Hacker News Reader

sandgianton First image of a black hole gets a makeover with AI

The point of the Event Horizon Telescope (ETH) was actually to attempt to produce something like an image of the black hole at the center of the galaxy, so in a sense it was a pretty-picture mission. The point of this study was to attempt to link up the observations with models in "picture-space".

There are certainly other, less pretty, ways to looks at the data, but producing something that can be intuitively understood, such as a 2D image, can also be helpful scientifically.

sandgianton First image of a black hole gets a makeover with AI

The title from AP is misleading. They are not using any "AI" but a simple principle component analysis and a Monte Carlo minimizer.

> In this approach, we apply principal components analysis (PCA) to a large library of high-fidelity, high- resolution general relativistic magnetohydrodynamic (GRMHD) simulations and obtain an orthogonal basis of image components. PRIMO then uses a Markov Chain Monte Carlo (MCMC) approach to sample the space of linear combinations of the Fourier transforms of a number of PCA components while minimizing a loss function that compares the resulting interferometric maps to the EHT data.

As an exercise in matching up MHD models with real data this can be an interesting study. Too bad the reporting is so off point, as usual. But I guess adding "AI" to your titles gets you more clicks these days.

sandgianton Scientists finally figure out why the water bear is nearly indestructible (2017)

I asked ChatGPT what it would respond and it produces something very similar, down to the same sentences, to the comment above. So this definitely looks like a ChatGPT response.

sandgianton Hyundai promises to keep buttons in cars

I bought a used 2019 VW instead of a new model for this exact reason. I'm afraid it's the last VW I'll own unless they roll back touch controls.

sandgianton EU will require Apple to open up iMessage (2022)

Of course, that makes sense. So you could revoke the key for each app individually.

sandgianton EU will require Apple to open up iMessage (2022)

I'm curious about this too. At least now I have fairly decent confidence that when I send an iMessage to someone, Apple protects their identity to whatever standard they have. Whatever trust I put in Apple, at least it's a single point of failure.

What happens if interoperability is enforced and messages have to be end-to-end encrypted? Wouldn't that mean that any side-loaded Android app would have to be able to get hold of my friend's private iMessage key?

On iOS I guess you could still keep the key private through Apple's SDK, but what about other platforms?

sandgianton The “Build Your Own Redis” Book Is Completed

In the first code snippet in "Introduction to Sockets", shouldn't it be `close(conn_fd)` in the while loop?

That confused me a bit. Otherwise this looks interesting, thanks for sharing.

sandgianton Show HN: I “wrote” a kid's book with ChatGPT and Midjourney

I agree with this, but also note that a lot of children's books written by humans (I suppose) are terribly bland. My daughter has some Disney books which has illustrations from the films, and the writing in those is exceptionally bland. Same goes with some wonderfully illustrated, but terribly written, books from the 50s and 60s. I'd actually much prefer reading her GPT stories than those books which we have on the shelf (I guess I should get rid of them actually).

sandgianton 1min high-intensity exercise 3x a week improves fitness as much as 3x aerobics (2016)

There might be someone able to better explain this, but one way it could work wold be to say that for small samples any bad assumptions you make (such as variables or measurements being independent) will affect the result more than if you had a larger sample.

The assumption I make to make this work is that dependencies are more likely to be drowned out by internal variation in a larger sample. So you get to pick which assumption you like better.

sandgianton Amazon packages burn in India, final stop in broken recycling system

The community was created in the 70s. Back then it was more political, today it's more practical. The age demographic is more or less flat (0-85) with a dip around late teens (we're in the countryside to most young adults choose to move to the city). We own the property together and cook and eat together five days a week. We have various shared workshops, depending on what the need/desires are at any given time. While we don't farm the land ourselves (not enough land), we try to buy local and organic produce whenever possible.

Cooking for many people, sharing tools, common areas, etc, saves a lot of money and resources. We are thus able to support a wide range of income levels. With the caveat that we're in a fairly equal country (Northern Europe), the community attracts a wide range of backgrounds, from unskilled workers to teachers and doctors. Everyone gets chores and responsibilities assigned based on personal priorities, skills and the needs at the time.

