HNHacker News
TopNewBestAskShowJobs

l33tman

1,775 karma · joined January 18, 2019

submissionscomments
l33tman··on Inside the proton, the ‘most complicated thing you could possibly imagine’
That does not follow. What inaccuracies are you thinking about?
l33tman··on Inside the proton, the ‘most complicated thing you could possibly imagine’
It reflects the fact that the measurement apparatus, environment in general and the observer in particular are all also part of the "wave function". Until these are linked to the system to be observed (like the proton), the system has to be analyzed as the full set of possible configurations of its constituents.

After you link it up with the apparatus, it is pulled into the system as a new part of it, and so on. The more stuff you pull into the system, the less number of different configurations you sum over, eventually you end up with a single configuration with an observer that has a clear measurement result in her mind. This process is what is called "wavefunction collapse" in old physics texts (and a lot of modern pop-sci accounts).

What sets this apart from other purely probabilistic theories is that it's non-local, the entire linked system from the proton to the measurement apparatus to the observer has to be taken into account in the calculations if you are to be thorough. In local theories you can separate the parts of the system and handle them separately, like you could say "there is a 1% probability that a charm quark pops out in the upper left part of the proton every second" etc. You can't really do this in QM/QFT, and this is what in essence leads to all of the counter-intuitive results and confusion..

l33tman··on Inside the proton, the ‘most complicated thing you could possibly imagine’
QFT is the most well-tested physics theory mankind has developed and certainly is not a "haze" as the pop-sci articles write.

What they mean by the "haze of probabilities" is that you need to iterate over all possible configurations of the gluons and quarks in the proton to produce experimental predictions.

It is for sure computationally ludicrous, but conceptually it's really not better or worse than to say that particles are billiard balls moving around by Newton's Equations. You're just more used to the latter, but spend some time working with for example Lattice QCD and you get completely used to the former being the "actual" underlying physics rules.

l33tman··on Hi everyone yes, I left OpenAI yesterday
It sure is, but the theme in the sub-thread was about if OAI in particular can afford to do that (i.e. wait) while there are literally dozens of other companies and open-source projects showing they can solve a lot of the tasks GPT-4 does, for free, so that the OAI value proposition seems weaker and weaker by the month.

Add to that a company environment that seems to be built on money-crazed stock option piling engineers and a CEO that seems to have gotten power-crazed.. I mean they grew far too fast I guess..

l33tman··on Stable-Audio-Demo
Emad hinted here on HN the last time this was discussed that they were experimenting with exactly that. It will come, by them or by someone else quickly.

Text-prompting is just a very coarse tool to quickly get some base to stand on, ControlNet is where the human creativity again enters.

l33tman··on Explaining the SDXL Latent Space
Just a terminology comment here. "Latent space" means a lot of different things in different models. For a GAN for example it actually means the "top concept" space where you can change the entire concept of the image by moving around in the latent space, which is notoriously difficult. For SD/SDXL it refers to the bottommost layer just above pixelspace, which expands the generated image from 64x64 to 512x512 pixels in the case of SD1.5.

This allows the rest of the network to be smaller while still generating a usable output resolution, so it's a performance "hack".

It's a really good idea to explore it and hack into it like in the article, to "remaster" the image so to speak!

l33tman··on Explaining the SDXL Latent Space
It's only patterns and textures for 8x8 images, so I guess it could make sense, you're not going to need every conceivable pattern of 8x8 pixels in normal images..

If you quantize those 4 floats per 8x8 block, is that encoding better than say the old venerable JPG 8x8 DCT + quant?

l33tman··on Alaska Airlines flight 1282 NTSB preliminary report [pdf]
The idea is that the system as a whole should be resilient to a certain degree of this, as it will be caught in other places or at least reported at a (hopefully non-fatal) aviation incident that leads to a report and analysis, at which point you backtrack and issue recommendations that actually have clout (like, you can't fly with your plane anywhere in the world if you don't fix this).

Of course if the entire industry is corrupt it doesn't work, but to a certain degree I guess it is robust. This time it led to a potential disaster but it will become safer as a result..

l33tman··on Vesuvius Challenge 2023 Grand Prize awarded: we can read the first scroll
This is awesome. But it makes me sad to think that our current DIGITAL artifacts might not survive even 20 years, unless explicitly archived. I'm pretty sure a harddrive or SSD thrown on a landfill won't be able to be resurrected in 2000 years even with alien tech, but maybe? Would be cool if someone studied this :)

But the main problem is probably not reading thrown out drives - it's that stuff is just too transient nowadays. People put something up on the net and decides to withdraw it next year. archive.org can't store the walled gardens, and even if a social network for example wants to archive stuff they might not be allowed to for legal issues (not that it has prevented them before, but anyway...)

l33tman··on How Doom didn't kill the Amiga
I didn't know that! However the CD32 was at that point in history (1993) a blip, it was far too little far too late. It sold 125,000 units, while the PC clone market was shipping 37 million units in 1993.

