Turun

The YouTube channel looking glass universe (highly recommended!) also has a video on how alphafold works.

Edit: https://m.youtube.com/watch?v=Or3iq4_9-wA

What about the month? That's gotta be worth something!

Unfortunately the average person prefers flowery language for some reason. So that's what OpenAI optimized for.

If you tell it to be precise and short it usually works fine.

You should report it to the Bundesnetzagentur, spam calls are illegal.

https://www.bundesnetzagentur.de/DE/Vportal/AnfragenBeschwerden/Beschwerde_Aerger/start.html

Server's down :(

The election system distorts the vote quite dramatically in the US, that's true.

However, depending on where you're from left has quite a different meaning from what Americans mean when they say left. E.g. the Democrats don't want to dismantle capitalism or expropriate real estate companies (the city of Berlin seriously looked into this option). Center left has a similar shift in meaning in other countries.

Yes, but at some point it doesn't matter. The AI is trained to replicate human writing. There will be a point where it becomes so good that the result is a perfect replica, where it is indistinguishable from human text. I.e. even a perfect detector will not be able to confidently declare it as AI written, not ever. Because there is no difference.

I bet AI detection is going to get a lot better over time.

I doubt it. ChatGPT 3.5 is good enough to rewrite small snippets of text with better phrasing, ChatGPT 4.0 can write a paragraph if given enough support. Good enough as in "the output is indistinguishable from what a human would have written.

Of course you can do even more with the currently available tools - and get found out.

There is a way to make AI generated text detectable: by slightly pushing the output towards a consistent pattern a detector can reliably judge long pieces of text as AI generated.
Imagine if the AI is biased towards consecutive words starting with consecutive letters of the alphabet (e.g. "a blue car" instead of "a navy vehicle".). Not strongly biased, but enough so that when there are 1000 words you can look at the probability of consecutive words starting with consecutive letters of the alphabet and get a clear result.

There are two problems though: this only works with proprietary systems and only with long texts.

I have compact view. I tap on a thumbnail to make the image "full screen". In short succession I tap on the screen once, then touch and drag, which zooms the image.

Since I have it set to dismiss/leave "full screen" via of images by swiping the image up or down I need the tap before dragging to zoom into the image.

It still works the same.

I don't understand the "that's no how PDFs work" criticism.

Removing data from the original file is the whole point of the exercise! Of course unique tokens can be hidden in plain sight in images, letter spacing, etc. If we want to make sure to remove that we need to degrade the quality of the PDF so that this information is lost in said lossy conversion.

So close and yet so far

Let's take the low life area and make it no life area