I tend to not really care for most new things, as most of it feels cheap, inauthentic or a scam to further the surviellance facist oligarchy state. Id be completely content with time frozen in 2004.
So, to be a little more positive, what are some new things that are actually good?
Note, to me, new is within the last 10 years.
I’ll start. The fediverse concept is neat.
Webb space telescope
Fully homomorphic encryption doing useful things
Higgs boson detected (oops, 2013)
Solar power and battery storage cheap enough to displace fossil energy and let ordinary people go off grid
New tacqueria in my neighboorhood is actually pretty good.
What more could you want?
And I’m only finding out about this NOW?!
Like what?
Private information retrieval with just one server! Nobody thought that was possible.
https://www.cs.cmu.edu/~csd-phd-blog/2024/piano-private-information-retrieval/
Shit like this is what makes me realize I’m an engineer and not a computer scientist. I get the logic, but not so much the practical math.
It seems like there’s a lot of unknowns for where PIR is best used, though. Like using DNS as their example has a lot of issues (that they mention). In what cases is it really best to download the whole dataset (temporarily) to just then go back and query the server on that dataset?
Though maybe I’m just not that creative of a thinker.
Single server PIR is kind of a theoretical surprise but good ways to do it with multiple servers have been known for a long time. See the Wikipedia PIR article. Yes it’s maybe a solution looking for a problem, but it’s way cool that it can be done at all.