David Björling

I like Asimov, and I love Clarke's storytelling. Reading his books, it amazes me how he seemingly predicted some of the technology we take for granted today. I can't help wondering if he may not in part have manefested his predictions, by inspiring the actual inventors. I have never read William Olaf Stapledon. A recomendation, I take it?

I wonder, are you planning to answer these two questions. You have no obligation to do so, obviously. Only if it feels constructive to do so.

Replying toThe Asteroid Setup That Demands an Explanation

I really enjoy imagining your last point, by the way ^^. I do not know if you meant to, but you paint a beautiful picture.

Replying toThe Asteroid Setup That Demands an Explanation

Ah! I'm glad I asked. So I had two guesses.

1) What if you could use the Cosmological Degradation as your entropy sink? What if you could tweak the asteroid sufficiently cleverly to make this work. The law about "entropy always increasing" would not be broken.

2) What if the structure of your setup was your gradient? What if the asteroid and its gravity well, the shell, the atmosphere and the energy conversion equipment would degrade in an edgecase like this in a way that could never be repaired from the energy converted. The law about "entropy always increasing" would not be broken.

Do you see anyting absolutely forbidding those two possibilities? And for 2, I would be interested in your intuition both for the asteroid setup, and in general.

Replying toThe Asteroid Setup That Demands an Explanation

Thank you, Anthony!

I learn some new things here. Mostly I learn about how much I do not know. I appreciate that. However:

Let us back away from what I have written as specific conditions.

Suppose we have a very advanced tech base to play around with. Anything not forbidden, they can build. Suppose we can jump into any time in the universe = any CMB temperature between say 3 to 3000 K. Suppose we can make an asteroid of any shape and any mass. Suppose we can add any type of atmosphere (our starting condition atmosphere).

Suppose we can shape the shell in clever ways, and attach it in clever ways at the most advantageous... (read more)

Replying toThe Asteroid Setup That Demands an Explanation

Good question!

Short answer is no. Here on Earth space "starts" at around 100 km. Above this we kind of have a vacuum. The same would be true for the asteroid (above a certain point the pressure is very low). As long as we release the lowered atoms above this point there would be no atmospheric pressure to fight against.

Replying toThe Asteroid Setup That Demands an Explanation

You raise a point about thermal equilibrium, and you're absolutely right that in a static equilibrium, temperature would be uniform throughout the gravitational field.

However, the system I'm describing isn't in thermal equilibrium as we "start" the setup (e.g. insert the atmosphere). It's in a state with continuous evaporative cooling. When the fastest atoms escape the gravity well, they remove more than the average kinetic energy from the atmosphere (that's why they can escape).

This is exactly analogous to evaporative cooling of water, just with gravity instead of intermolecular forces. The key is that the CMB at 5K provides continuous heat input to compensate for this cooling. So we have:

1. Fast atoms escape → top... (read more)

Replying toThe Asteroid Setup That Demands an Explanation

You're right that it superficially resembles the Earth atmosphere example, but there's a crucial difference that makes it more interesting.

In the Earth case, you're fighting against atmospheric buoyancy. The air you're trying to drop is surrounded by denser air, so no net energy gain.

In my setup, the helium that accumulates near the shell has escaped the asteroid's atmosphere entirely and is drifting in near-vacuum. When you capture it at the shell (e.g. 31 radii out, gravity ~0.1%), you can lower it through essentially empty space (no buoyancy to fight against).

The key insight: atoms that barely escape arrive at the shell with near-zero kinetic energy, creating a density enhancement. Most will arrive radially... (read more)

Replying toThe Asteroid Setup That Demands an Explanation

I add a rather dumb example too (please fill in the obvious blanks):

Suppose you threw tennis balls into the air, several balls each milisecond. Suppose you placed a jagged roof at hight h. Without the roof the tennis balls would travel to height 2*h. Suppose the jagged roof scatter tennis balls in all concievable directions. Would you accept that you get an accumulation of balls in the vicinity of height h compared to what you would get without the jagged wall?

