wunan

Replying toAre We In A Coding Overhang?

wunan2mo

Are We In A Coding Overhang?

Not necessarily. You can have a superhuman centaur (AI+Human) system without the AI alone being superhuman.

Replying to$500 Write like lsusr competition

wunan2mo

$500 Write like lsusr competition

By "high-effort" I mean that I feel posting them on Less Wrong makes the world a better place.

That is an extremely unusual usage of the term "high-effort."

Replying toWhy do LLMs so often say "It's not an X, it's a Y"?

wunan2mo

Why do LLMs so often say "It's not an X, it's a Y"?

More specifically, they may be emulating the Kenyans who were hired to create much of the training data. "I'm Kenyan. I Don't Write Like ChatGPT. ChatGPT Writes Like Me."

Note: I can't verify that the post I linked is legitimate. For all I know it could be generated by ChatGPT instructed to emulate a Kenyan writing about ChatGPT. HN discussion here.

-1

Replying toNorth Sentinelese Post-Singularity

wunan2mo

North Sentinelese Post-Singularity

Why?

Replying toNorth Sentinelese Post-Singularity

wunan2mo

North Sentinelese Post-Singularity

That is easily patched via full-dive VR. You can keep your biological body. No upload be necessary.

Replying toNorth Sentinelese Post-Singularity

wunan2mo

North Sentinelese Post-Singularity

Have you read Friendship is Optimal? Is that outcome unappealing to you in a way which can't be easily patched (e.g. removing the "become a pony" requirement)? Do you think it would be unappealing to almost everyone in ways that can't be easily patched?

Replying toHave there been any rational analyses of mindbody techniques for chronic pain/illness?

wunan2mo

Have there been any rational analyses of mindbody techniques for chronic pain/illness?

Spencer's post isn't loading for me and I don't see any post about this on his facebook feed -- is the link right?

Replying toHave there been any rational analyses of mindbody techniques for chronic pain/illness?

wunanDec 08, 2025

Have there been any rational analyses of mindbody techniques for chronic pain/illness?

Specifically for ME/CFS and Long Covid, I recommend s4me.info. Pretty much all of the major studies on the mind-body methods already have threads there with discussions. The tl;dr is that they are extremely low quality studies in ways that will not be a surprise to anyone familiar with the replication crisis and these techniques very likely do not work for ME/CFS or LC.

This blog is good as well: https://mecfsscience.org/

wunan2moQuick Take

Anthropic researchers estimate that Opus 4.5 provides 2-3x speedup to their research, if I'm reading this correctly. This seems very important and I'm surprised I haven't seen more discussion of it.

Twitter thread: https://x.com/HjalmarWijk/status/1993752035536331113

Unrolled/without login required: https://twitter-thread.com/t/1993752035536331113

@HjalmarWijk
Nov 26
Anthropic says in their system card that *all* their AI R&D evals are close to saturation, and report a median self-reported uplift of 2X (mean over 3X!) for power users. They provide very little evidence ruling out imminent dramatic AI R&D acceleration.
I personally suspect that their self-report uplift numbers are inflated and that agent time horizons are still limited. But if taken at face value, then even the most aggressive scenarios (e.g. AI 2027 or https://blog.redwoodresearch.org/p/whats-up-with-anthropic-predicting) would have underestimated progress.

I didn't quote the whole thread, there's more if you follow the link.

•••

Replying toI'll be sad to lose the puzzles

wunan3mo

I'll be sad to lose the puzzles

I recommend Deep Utopia for extensive discussion of this issue, if you haven't already read it.

If you have Long COVID or ME/CFS, or want to learn more about them, I highly recommend https://s4me.info. The signal to noise ratio is much better than on other forums for those topics that I've found. The community is good at recognizing and critiquing low vs high quality studies.

As an example of the quality, this factsheet created by the community is quite good: https://s4me.info/docs/WhatIsMECFS-S4ME-Factsheet.pdf

What is the research speed multiplier of the most advanced current LLMs?

wunan

In Sam Altman's recent interview on Bill Gates' podcast, he said AI tools currently can speed up a programmer by about 3x. It's unclear whether he's referring to publicly available tools like copilot, or internal tools at OpenAI.

Does anyone know whether he's referring to private or public tools? Is that 3x figure based on any public research?
More speculatively, if programmers are sped up by 3x, how much are AI researchers sped up, particularly using whatever internal tools are currently available at OpenAI? The job involves more than just programming, but many non-programming tasks can be sped up as well.
Even more speculatively, how much do you predict GPT-5 will speed up researchers at OpenAI?

Avoiding "enlightenment" experiences while meditating for anxiety?

wunan

How can I meditate a lot (3-6+ hours/day) while deliberately avoiding any non-dual/awakening/enlightenment/etc experiences, so that I can just get the anti-anxiety benefits?

For context, the reason I want to do this is that I have chronic fatigue syndrome and recently developed severe anxiety and panic disorder on top of it. Due to the CFS, I already needed to spend a large portion of the day sitting still with my eyes covered, but the anxiety has made this much less pleasant, so I think meditating can help.

I've read The Mind Illuminated in the past (years ago) and fairly quickly got up to stage 4/5, then decided I didn't want to go further after reading various sources on awakening/stream entry/etc.

Currently I'm planning to do the techniques in TMI up to stage 4 but nothing past that. Are there better meditation techniques than those described in TMI for someone who explicitly wants to avoid awakening and just wants the anxiolytic effects?

Any other advice is also appreciated.

COVID contagiousness after negative tests?

wunan

If you've recently had COVID, how do you know when it's safe to stay with an elderly or immunocompromised person? If you've had two negative tests 48 hours apart, what is the risk that you're actually still contagious? Is it more like 1%, .1%, .01%?

If compute is the main bottleneck to AI progress, then one goalpost to watch for is when AI is able to significantly increase the pace of chip design and manufacturing. After writing the above, I searched for work being done in this area and found this article. If these approaches can actually speed up certain steps in this process from taking weeks to just taking a few days, will that increase the pace of Moore's law? Or is Moore's law mainly bottlenecked by problems that will be particularly hard to apply AI to?

I have a potential category of questions that could fit on Metaculus and work as an "AGI fire alarm." The questions are of the format "After an AI system achieves task x, how many years will it take for world output to double?"

wunan's Shortform

wunan

This is a special post for quick takes (aka "shortform"). Only the owner can create top-level comments.

LESSWRONG
LW

LESSWRONG
LW

OpenAI charter

Avoiding "enlightenment" experiences while meditating for anxiety?

COVID contagiousness after negative tests?

What AI newsletters or substacks about AI do you recommend?

wunan

wunan

What is the research speed multiplier of the most advanced current LLMs?

Avoiding "enlightenment" experiences while meditating for anxiety?

COVID contagiousness after negative tests?

What AI newsletters or substacks about AI do you recommend?

wunan's Shortform

OpenAI charter

wunan

OpenAI charter

Avoiding "enlightenment" experiences while meditating for anxiety?

COVID contagiousness after negative tests?

What AI newsletters or substacks about AI do you recommend?

wunan

wunan

What is the research speed multiplier of the most advanced current LLMs?

Avoiding "enlightenment" experiences while meditating for anxiety?

COVID contagiousness after negative tests?

What AI newsletters or substacks about AI do you recommend?

wunan's Shortform

OpenAI charter