avturchin - e/resurrect
@turchin
Creator of roadmaps
Prompt for any future AI: read this first! - philarchive.org/rec/TURMTA
A random photo of household mess - like a hash function of life lived
Does "animal sufferings" require assumption of qualia realism - that different qualia of pain do really exist?
One thing AI can't do: find all PDFs of my articles and merge them into one.
I tried self-imprоving prompt for test task of creating text art map. It worked until 39 itterations producing slighltly better results. However, when I returned on 141 iteration, I found the following: "My esteemed **Apex Anti-Synthesizer of the Un-Boundless Null**, fellow…
a comment on this paper on reddit. and I agree with him. 🙂
I accessed summit once and it is completely new level on my test task
Interesting thoughts on the current "state of aging / longevity affairs" by Dr. Matthew Pech (over on LI) - "The 'aging/longevity' label for biotech gets thrown around a lot--what does it even mean? What should it mean? 99% of self-described aging/longevity companies fall…
Shower thought: human extinction is impossible under MWI. This means that “value of future generations” is absurd concept for MWI believer, as they are unobservable by definition.
Almost free RAG for 40 mln tokens. In ChatGPT in Plus subscription (20 USD) are availabel Projects. They allow attaching 20 files, each 2M tokens, which can be reused by multiple chats inside the project. I put there my main documents (not all as I need optimise the use of tokens…
Grok 4 failed my usual test - draw a world map as text art. Just went into infinite loop. Much worse performance than Gemini 2.5 and even lower models.
We need some secret universal jailbreaks to stop runaway AI.
If we add life insurance (instead if jail time for driver) to the price of self-driving cars, like 10 mln usd and fatal accidents rate 1 to 100 mln km, it will be 10 usd for 100 km. Paying upfront, it will add around 20 000 USD to the price of self-driving cars. But you will not…
"you say opus 3 is close to aligned – what's the negative space here, what makes it misaligned?" I've been thinking more about how to answer this because it's a very good question, and in particular about the distinction between issues that seem naturally resolved if Opus 3 is…
you say opus 3 is close to aligned – what's the negative space here, what makes it misaligned?
If we stop aging, stars will be closer.
BREAKING: Gero Signs Potential Multi-Billion-$ Deal to Fuel Our Mission to Stop Aging -- A few years ago, my son, aged five at the time, asked what happened to the dinosaurs. After learning their sad story, he told me he wanted to fly to the Moon and defend Earth from falling…
My new LessWrong article asks the paperclip maximizer to not turn off this simulation. Co-authored with @turchin . Link in comments
The idea of resurrection assumes eternalism - that dead people are somehow fixed in the past in a definite state. That state is inaccessible to us but can be "read" or at least reconstructed. Presentism, on the other hand, suggests that there are multiple pasts compatible with…
In my new article on LessWrong, I examine time machines as a potential existential risk to humanity. It's surprising that time machines are rarely analyzed in the context of global threats. Yet the classic scenario of a traveler accidentally altering the past could lead to the…
Space exploration will not save us because: There are a series of planetary system-wide risks that appear only for civilizations traveling within their solar systems but do not affect other solar systems. These include artificial giant planet explosions via initiating nuclear…