We've just had survey at work about genAI, a large part of the devs use Linux and LLMs, yet nobody could explain how genAI works.
I was very surprised.
msage
joined 2 years ago
You would be surprised.
An awful lot of people, including Linux users, are into genAI. And not many understand anything behind it.
That's the neat part.
Ever since 'journalism' has been reduced to reposting social media posts, there is very little reputation to be had.
The community notes appear many hours after the original has been posted, meaning that majority of interactions will be before any notes will be attached. And even the process of choosing the community notes is not transparent, so you can never be sure that the note does its job even after it'd added.
A lot of different comments, and I know this is a greentext, but I wonder what people here think about americans who have 3 jobs and can only barely pull through?
Such people have nothing more to lose, everybody constantly keeps forgetting about them, the homeless, sick people... the list goes on.
And I'm not saying acceleration is good, but for the lost people, nothing good is coming from any side. And the pressure keeps getting worse. Even from the democrats.
So what is the solution, achievable as the world is now?
The current drive behind AI is not progress, it's locking knowledge behind a paywall.
As soon as one company perfects their AI, it will draw everyone to use it, marketing it as 'time saver' so you don't have to do anything (including browsing the web, which is in decline even now). Just ask and you shall receive everything.
Once everyone gets hooked, and there won't be any competiton left, they will own the population. News, purchase recommendations, learning, everything we do to work on our congitive abilities will be sold through a single vendor.
Suddenly you own the minds of many people, who can't think for themselves, or search for knowledge on their own... and that's already happening.
And it's not the progress I was hoping to see in my lifetime.
Really depends on many factors. If you have everything in RAM, almost nothing matters.
If your dataset outgrows the capacity, various things start to matter, based on your workload. Random reads need to have good indices (also writes with unique columns), OLAPs benefit from work_mem, >100M rows will need good partitioning, OLTP may even need some custom solutions if you need to keep a long history, but not for every transaction.
But even with >B of rows, Postgres can handle it with relative ease, if you know what you're doing. Usually even on a hardware you would consider absolutely inadequate (last year I migrated our company DB from MySQL to Postgres, and with even more data and more complex workflows we downsized our RAM by more than half).
Postgres is so quick if you know how to use it...
Gentoo is the only way.
Me, for instance
I've seen people on the internet suggesting Darktable as a solid Lightroom replacement... I don't know anything about photo editing, but am curious - how bad is it?