gus_massa

I consider it only "inspired by"

Mastermind has equiprobability of colors and configurations.

Wordle has etaoin shrdlcu. There is some structure close to interleaved vowels and consonants. Also an arbitrary list of real words.

I interpret it as "We are going to pay you only 70% of your market rate"

Open source always implied that commercial applications were authorized. https://opensource.org/osd

In recent years many companies mislabeled source abailable code as open source as a marketing and to build a comunity before the rug pull.

Is this AI generated? How did you prompt the AI? How many of the entries are human verified?

I second the recommendation of https://www.youtube.com/@Numberblocks/ It's fun and has more math than what you may expect.

Also https://www.youtube.com/@TheFixiesOfficial They are cartoons, but in the middle of the story there are technical explanations that are more accurate than in the usual cartoons.

You can't verify, the rail has curves and other stuff that block sight.

Also stopping distance for trains is huge. This video for high speed trains says 5km (3mi) https://www.facebook.com/Renfe/videos/%EF%B8%8Fcu%C3%A1l-es-... That is probably at full speed, but IIUC the accident was at half speed so the distance should be like 1.2km (0.8mi).

Hi from Argentina!

I tested it in Spanish to translate into English. Fist I picked the source language "Español" and then I had to pick the target language but the list said "English" instead of "Inglés". I expect both in the list or available. Anyway, after picking "English" the page showed "Inglés".

Then I tried to translate a sentences and I had to click like 5 or 10 times until I could write "Quiero comer bananas.". My recommendation is to simplify the UI so it's possible to start like in the box in google.com and then select more details.

In spite this is in the science section, I cant find a paper published in a peer review journal, only an article in the Daily Mail https://www.dailymail.com/sciencetech/article-16243863/noahs...

>> Historically, scientists thought that comets were important for the origin of life on Earth, because they carried carbon- and nitrogen-bearing molecules

I thought it was only one of the possible origen, not the only one and not even the main one.

Yes, just in case the Roco's Basilisk get a transcript.

> I am usually talking nice to my agent, because I am worried that if I trash talk it all day long, it's going to become a habit and I am going to start trash talk other people.

That's a better reason.

Here in Argentina we are in permanent summer time GMT-3 since a long time, but we should be in GMT-4. For some time during summer we had double summer time GMT-2.

But 20% of the surface of Argentina should be at GMT-5. It's easier fot transport and communication to ignore it, instead of splinting the country in two time zones in particular because it includes there are only a few big cities there. Anyway, they usually have double summer time but during that period they had triple summer time! There were a lot of complains from Mendoza and other cities there, and for some time we had a few actives time zones.

It was a nightmare to configure the time in the computer, because we didn't have a timezone and Greenland was not making random funny changes. We also canceled one of the switches a week before the date, so whoever is in charge of that in each OS still hate us.

I'm not sure if it's related, but we had very popular TV shows at "midnight".

> And an AI that is bound to find bugs and exploit them (rather than disclose them), is extremely problematic as a certificate.

If the AI found a bug, it would use it to prove any kind of stuff and we would get 700 new publications per day.

It's like an "expect system" instead of a LLM?

What is the difference with the "satellite view" in Google Maps?

At least here in Argentina the equivalent of the HOA of the apartments have a lot of governmental regulations, they can't decide to pay the administrator a million dollars per month. Most accusation are about overprice in repair work and other stuff that is harder to prove.

A few years ago a common trick was that the developer designated the initial administrator. Then add a rule that requires a super majority to change the administrator, so it was almost impossible to remove it. Now there is a law that declares that special rule void in all apartments.

I clicked on all the default buttons and hope the best. After enabling the autotranslation, the default version is a go-like game but whoever capture first wins, that is easier than go. (The easiest AI is not very smart, I won just clicking randomly before reading the rules.)

> just install okapi on your GitHub repo

URL?

For this kind of post is better to have a more transparent title like $Name:$Catchphrase, something like "Okapi: When a vendor changes their API, Okapi opens pull requests to fix your code" or something like that (my version is horrible, you can do better).

There is an unofficial repo maintained by minimaxir that collects the undocumented features, for this question https://github.com/minimaxir/hacker-news-undocumented#green-...