We have a "rule-book" that defines a democratic voting system for making decisions. We hold meetings every other week. I would lie if I said it was always pleasant or easy, but it does work surprisingly well. I think it helps that the place is around 50 years old, so we've had the chance to learn from mistakes and update the "rule-book" along the way.

I don't think it's for everyone, and I'm not sure how well this scales beyond 50-100 people, but I do think that you can be happier, healthier and wealthier living in suitable small-scale communities, sharing everyday tasks and tools. Instead of everyone owning their own everything.

sandgianton Amazon packages burn in India, final stop in broken recycling system

We live in a community of ~100 people. We have a room full of other people's left over things that we can take (or add to) for free. 90% of the gifts and clothes for our, and other people's children, we source from that room. We are able to give away a lot more and higher quality stuff than if we would have to buy it and the kids are just as happy.

It's incredible how much stuff is produced just to be thrown away. But I guess it keeps the wheels turning? I often wonder wether this is the best system there is. Something tells me it's not.

sandgianton Madison Square Garden Uses Facial Recognition to Ban Its Owner’s Enemies

I have a VW from 2019. I think it's the last generation to only have Bluetooth connectivity. Still, every time I talk to an official VW employee they ask me to install their app and update the car firmware, so they can spy on me. I politely refuse.

I'm sure we'll figure out the regulation sooner or later, but for now it looks like things will get worse before they get better.

sandgianton The rise and fall of peer review

This argument makes no sense. Why would "less than average" referees block a paper? And why would editors select referees from the population at random? I don't know about other fields, but that's certainly not how things work in physics.

I would say it's more like a club where everyone knows each other. If you are authoring a paper you often know who the potential referees are (those that published most in the field), and you politely ask the editor to exclude referees that are in open conflict with you. Editors usually abide and you get a reasonable review of your paper.

I agree about most other issues raised about peer review in general and related incentives, but the issues raised here are not some that I recognize.

sandgianton How does GPT obtain its ability? Tracing emergent abilities of language models

I don't think anyone looks a biological systems and thinks they are well described by P(reproduction|environment). ChatGPT, on the other hand, is P(text|internet) more or less by definition (it's more than that, but not much more).

The fact that we observe what looks like emergent structures such as "understanding", "knowledge" or "reasoning" is fantastic, but it is not in any way incompatible with a P(text|internet) model simply "mimicking" humans.

I agree that exploring the inner workings of these emergent features is interesting in it's own right, but all that glitters is not gold.

sandgianton Mac OS 9

I love projects like these! I don't have anything to add to what others have already said about nostalgia and simpler times, except I watched the Civilization intro and almost started a play-through. The original Civilization tile set is so pleasant to look at.

sandgianton Using ChatGPT as a Co-Founder

It's even more astonishing (depressing?) to me how little credit people give humans these days. If the consensus is that ChatGPT pretty much passes for a human, my definition of a human is very different from that of the consensus. At the very least I'd like my humans to exhibit even minimal capacity for emotional expression as well as consistency and honesty.

I can't even believe I'm even spelling this out. What is the point of associating anything even remotely human with a chatbot like ChatGPT? I really don't get it. What benefit do we get from this? If anything is confuses us in ways that makes it more difficult to understand and argue about AI specifically, and technology generally.

ChatGPT is being grossly misinterpreted, and it's a shame, really, because it's such a cool piece of software.

sandgianton ChatGPT passes the 2022 AP Computer Science A free response section

So let me understand this. You're saying that the primary feature of human brains is deductive reasoning, and that deductive reasoning is equivalent to calculating chess moves. Since ANNs are a lot better than humans at calculating chess moves, they must be an analogy to human cognition? Can we infer from this that ANNs are already cognitively superior to humans?

