monitron

It's interesting that they used the Planet Express building from Futurama as one of their examples of 3D-inconsistency, because I'm pretty sure the exteriors are in fact computer-generated from a 3D model. Watch the show and you can see the establishing shots usually involve a smooth complex camera move around the building.

I just want to meet the dev who wakes up in the morning, sits down with their cup of coffee, rubs their hands together and says "Boy, I just can't wait to make a website where scrolling feels like trying to run across an ice rink!"

Yep. I am a current Tesla Model 3 owner, really like my car, but plan to switch to another brand after running this one into the ground. Elon's behavior is awful and dangerous given his fame/money, and luckily, unlike in 2017, there are now other good EV options which are getting better all the time.

This is fun and has a great symmetry but my brain breaks at the idea of the infinitely fast and large computer that would be required to run it all :)

I think the coolest thing about this is the way time works. One step at each level is computed using a large number of steps at the level below. So as you zoom out you are sinking into an exponential vortex where each step takes days, years, millennia to change.

But ingeniously, the simulator smoothly scales its speed as you zoom, totally erasing the fact of time. I wish there was an indicator somewhere of how much game time was going by per real-life second.

EDIT: ...but in order to do that you'd have to declare one of the levels to be the "bottom" level, the one that runs in real time, and that would ruin the fractalness of it all...

Yes! The delightful reason why this configuration can repeat within itself is that this Game of Life pattern is a machine with the rules of Game of Life built into it. It's calculating and displaying the game step-by-step. If you watch the machinery you can see the "data" flying around between cells and then the cells turning on and off.

This pattern isn't unique – you could certainly come up with infinitely many more machines in Game of Life that do the same thing. But yes, most Game of Life patterns aren't fractal like this.

I know nearly nothing about this field and yet at least one big reason is obvious to me:

A country set up for global trade is not ready to lose access to that trade. They depend on things not produced locally, and depend on the proceeds of exporting things they produce in surplus. It would take decades to adapt by returning to a self-sufficient economy. And some goods (phones, for instance) could barely exist without global trade, so an embargoed country will lose access to those things inevitably.

There is a set of radio buttons in the top right corner you can use to choose a different one :)

It happened in my neck of the woods (Maryland) a few years ago: https://www.washingtonpost.com/dc-md-va/2022/11/27/plane-cra...

Right now this display has the problem that the darkest color you can achieve is the color/brightness of the light coming through the screen. A sufficiently bright background will wash out the image entirely.

Why not put a grayscale LCD panel behind the MicroLED screen to allow the display to be locally opaque? Control it with an alpha channel for the whole display. Then you could have windows or objects appearing to float in space, with full dynamic range available.

I am having good luck with Eaton so far. Rock solid and their network monitoring is noticeably less painful to use than APC's!

The thing that made it unexpected in my case is that “borders” is an English word and was not previously known to me as a name. I was surprised and amused when I learned it was an eponym. That puts it in the same category as the other examples on this site, at least from my perspective.

Am I being detained, Mr. Policeman?

Borders Books (for anyone who remembers it) was named after founders Tom and Louis Borders. Louis went on to found Webvan (for anyone who remembers that).

"We designed [the Quest 3] to weigh 120 grams less" implies that the AVP is the benchmark by which other headsets are measured. It's funny to say you "designed it to weigh less" than a product that wasn't even unveiled yet.

Thanks so much for the thoughtful and insightful response!

I'm not sure a nail gun is the best analogy; it's a very simple tool that does one thing exactly as "instructed," and can't do any of the "thinking" for you.

I’m curious…if you’re writing a path tracer for fun, does using Copilot not take the fun out of it? If it’s already been done a thousand times, it feels like the point has got to be to write your version, otherwise why not just copy-paste or pull in a library? If you use Copilot are you not giving up the joy of building it with your own hands? What’s left at that point?

This is really neat. I wish it had an alternate color mode, as when the creator chose the endpoints of their scale, they managed to pick the two colors that are least distinguishable for me and about 5% of people!

Fascinating. How did you train yourself to improve this?

From a Firesign Theatre album, 1969:

“That's okay, I brought an Erector Set.”

“Throw a towel over it!”

“Do some pushups, Pablo, Maybe it’ll go away.”

> This doesn't seem that human: https://www.theregister.com/2023/12/01/chatgpt_poetry_ai/

Agreed; that’s why I was very careful to say “one approach.” I suspect that technique exploits a feature of the LLM’s sampler that penalizes repetition. This simple rule is effective at stopping the model from going into linguistic loops, but appears to go wrong in the edge case where the only “correct” output is a loop.

There are certainly other approaches that work on an LLM that wouldn’t work on a human. Similar to how you might be able to get an autonomous car’s vision network to detect “stop sign” by showing it a field of what looks to us like random noise. This can be exploited for productive reasons too; I’ve seen LLM prompts that look like densely packed nonsense to me but have very helpful results.

Okay. I think you might be yelling at the wrong guy; the conclusion you seem to have drawn is not at all the assertion I was intending to make.

To me, "acting like a human" is quite distinct from being a human or being afforded the same rights as humans. I'm not anthropomorphizing LLMs so much as I'm observing that they've been built to predict anthropic output. So, if you want to elicit specific behavior from them, one approach would be to ask yourself how you'd elicit that behavior from a human, and try that.

For the record, my current thinking is that I also don't think ML model output should be copyrightable, unless the operator holds unambiguous rights to all the data used for training. And I think it's a bummer that every second article I click on from here seems to be headed with an ML-generated image.

The LLM is trained on human input and output and aligned to act like a human. So while there’s no individual human involved, you’re essentially trying to social engineer a composite of many humans…because if it would work on the humans it was trained on, it should work on the LLM.

Wow, that’s mind-blowing. Imagine if the folks back at Bell Labs could have known that was coming.

I love that. It’s funny, though – I know the transistor is “little” compared to the vacuum tube, but look at it next to an actual brain cell and it’s enormous!

I polled my kids. All of them down to the youngest (19) enjoyed Holy Grail a lot. My 24-year-old says his friends reference it in conversation regularly. We’re not totally out of touch yet!!

Wow, thank you for this. Today I learned why the slash-and-underscore ASCII art I came across looked so underwhelming on my PC, and that others were seeing it in all its glorious cohesion.

Thanks for sharing. Your uncle sounds a lot like my uncle, actually :)

It’s easy to miss, the 990 is a maze! I just happened to get lucky – my work had me doing a project on 990 data recently.

Out of their $443,020 expenses in 2021, the biggest line item expense was $187,262 for "CONSULTANTS." Their next biggest was $169,383 of merchant processing fees. It will be interesting to see their Form 990 for 2022, which is presumably the "last year" Ben Fry is talking about..