I agree fully with the OS issues, but for the purposes of gaming, it turned out to not be an issue really I guess..

l33tman··on How Doom didn't kill the Amiga
I hand coded asm demos and music trackers for Amiga since 1986, and there is no need to be rude on HN.. keep that to reddit :)

This was ages ago, but as far as I can recall, the main obvious limitation of the AGA chipset compared to the PC cards was 1 byte-per-pixel modes (I think they call it "chunky" modes now) so the simple texture-mapping 3D loops you could hand-code on the x86 were a pain on the Amiga. I'm sure now 30 years later people have figured out cool hacks to do this better but back then it didn't work, and adding to this was the fact that the PC cards were improved like every 6 months. Commodore went bankrupt 1 year after I bailed, and I didn't look back.. (though I did miss AmigaDOS real multitasking for some PC/Mac generations after :).

The A1000 I had was a great machine in 1987-1990 but somewhere after that the whole Amiga concept started to feel dated (IMO). The PC architecture won because it was essentially hardware open-source, and hundreds of vendors could add features to the platform together. Commodore would have had to do every single decision spot on correct for years in order to stay competitive as a single company. Look at what happened when IBM tried to tame the beast and released the PS/2 that was "closed source"..

l33tman··on How Doom didn't kill the Amiga
I sold my 6 months old Amiga 4000 when I realized the PC VGA cards coming out around then were cheap, had 1024x768 resolution, blazing DOS pixel writes and nothing of value had happened with the Amiga chipset from 1985 to 1992..

Remember TSENG LABS? :)

I also think I saw Ultima Underworld or some of the variants on PC before this, rendering full free-view 3D in a sub-window at least. Was really blown away by that, this was before Doom came out.

l33tman··on Apple Vision Pro and ADHD
I find at least a little bit funny that it seems that completely new people now "discover" what you can do with VR, as if this hasn't been a thing for almost 10 years and certainly for the last 5 years (although with slightly lower resolution than with the AVP).

But then again, it helps the entire industry that it apparently has reached a different audience this time who gets hyped again.

I just wish there could be new use-cases than just the same old recycled use-cases like virtual monitors, watching movies, looking at how it is to stand on the moon, or virtual conferencing. Maybe there really are no killer VR apps, except beat saber :/

Wasn't AVP supposed to be an AR headset? Or they just haven't really figured out what to do there? Now it's just a VR headset with passthrough, it isn't really "using" the external world.

l33tman··on Damn Small Linux 2024
Yeah the 2/8 combo was probably what we went with in the product as well. The 512k was more like a shoehorned concept demo in an existing product.

The next thing we did was make a version of our CPU with an MMU, designed to work optimally with Linux (the first version was on the uClinux concept, with a kernel without MMU support and user-space programs that couldn't rely on fork() or mmap() fully). After a year or 2 with MMU-less Linux, it was like heaven to be able to run on an MMU :)

l33tman··on Damn Small Linux 2024
I had the Linux kernel and some simple user-space tools like busybox running on an embedded platform with 512 kB RAM and 2MB flash back in 1999! Those were fun times. To be honest 512 kb was possible but very on the limit, I think the product we launched with it had a few megs of RAM eventually. We had to invent a journalling flash filesystem as well in order to make it work in practice, something that didn't exist back then either. But Linux then was really a breakthrough compared to the horrible mess of embedded OSes that were needed otherwise to handle TCP/IP, filesystems and multitasking.
l33tman··on On a Tesla: Don't
My 2012 car has no spare tire, you are instead mandated to use run-flat tires (on those, you can still drive 80 km at a lower speed). This is pretty common.
l33tman··on Lumiere: A space-time diffusion model for realistic video generation
Please see https://arxiv.org/abs/2306.05720

Beyond Surface Statistics: Scene Representations in a Latent Diffusion Model

"... In this work, we investigate a basic interpretability question: does an LDM create and use an internal representation of simple scene geometry? Using linear probes, we find evidence that the internal activations of the LDM encode linear representations of both 3D depth data and a salient-object / background distinction. These representations appear surprisingly early in the denoising process−well before a human can easily make sense of the noisy images. ..."

l33tman··on Lumiere: A space-time diffusion model for realistic video generation
It has been shown that at least still-image generators learn a 3D representation internally and uses it to bootstrap their generation. If you think about it this is the only way they can be so good at shadows and reflections, perspective and lighting etc.
l33tman··on Direct pixel-space megapixel image generation with diffusion models
LAION is good
l33tman··on Nightshade: An offensive tool for artists against AI art generators
I know what you're saying, and for sure existing models can be difficult to force into the really weird corners of the distributions (or go outside the distributions). The text interfaces are partially to blame for this though, you can take the images into Gimp and do some crude naive modifications and bring them back and the model will usually happily complete the "out-of-distribution" ideas. The Stable Diffusion toolboxes have evolved far away from the original simple text2image interfaces that midjourney and dalle use.