Replying toThe Asteroid Setup That Demands an Explanation

David Björling7mo*

Okay. So, I will try to break this down into sections. We have what I believe are agreements, questions and possible disagreements.

To set the stage, let me start with agreements (most of the times followed by a “however”):

Agreements

The standard cheeky description: I 100 % agree. However, I think speaking of structural degradation may serve as a guide for finding ways of using up structure as a means of converting a fraction of heat used in the breaking of structure into useful energy.
Big-O: I 100 % agree, and I find your intuition “One of the things that I think you'll find approaches zero more quickly than most is the rate at which it

... (read 807 more words →)

Replying toThe Asteroid Setup That Demands an Explanation

Thank you Anthony! Truely!

There are many things in what you write that are unquestionably worth looking into.

As I stated earlier, I really think your knowledge as a materials scientist is invaluable for helping me understand this. I do have both questions, and objections. But I do not want to waste your time. If (and as long as) you are interested, I could write down some of my thoughts. Right now, I just fear, I might have presented something that was percieved as ugly, and improper.

If you think a bit of back and forth may be interesting, please let me know. I am quite confident you would find some of my objections and questions at least partly valid. ^^

Planet X, Lord Kelvin, and the use of Structure as Fuel

7mo

Introduction

I recently made a post named Planet X, Lord Kelvin, and the use of Structure as Fuel. I got some very insightful engagement from AnthonyC. This follow up post also contains a thought experiment. One that does not break the second law of thermodynamics, but where it may hard, and potentially useful, to find out why it does not (at least it was hard for me).

1. The Asteroid Thought Experiment

One way to summarize the Second Law of Thermodynamics is: “You can never build a perpetual motion machine.” Or more precisely: “No process can extract usable work without, on average, increasing total entropy.” Now, consider the following:

The setup:

An asteroid with helium atmosphere, surrounded

... (read 1421 more words →)

-2

7mo

On the nature of boiling

Few people know this, but boiling is a cooling effect. If you somehow lower the boiling point of water below ambient temperature, you will get boiling water, as it quickly cools down to its current boiling point. The easiest way to do this, is to create a partial vacuum with a vacuum pump. A glass of water inside the bell will start boiling at room temperature, as pressure drops.

This is a fun demonstration I have shown students. I always ask “Is there anyone brave enough to get boiling hot water poured in their hand?” There is always someone. The shock is universal, each time newly boiled water is... (read 892 more words →)

Neural Dimensionality

Output Convergence: A Safety Metric for Understanding AI Epistemic Stability

7mo

Background

In a raw, newly trained LLM you will have a haphazard structure, where the exact same block of text may map onto lots of different neural activation patterns in the first layer of your LLM. Different seeds may lead to radically different activation patterns. This may be partly based on context that may lay outside the text itself, and partly based on lingual patterns from many, many humans. Different humans might e.g. respond to the very same question.

However: For the human brain I believe there are certain contextual labels that every block of text can be put in relation to. One might view these contextual labels as a kind of dimensionality. If... (read 3591 more words →)

LLMs Can Amplify Genocide Rhetoric — I Found Failure Modes, Tested Fixes, and Now Seek a Collaboratoritled

7mo

1. Introduction: Why This Matters

Over the last couple of months, I have come to realise how unsafe AI models are. Anthropic’s Claude is perhaps the safest available today, and yet it can be “fooled” into creating unsafe content by a fig-leaf-thin pretence of being a “writer”, or a “scientist” or a “best friend”. Even more troubling, it sometimes keeps going after seemingly realising the pretence is bogus.

I believe a large part of this problem stems from a failure of epistemic consistency. This post introduces a diagnostic metric I call Output Convergence, which aims to measure and eventually improve epistemic stability in large language models (LLMs). I also introduce a partial mitigation strategy... (read 1683 more words →)