Keep the programs in RAM so I can switch between them. Even my Pixel kills the game when I navigate away from it. It just thinks I have way too much shit open. So my phone has 12 GB of RAM and it’s not nearly enough to multitask

I have 67 tabs open in Firefox at this moment and maybe I don’t want them to reload when I come back to them from Lemmy

No it doesn’t, when have you ever taken LiberuxNexx two times daily? It sounds like some guy’s username on X

That’s a good one, but it’s only for Japanese.

Of course, you can’t easily extend that to other languages unless you have conjugation/declension tables. When I want to learn a word I need to be adding the base form to Anki, not the actual word said

You’re talking about learning with texts

It’s not great for languages like Korean where you might have a lot of different conjugations that will be detected as new words

Maybe it should be.

StarCraft would not be so hard to make. But nobody did that, even though 0 AD exists to clone age of empires 2

It even works as a 2D game so no modeling experience necessary

I run it in Firefox, though

Open source language learning only has Anki. Everything else is in an enbryonic stage.

There are so many low hanging fruits. Add-on to look up words in subtitles and add it to Anki. Luo dingo clone that’s a bit less tedious (without having to write so much of your native language). Clozemaster clone (unless someone knows how to set up Anki to do this)

So are fiat currencies

I’m a real adherent of disenshittificationism

Bezos doesn’t run the business anymore. He’s no longer the CEO. While he exploited the workers during his tenure, as a customer I had no complaints when he was CEO. They had only started to enshittify recently

Long before the anticipated sanctions, Liang acquired a substantial stockpile of Nvidia A100 chips, a type now banned from export to China.

That’s how

You can see with image generation progress was extremely quick

It’s irrelevant because it wasn’t a precursor technique. The precursor was machine learning research, not other image generation technology

Creating abstract art by moving pixels around is not anywhere close to what we mean by image generation. At no point did this other software generate something from a prompt

Image gen did not exist in any way shape or form before. Now we’re getting video gen like a few years later.

Let’s not forget we started by playing the game of Go better. My prediction as a hobby Go programmer (the game, not language) in 2015 would be that better than human AIs would be there by 2020 and they got there by 2016.

Before the AlphaGo match with Lee Sedol people predicted the AI would just put up a decent fight since a previous version played questionably against a weaker player. It blew one of the best players ever out of the water, losing only one game of the series.

Future matches even against the world #1 with the better models showed it to be invincible against humans

You’re making the same mistake. You’re looking at the current capabilities and predicting a human speed of improvement. AI is improving faster.

https://github.com/mozilla/TTS

Also tortoise tts and a few other options

Yeah, I’m just testing it out. For a true Duolingo experience it would need fill in the blank and audio

We’re talking about, say, learning Spanish and Duolingo be like “now translate this very long and overly specific sentence to English”

Then you end up trying to construct the English sentence even though you’re learning Spanish

Here’s an example where I think my sentence is perfectly fine, but it just expected a different word order. It expected me to put If at the beginning, but I didn’t notice it was capitalized.

Korean doesn’t even have capital letters, why is it doing some gotcha about English capitalization when I already know English?