getting ready to give a presentation about my bit of IBM : https://www.ibm.com/client-engineering

I'll be describing some of the projects we've worked on with customers this year. It's a technical audience, so I've even drawn diagrams to help explain some of the fun things we've built 🤯

Mastodon will update to 4.0 soon! Here are the most exciting new features (in addition to fixes & changes):

- ability to follow hashtags
- ability to filter by language
- ability to translate posts
- ability to post HEIC files (i.e. your iPhone’s native photo format)
- define user roles (big deal for admins)
- define sign-up process on your Instance (admins again)

See all changes here: https://github.com/mastodon/mastodon/releases/tag/v4.0.0rc1

Thanks, @Gargron and @Mastodon Team. #Mastodon4

"If you allow the most intolerant voices to be as loud as they want to, you’re going to shut down voices of different opinions as well. So allowing free speech by just allowing all speech is not actually leading to free speech, it just leads to a cesspit of hate.

I think that is a very uniquely American idea of creating this marketplace of ideas where you can say anything you want completely without limits."

~#Mastodon founder Eugen Rochko

https://time.com/6229230/mastodon-eugen-rochko-interview/?utm_source=pocket_mylist

Btw: Mastodon lets you use two-factor authentication (e.g. with Google Authenticator, Yubico Authenticator, ...)

I'd recommend making use of it to protect your account.

#mastodon #security

Braess's paradox is one of the counter-intuitive consequences of Nash's equilibrium.

When a system is moved from its state of equilibrium, the new state of equilibrium may be worse than the previous state, even if the intention was to improve the cost function for everyone.

Take the case where 4000 cars have to go every day from START to END. There are two routes available, one that goes through A and one that goes through B.

The START-A leg takes t=a/100 minutes, where a is the number of cars going through A, and A-END takes a constant time of t=45 minutes.

Similarly, START-B takes t=45 minutes, and B-END takes t=b/100 (b being the number of cars going through B).

Since the total duration of the two routes is the same, in a state of equilibrium half the cars would go through A, and half the cars would go through B. If we have a total of 4000 cars, both the START-A-END and START-B-END routes will therefore take 2000/100 + 45 = 65 minutes.

Suppose that we decide to break this equilibrium, and make commuting faster by building a new fast route between A and B. Suppose that it's a super-short but super-wide highway and the time required to cross it is practically zero. How will the new equilibrium state look like?

All the drivers may now pick the START-A leg, because it takes 4000/100=40 minutes, compared to START-B which is guaranteed to take 45 minutes. Then they would commute from A to B (which takes approximately zero), and then B-END, which takes 4000/100=40 minutes, compared to A-END which takes always 45 minutes.

So, in this new state of equilibrium where all the cars choose the new optimal route, compared to the previous 50/50 split, the total time will be 40+40=80 minutes. 23% worse than the initial state, even though the initial idea was to speed up commuting times for everyone.

https://resources.mpi-inf.mpg.de/departments/d1/teaching/ws12/ct/Braess-paradox.pdf