I don't get why they test it with half life 2. Almost all those examples look better with what they're demonstrating turned off, and it already had really good lighting for the most part, so the difference that RTX adds is very subtle.
They should focus on building support for more games that had simpler real-time lighting instead of leaving of leaving it to the community. I tried using thier tools with Far Cry 1 and it was far too complicated and janky to get working.
I picked up some firearms a few years back from a lady that wanted to get rid of her deceased husband's collection fast and for cash. One of them was not only in horrible shape, unsafe to fire, but it was illegal. Broke it down, recycled what could be, sold the few parts that were usable, then trashed the rest.
How'd her husband die? Gunshot wound?... 😅
Sounds like you got rid of the murder weapon for her 😅 (j/k)
You don't have to explain the process to observe the difference. But it could be useful to understand the process to help figure out ways to mitigate it. I think we'll have to wait for future research to find that out though.
God knows why they felt the need to scan old catalogs
For the AI training data probably 😅
Also they started doing it way back in 2005, when they were just doing whatever thier engineers thought would be useful and ensure they kept market share of search. (before Prabhakar Raghavan took over)
Maybe it's colloquially short for light-parsecs 😅 Like how people say "gigs" but might be meaning either gigabytes or gigabytes/second or a concert and you work out which from the context.
It's usually used in explaining that a rich person can be sad, rather than to say all poor people are always happy.