That's not a map of Twitter users, be careful when reposting things without checking! www.reddit.com/r/WhitePeopl...

Why do you think social media only contains people? Why are you not considering the effects of anonymity and the lack of restraint when you're in a different country with potentially different laws in place? Why don't you address the actual social media harms mentioned in this research?

Subtitles are great, better chance of people watching if it works with sound off as well 👍

That word-by-word highlighting is just distracting and annoying though: feels like the video equivalent of a <blink> tag!

"Convenient" is perhaps an exaggeration: enforced key quoting, no comments, no trailing commas, poor numerical support... if you're after _convenience_, formats like YAML seem a better fit? JSON is great for software, less so for people.

I think that was a different company - en.wikipedia.org/wiki/Monolit... - but they both rock...

Bubble sort is _almost_ never the right choice, so LLM is leading you astray unless you really wanted that for some reason... there are plenty of better algorithms from en.wikipedia.org/wiki/Sorting...

If you're claiming that "I don't know" is lying, then perhaps the misunderstanding lies with you?

If the pedals are also confidently wrong about what colour the milk is, then sure - some level of concern seems justified. Even a pedal-powered LLM can be trained to say "I don't know", though.

Are those PHDs confidently stating a number which isn't based on any actual reasoning or facts, or would they explain _why_ they can't answer and what information they'd need in order to do so? And if the former... which field are they in?

Worthy cause, but not a great survey - for a start, no contact information or details on the page 😟 Example problems include the questions like "The maximum fine the Electoral Commission can levy" (a 50k+ fine is going to make a difference on a 500m donation??), the lack of any links for context.