I feel like something is missing in this equation. :)

sandgianton Building a Virtual Machine Inside ChatGPT

Fair enough. I can see how for a limited domain of understanding, it can be a reasonable analogue for compression. I still don't see the benefit of reducing the domain of a word just to be able to put a new technology on arbitrary scale of "intelligence". I think we should be able to appreciate the power of ChatGPT-like models without lossy compressing of our vocabulary in the process.

sandgianton Building a Virtual Machine Inside ChatGPT

Yes, it's brilliant! I love how novel things like these always end up on their own axis of reality, while everyone (including me) is trying hard to project them onto whatever predisposed axes we're hauling around.

sandgianton Building a Virtual Machine Inside ChatGPT

At the point before we actually understand how the models work and realize this has nothing to do with awareness. On the upside, we can learn a lot about ourselves when we realize this.

sandgianton Building a Virtual Machine Inside ChatGPT

On the contrary, it's important to realize that these models have no sort of understanding in any definition of the word. We can always decide to redefine the word "understanding", but that doesn't seem like it would provide any sort of benefit. Perhaps outside of avoiding a difficult conversation about the (huge) gap between ML models and consciousness.

sandgianton Building a Virtual Machine Inside ChatGPT

Thank you for posting this. I'm always a bit conflicted when "AI" news goes viral like this. One the one hand it's fantastic that the excellent work of OpenAI and many others are getting the attention it deserves, but at the same time it's a bit disheartening to witness the lack of understanding of how the models work, and how it relates to research into artificial intelligence more generally.

I guess that could probably be said about most research fields though. We've just not reached novelty-saturation in machine learning sciences yet. Let's enjoy the attention while it lasts!

sandgianton Building a Virtual Machine Inside ChatGPT

I'd love to get on the hype train here, but I really need for someone to explain to me how ChatGPT is more than just a toy. What is this going to be useful for exactly?

Another kind of knowledge exploration perhaps? If they find a way for it to not produce factually incorrect output. Isn't that a big if?

sandgianton Building a Virtual Machine Inside ChatGPT

Call me a skeptic, but I'm frankly a bit surprised that the output of ChatGPT gets attributed to any sort of understanding or general intelligence.

Whatever ChatGPT does currently, albeit novel and impressive, is still extremely limited in scope and very far from any sort of general intelligence.

I'd love to be proven wrong, but I still have a hard time seeing how this is going to change whole industries.

sandgianton Puter

This is great! Had a real good time completing Panda Love with my daughter. Thanks for sharing!

sandgianton The Death of Quanta Magazine

Looks like Quanta Magazine changed the title of the article after publication. I guess they did get some push back on the original article.

Original title (dec 1.): "Physicists Create a Wormhole Using a Quantum Computer" [1]

Current title (dec 2.): "Physicists Create a Holographic Wormhole Using a Quantum Computer" [2]

[1]: https://web.archive.org/web/20221201021130/https://www.quant...

[2]: https://web.archive.org/web/20221202022713/https://www.quant...

sandgianton FTX’s balance sheet was bad

> What Sequoia was reacting to was the scale of SBF’s vision. It wasn’t a story about how we might use fintech in the future, or crypto, or a new kind of bank. It was a vision about the future of money itself—with a total addressable market of every person on the entire planet.

> “I sit ten feet from him, and I walked over, thinking, Oh, shit, that was really good,” remembers Arora. “And it turns out that that fucker was playing League of Legends through the entire meeting.”

> “We were incredibly impressed,” Bailhe says. “It was one of those your-hair-is-blown-back type of meetings.”

https://web.archive.org/web/20221027180943/https://www.sequo...

sandgianton DreamFusion: Text-to-3D using 2D Diffusion

I think this is a misunderstanding of how these models work. The model does not understand anything at all. It's computing correlations. I could spend all day computing correlation without ever understanding what the correlations correspond to in the physical world. The correlations could still amount to a useful description of some physical phenomenon, but their interpretation requires much more than just the ability to compute correlations in some limited space.

If we elevate "correlations" to mean "understanding", we're quickly going to run out of words.

sandgianton Bikes, not self driving cars, are the technological gateway to urban progress

Sure they could. There are bikes made specifically for obese and elderly people.

If you prefer driving your car, that's fine, but don't pretend that biking is somehow reserved for young fit people, that's simply not true.

Go to any northern European city and see for yourself.

sandgianton It’s time to let the five stages of grief die (2019)

In reality Kübler-Ross developed her stage model after interviewing many individuals with life-threatening illnesses. It was only the experiences of these patients that she attempted to model.

I actually didn't know this, but that makes a lot more sense.