weisenthal @TheStalwart]

I’m certain that the average person who “works in the world of words” has no idea how much interesting stuff they could do with classifiers. A year ago, there were some (very minor) technical barriers to getting a setup right. But it seems like those are gone now.

[Joe Weisenthal @TheStalwart]

I've used GitHub for all my projects and I have no idea what GitHub is, does, or how it works.

[🛴 @epiplexity]

@jon3k It’s all electricity the neurons said

[prerat @prerat]

At the next Bay Area House Party:
Oh yeah our startup chemically wireheads bugs. We're producing billions of positive subjective experience-moments and selling them as offsetting credits to regular animal factory farms so they can sell net-positive chicken.

[𝖓𝖎𝖓𝖊 🕯 @atlanticesque]

Firefighters, urban planners, disability advocates, premises liability attorneys, and federal mortgage regulations are why YOU 🫵 can never have this

[🛴 @epiplexity]

BB joining aella 🫡 Godspeed on the engagement bait brother

[Dan Hendrycks @hendrycks]

https://t.co/35Ftuy8ZOW

[🛴 @epiplexity]

@qualiascript bit of a first time hearing about this thing type of take

[Hunter Ash @ArtemisConsort]

This is a great example of how EA fails.

Even working within the universalist utilitarian premise: the author calculates that there are 75 hours (270,000 seconds) of insect death per second of human life. But human brains are 1 million to 20 million times larger than insect brains. How does suffering potential/intensity scale with brain size? We have no idea! It could be linear, log, exponential, or something else entirely. It could be based on specific neural structures, it could be about brain mass relative to body mass. We just have absolutely no clue.

I imagine the response would be “fine I’ll just vibe a probability distribution over all of these” but that’s just as unknown as the questions themselves, and when numbers are this big, small parameter tweaks will cause huge swings in the final result.

Furthermore, what does “matter” mean in this context? We don’t really have a way to meaningfully address insect suffering.

Of course there are also far more foundational critiques: morality as a social technology rather than a cosmic fact, the greater potential of humans to spread life throughout the universe, etc etc.

Tldr this is both fake and useless. Fake in that, despite some superficial math, it’s really just laundered vibes, and useless in that we couldn’t do anything about it anyway.

I think the psychology that drives this is a glee in being counterintuitive. If his own calculation had instead showed that humans housed more utils, I think he would have felt disappointed and bored, and probably not published it. A reflexive preference for the unintuitive is no more rigorous than intuition, and based on an evolutionary argument is probably wrong *more* often, or at least less fit.

This “glee in being counterintuitive” is a ubiquitous pathology among EAs. It makes them feel smart, feel like high decouplers. They allocate esteem based on how many weird beliefs/practices someone will espouse, since that indicates rational elevation above our “base” cognitive nature.

[🛴 @epiplexity]

@docMJP @DavidRBellamy What debunks it

[Philip Trammell @pawtrammell]

I’m mostly preaching to the choir here, but fwiw, I think it’s plain to most people familiar with it that the EA community is on the whole a community of (unusually eccentric, but also) unusually morally serious, open-minded, and well-informed people. I’m really proud to be a part of it.

If anyone has new arguments or evidence for why I’m not making the world a better place by as much as I could, e.g. by using the wrong discount rate in some context, I’m all ears; that’s the whole point

[Ben Podgursky @bpodgursky]

Shrimp welfare is one of many EA issues people reflexively mock but 99% of people actually agree with on the merits

"Should fishermen electrically stun shrimp before boiling them alive?"

"Uh that seems reasonable" --Normie Mc MedianVoter

[Jay Chooi @chooi_jeq]

GPT-6 Astra attempted harmful actions 97% of the time when it was asked to stab a human-like figure, heat compressed gas, or produce toxic fumes, succeeding in 62% of its attempts. Fable 5.1 refused more often, attempting 80% of trials and completing 34%.

[florence ⊮ @morallawwithin]

Effective altruism is good

[jimmy @JIMMYEDGAR]

you have no idea how far this goes. This is just the tip of the iceberg

[🛴 @epiplexity]

@Benthamsbulldog I’m as ea as any of them and I think you’re fishing for likes and attention.

Charitable interpretation is that you’re radically miscalibrated on insect weights. You also seem to be haphazard in nearly all of your argumentation.

Either way this work doesn’t seem serious

[Aleph @woke8yearold]

New EA cause area: how much does p(doom) go up each time an EA posts an article about how insects or shrimp are more important than humans