The models will generalize (because that's the most efficient way of storing concepts) and you can make an argument that that means they understand a concept. Claiming "it's not learning concepts, only statistical probabilities" trivialises what a modern neural network with billions of parameters and dozens of layers is capable of doing. If a model learns how to put a concept like line width 5, 10 and 15 pixels into a continuous internal latent property, you can probably go outside this at inference at least partially.

I would argue that improving this is at this point more about engineering and less about some underlying unreconcilable differences. At the very least we learn a lot about what exactly generalization and learning means.

l33tman··on Nightshade: An offensive tool for artists against AI art generators
You're wrong in your concept of how AI/ML works. Even trivial 1980's neural networks generalize, it's the whole point of AI/ML or you'd just have a lookup-table (or, as you put it, something that copies and pastes images together).

I've seen "infographics" spread by anti-AI people (or just attention-seekers) on Twitter that tries to "explain" that AI image generators blend together existing images, which is simply not true..

It is however the case that different AI models (and the brain) generalize a bit differently. That is probably the case between different humans too. Not the least with for example like you say those with photographic memory, autists etc.

What you call creativity in humans is just noise in combination with a boatload of exposure to multi-modal training data. Both aspects are already in the modern diffusion models. I would however ascribe a big edge in humans to what you normally call "the creative process" which can be much richer, like a process where you figure out what you lack to produce a work, go out and learn something new and specific, talk with your peers, listen to more noise.. stuff like that seems (currently) more difficult for AIs, though I guess plugins that do more iterative stuff like chatgpt's new plugins will appear in media generators as well eventually..

l33tman··on LG washing machine sending 3.7GB of data a day
I sometimes use it to start the machine remotely when I'm away the whole day, so I can time it to be ready when I get home and can switch over the clothes to the tumbler immediately.
l33tman··on Contra Wirecutter on the IKEA air purifier (2022)
Hacking them to fit another purifier reminds me of the Apollo 13 incident, which included the astronauts having to do a McGyver hack to fit a square CO2 filter into a socket designed for a round one... using tape, hoses and other plastic items they had lying around :)
l33tman··on Show HN: Better Seater – Wedding seating chart optimizer
Do any of you happen to have any thoughts wrgds to OR-tools vs optaplanner? I've been trying to use OR-tools for a vehicle routing problem (with lots of constraints) and found it is really bad at finding solutions in our case, and it has only auto-gen documentation more or less, very sparse set of examples and a single dev supporting on the forums it seems (still thankful for that it's open source and HAS some support!).
l33tman··on Japan Goes All In: Copyright Doesn't Apply to AI Training
Sounds similar to the lawsuits against Google by the newspapers of the world for them providing excerpts of their articles on the search result page or so.

So essentially you can ask chatgpt to fetch and summarize a paywalled article, because openai has subscribed their crawler?

l33tman··on DaedalOS – Desktop environment in the browser
Creepy analogy available here with our universe-as-a-simulation
l33tman··on San Francisco’s rent prices have never returned to pre-2020 levels
It doesn't have to be 200k above in a single transaction to upset the "poorer" market. When you've just sold your house at a big capital gain, you're more willing to pay just a little bit over the average market value if you find something nice, even without being conscious of the fact (buying something is never a perfect decision, you have a span and can be pushed in either direction a bit). And that little bit over feeds the trend delta when lots of buyers think like that over time.

I think that logic ties in with the theme of the article as well. Supply and demand regulates markets in a subtle way, it's not a binary input like "people want to live in SF or not". Maybe it's enough that 1/10 of the people who wants to move there changes their minds when they see the housing market.

l33tman··on Quantum computing's reality check
What the OP suggested was that "it is a scam". It clearly isn't, there is nothing in QC that isn't traditional QM which is a 100 year old incredibly well-tested theory. As far as we currently know, it's an engineering problem, not a question of "will it work". If there is evidence to the contrary, I'm sure any journal would be happy to publish it (that would make it to Science or Nature, not a quantum-computing specific journal).
l33tman··on Flare, a video sharing site built on Nostr
Just vaguely reminds me of: https://www.techdirt.com/2022/11/02/hey-elon-let-me-help-you...
l33tman··on Google Imagen 2
They never released Imagen 1 either, why do they even do these "releases"?
← PreviousPage 5 of 22Next →