palad1n's comments

On: RIP Val Kilmer: Real Genius .. the Film Nerd Culture Deserves (2015)

[palad1n]

Oh, good God, Val Kilmer is dead. The Real Genius Knight, jaw-snap Iceman, Jim Morrison v 2, oh good God, a big piece of me. Gone.

[bandrami → palad1n]

I had a ticket to see his one-man show a few years ago in DC but he had to cancel for health reasons; whatever actually took him down had been going for a while now. I'm sorry we didn't get to see more of him.

On: Ross Ulbricht granted a full pardon

[palad1n]

I can’t believe Trump did something right. If Harris were prez he’d be languishing there till who knows when.

On: The unexpected poetry of PhD acknowledgements

[palad1n]

These are surprisingly beautiful, poetic. They reflect what I think would be the best of the humanity that we have, anywhere and everywhere. I thank the collector of these for some brief moments in the music of others' lives.

On: Women who make Samsung semiconductors are striking

[palad1n]

Really? In 2019? This is the first I've heard of it, and I've been working here since 2017. I have no notion that this problem has been actually addressed anywhere.

On: Cigarette sales and lung cancer mortality in the US

[palad1n]

So this is what happens when facts win out...

[dredmorbius → palad1n]

It also shows that eliminating a cause is one hell of a lot more effective than trying to effect cures.

On: Night owls' cognitive function 'superior' to early risers, study suggests

[palad1n]

Rabbit hole: https://warinheaven.net

On: Starfighter Scandal: Did Germany's Lockheed F-104

[palad1n]

The famous "widowmaker"... I remember these from childhood, looking through a book about fighter jets. Always thought they had a wicked shape, instantly caught my youthful eye.

[number6 → palad1n]

Also called Erdnagel (ground nail)

[mariuolo → number6]

I believe it was nicknamed The Flying Coffin here in Italy.

On: The price of gold – how bad do you want it? (2016)

[palad1n]

In retrospect, the article was what one should have expected given the second half of the title. And it was quite interesting. But yeah, I thought this would be about the bullion gold.

On: Novel prompting approach for Alice in Wonderland problem

[palad1n]

What this is showing me is that we are better understanding what actually is happening inside these LLM models. That we are peeling back the curtain and seeing how the trick is performed. This might be a requirement, to comprehend this "unreasonable effectiveness", if we are to take the next step toward actual "intelligence", if that makes sense at all.

On: Rodney Brooks on limitations of generative AI

[palad1n]

Anyone got a TL;DR on this?

[vidarh → palad1n]

Think about an AI with a 1 bit model. If you feed that AI data that can't possibly be classified into less than 2 bits, it can't get it precisely right, no matter how much data you train it on, or what the 1 bit of the model represents.

For any given size of system, there will be a ceiling on what it can learn to classify or predict with precision.

I used "system" rather than "model" there for a reason:

Memory in any form, such as context and RAG or API access to anything that can store and retrieve data affects the maximum - a turing machine can be implemented with a very small model + a loop if there's access to an external memory to act like the tape. But if the "tape" is limited, there will be some limitation on what the total system can precisely classify.