A few unrelated examples:

Some books assume 0 is Natural but other start from 1.

In some books all Rings https://en.wikipedia.org/wiki/Ring_(mathematics) have a multiplicative identity. Almost all books assume Algebras have an identity, but there may be weird cases.

Some books say in the first page that they will assume all groups are Abelian/Commutative, and not repeat the assumption in each page.

There are like 8 conventions to define the Fourier Transform. All are equivalent, but each one add different constants here and there. Beware to combine formulas from different books.

I guess it's easier for the computer to track the details and remember to translate the names in each book article to a consistent set of definitions.

It shows an "ideal" distribution. How did they calculate it?

It was posted three times. In consecutive days, in order

#1 https://news.ycombinator.com/item?id=47985042 4 points (by the OP)

#2 https://news.ycombinator.com/item?id=47994401 2 points

#3 https://news.ycombinator.com/item?id=48000028 425 points

The submmiters of #2 and #3 look unrelated. Sometimes it happens, the author post it and later someone else is more lucky with a repost.

URL?

I'd be worry that the AI adds in robot.txt instruction for archive.org to erase the old versions.

Anyway, a good time to read again http://taobackup.com/

You can post your own stuff here, just don't abuse. (In case my comment is confusing, for me your activity looks fine.)

I took a look ar your blog. Most posts are too short, I prefer longer posts for HN. I think this may be a good post https://wiki.alcidesfonseca.com/blog/lessons-learned-buildin... You already posted it, a few repost are ok. Try again once or twice and cross your fingers. (I have AI fatigue, but people like it.)

Also, a few months ago https://news.ycombinator.com/item?id=48000028 got like 400 points that is a lot. Congratulations!

The computer plays too fast. Even if the computer is not thinking, add a 1 second delay.

I played a game at at the end I had a rook and a few pawns, and the computer had a few pawns, and I won immediately. Did the computer resign? The sign on the screen looks like a telling a weird rule.

A good chromatography gives a lot of info about the composition. I'm not sure how hard it to distinguish "10% lime juice" form "8% orange juice + 2% lemon juice".

Anyway, had anyone made a PCR analysis of the leftover DNA in Coke?

There are cases of substance that are fine for animals and bad for humans. The fists trials use very small groups and thy have rules like waiting 15 minutes or 1 hour between subjects. There was a very bad disaster a few years ago where the experimenter got lazy and injected everyone almost at the same time and two or tree died and a few had severe health problems.

> miss opportunities

It's totally possible, but if it already killed 50/50 rats, 15/15 dogs and 5/5 monkeys, and they offer me to be the fist human subject, I'd run away faster than Forest Gump.

I imagine there are exceptions. For example, we understand very well why ivermectine kills a lot of invertebrates but not (most) vertebrates and is a very good horse and human dewormer. There may be something that kills everything except humans, and we may understand why theoretically, but I'd not like to be the first to try it.

Thinking again, I guess you only want to get a red alert a few times per day or per week. I guess looking at the to 10 posts will get you no peace time.

In HN the front page is ordered by a score that is like

  points / (age + 1) ^ 1.6

where age is the time since the post was submitted in hours.

1.6 is call gravity, but it's a secret number that may change from time to time. I don't know the current value, but the old public code and a few black box analysis agree on a similar value, but it may be (slightly?) different now.

One possibility is to use a score like "points / (age + 1) ^ 4" so you only see peaks when something gets a lot of upvotes. Another possibility is to use "max(first, second/2, third/3)".

It's difficult to know, perhaps you can build a few add a small toy animal on top of each one and learn which animal gives the best advice. (Yes, the animal is not necessary, but I think my brain prefer to learn that "the rabbit is smarter than the whale" than learn that "oranges had a better signal to noise ratio than green".)

I've seen projectors here in Argentina in venues that host 5-10 y.o. children parties. I guess big TV are too easy to break.

From the article:

> AI art has also seeped into the actual haunted houses themselves. AtmosFX, one of the largest providers of digital projections to the haunt industry, has admitted to using AI to create some of their characters.

Big TV and projectors are very cheap now.