hulium

My last sentence. Or which parameter do you think was modified?

The problem is that the parameter was unspecified in the original problem. Your answer is equivalent to the shaky answer referred to in the paper. Without knowing that the host selects between 2 goats with 50/50 chance, you cannot give a general answer.

Last time I claimed this was the correct answer, I was linked to the Monty Crawl problem. https://www.probability.ca/jeff/writing/montyfall.pdf

Isn't it a bit unnecessary to add blur when the website is almost completely black? Maybe put some screenshots on the page.

Were there any interesting non-neural approaches? I was wondering whether there is any underlying structure in the ARC tasks that could tell us something about algorithms for "reasoning" problems in general.

Unexplored comes to my mind, it differs from other games with procedural generation in that it generates a graph for the gameplay first and builds the levels around it. It's not necessarily ground breaking, but it has a special feel to it as objects are placed with purpose.

Computer Science becomes MORE interesting as computers become more capable, not less. There are so many things we could be working on, but we still waste so much time on boring libraries, configuration, implementation details that we simply don't get to experiment enough.

Just like nobody programs on punch cards anymore, learning details of a specific technology without deeper understanding will become obsolete. But general knowledge about computer science will become more valuable.

I think the cool part about this is that you can easily build a user interface from physical objects, so you can skip some of the digital frontend and design and focus more on the computational backend. This frees you up for more explorative, improvised coding, which is great for research and art but maybe also many other situations where fixed user interfaces get into the way too much. Based on what they show about their biology lab project, it seems to be surprisingly useful.

Yes, you are right. In this game, you can know the answer after 6 guesses, but then you also have to tell him, which counts as the 7th guess.

> Given that 7 guesses covers 128 numbers

I might be confused, but don't 7 guesses actually cover 255 numbers? I think you have to count all nodes in the search tree, not only the leafs, because you can get the correct number before reaching a leaf node.

Or more generally k guesses cover 2^(k+1)-1 numbers, e.g. with one guess you get the answers correct/high/low, which can cover 3 numbers)

Maybe there is a mistake in my thinking, because this would mean you can cover 127 numbers with 6 guesses so you could not lose the original game.

Edit: My mistake is that you still have to explicitly guess even if you know the precise answer already, so you cannot cover 3 numbers with 1 guess. This means 7 guesses cover 127 numbers.

I don't think the website where you download the PDF from matters much, the journal claims to be peer reviewed https://www.iospress.com/catalog/journals/algorithmic-financ....

> Note that the "journal" editor is the author of the paper.

Interesting, I didn't notice that. The journal does have other editors too, see the first page of the PDF.

Related is the theory of the "Lemon Market": https://en.wikipedia.org/wiki/The_Market_for_Lemons

If the buyer cannot estimate the quality of the product before buying, the sellers will reduce quality or be forced to by competition. Buyers lose trust and the market can collapse.

The link leads to a journal published paper, why do you think it not peer reviewed?

awk is also not hard to understand, scroll through the Wikipedia page for a few minutes https://en.wikipedia.org/wiki/AWK#Structure_of_AWK_programs

It runs an action for each line in the input (optionally filtered by regex). You get automatic variables $1,$2... for the words in the line split by spaces.

The syntax is almost like a simple subset of Javascript. Builtin functions are similar to C standard library.

If you have input in text that is separated in columns with a delimiter, and you want do simple operations on that (filter, map, aggregate), it can be done quickly with awk.

That's all you need to know about awk.

It is not fully clear from the article, but it seems like their new model is enabled only for Japanese-English, Chinese-English, German-English and only on paid DeepL Pro at the moment.

If you set the screen brightness and background color right, you can make a screen look like paper or maybe an e-ink display. Unfortunately, the illusion can break within minutes if clouds are moving in the sky. Also the screen doesn't reflect like paper if viewed at different angles. Nevertheless, I wonder how well it would work to do such a calibration with a sensor.

That's the idea, but what makes it a bit more complicated is that a string can contain a prefix of itself, like "mama" in the article, or in the worst case "aaaaaaa".

There is also the open source FinGPT, that is claimed to beat GPT4 in some benchmarks at a fine tuning cost of $17.25.

https://github.com/AI4Finance-Foundation/FinGPT

If the the door contains a car, then it doesn't matter whether you switch or not. If it is a goat you get the problem as intended.

You still have the intended problem as a subproblem, so this is just a distraction and not a serious objection.

> opens another door, say #3, which has a goat

There is no randomness in this sentence. The door has a goat, not a car.

Then again, you might read it as an example outcome. But even if the door was chosen randomly, it would not justify the 50%/50% answer.

Independently of that, assuming that the host chooses the door with the car and the goat with 50% probability each is the same mistake that confused so many, you cannot always assume uniform probability only because there are 2 options.

(I wanted to point out the flaws in your thinking because this has also been a confusing problem for me)

I did this in public networks and it did work, but I am not an expert in how they are set up. When do you think it would not work?

Thank you, I think I didn't get the idea across properly.

None of that is necessary because you are not exposing anything to the internet. We are still talking about local sharing. Android does not have a firewall so there really is nothing to configure. But if you think it cannot be possible, try it for yourself.

An app like "Share via HTTP" does that for you.

https://play.google.com/store/apps/details?id=com.MarcosDiez...

Which of the steps that I described required advanced technical knowledge?

Edit: Ah maybe you thought setting up the server is an additional step. It is not. Share via HTTP is the server. It runs directly on your phone. The URL points directly to your device.

I am not sure you understand. You click share, then Share via HTTP, the other person scans the QR code, they have the file. Setting up the server is no effort at all.

Only if you are outside without Wifi you have to do another step, setting up a Wifi hotspot, but that is not hard. And in most situations I have been in both devices were in the same network anyways.

You can create a hotspot and start a HTTP server. Scan a QR code to connect to the hotspot, then a QR code for the file URL. It's easy and works with any device with Wifi access.

For the server, on Android I use "Share via HTTP". On desktop running a lightweight HTTP server is easy too.

Edit:

Apparently that sounded complicated, here is how I can share a file from my phone to any other phone in the same Wifi.

- Click Share -> Share with HTTP.

- The app opens and shows a QR code.

- Other person scans the code. Their browser opens the file and they can do whatever they want with it.

If you have no Wifi:

- Open QR code for hotspot

- Other person scans QR code for hotspot, then for file.

If you still think it is too complicated, surely there is an opportunity create a simpler UI! From a technology perspective it does not have to be complicated.

The whole thing is quite confusing at first for such a simple tool. The splash screen says "Create amazing mockups" and "Craft beautiful presentations", somehow that translates to "Frames and backgrounds for marketing images".

> the "app gap" was one of the main problems. High-profile apps like YouTube never became available on the platform

There were great native third-party YouTube clients in the official Store on Windows Phone. On the contrary, I missed my Youtube client after I had to switch back to Android (until I found NewPipe).

The linked article is not about AlphaFold itself, but about researchers verifying AlphaFold's predictions. So the article directly addresses your question. Unfortunately HN only shows the first half of the headline.

> Shoichet agrees that AlphaFold predictions are not universally useful. “There were a lot of models that we didn’t even try because we thought they were so bad,” he says. But he estimates that in about one-third of cases, an AlphaFold structure could jump-start a project. “Compared to actually going out and getting a new structure, you could advance the project by a couple of years and that’s huge,” he says.