Many moons ago, I was making mod for a game and had the idea to publish it on Nexus Mods [0] so that I didn't have to bother setting anything up once I actually wanted to publish the initial release of the mod. It was not at all in a working state when I made the public page for it.

Imagine my surprise when I wake up the next day and have thousands of views on the page and a dozen comments berating me for publishing a mod that doesn't work...

Ever since then, I have had problems with working with the garage door up, even though I know that it's totally acceptable on GitHub. It's habit by now to work on everything with the garage door down, just in case...

[0]: https://www.nexusmods.com/

[0]: https://grayjay.app/

The generated palette seem to be a great indicator of whether the model understood the prompt or not.

I Haven't checked out the Python SDK yet, but it seems very interesting!

I'm curious to know if there is any reason for why you picked Gemma 1B for the Expressive model. Did it generate more cohesive parameters than other 1B models? Or was it just the first one you picked?

>But what if none of that accepted wisdom is true?

Is that really generally accepted "wisdom" though? I thought it was just a sliver of young men that had those ideas. I wouldn't consider that as an accepted fact (or "wisdom", as they call it).

You don't even have enough karma to downvote comments though..?

Even for propaganda, I am constantly made aware of my propaganda immunity being subpar for all different kinds of propaganda. Often it's just subtle seeds of propaganda that impact the choice of words that I use to be something different than what I really believe in, and sometimes it is more serious and deeper cases of propagandisation. Very unfortunate, but each time it shows me why I should be critical of everything that I read online.

I'm assuming this is only works for the web, and not for mobile apps?

What's the support for Material 3 like?

I'll definitely try it out later when I am on desktop!

Btw, the Hero Description (the text that among others says "Perfect SEO") overflows on my phone. Might want to get that fixed. Oh, and the "View on GitHub" button doesn't actually seem clickable?

Even if a pupil does pay, they will either be too lazy to wait the nearly 10 minutes it takes for the AI to do its research, or they actually care about getting good grades and therefore won't outsource their research to AI.

I would wager that they're still going to maintain their own version of the World Factbook, and just simply not share it. This would allow them to cut out the very costly review step that I talked about.

Now whether that's a good decision or not is a completely different question.

Now, I have yet to see any cases of the CIA doing this to the World Factbook, since that would tank its credibility, but I also don't browse the Factbook too often.

Since Wikipedia articles in local languages tend to also cite sources in the local language, manually written articles virtually always contain exactly the correct technical term for different subjects.

For instance, If I want to figure out what a "Universal joint" is in Swedish, I just find the Wikipedia page for universal joint and hit the translate button, and voilà, I see that a U-joint is called "Kardanknut" in Swedish.

If I had instead tried to translate with Google Translate, I get "Universalkoppling", which (from what I can see) is a type of screw.

With AI, I worry that entire articles will be directly translated without anyone first making sure that the articles are using the correct local terminology.