> 16GB RAM is also a practical capacity for general tasks such as web browsing, office work
16GB RAM to read and write text and see images... I think it just says it all.
More could be said about the bloatware, telemetry, surveillance, and adtech built into Windows.
Hard to imagine that 1GB was "a big deal" in my lifetime.
Only when the resources are limited do we see optimization being done seriously.
Sure I don't need 32gb to do all that, but who knows what fun things will come from having that extra budget?
For 90% of browsing like searching and reading articles or watching video or using email, all of those worked on the almost same functionality 10 years ago with a fraction of the memory.
An example I have is a food orders website, which takes 4 seconds and 10MB to load, where all I want to do is go unsubscribe for the weekly order (where I need to unsubscribe every week because it's opt-out, I guess it's better for them). This page is crap because it can, not because it brings some value it couldn't without using all that compute power.
Then as a software engineer, many colleagues literally say "memory is cheap" very often, which kind of hints towards the fact that they don't care about optimizing for memory :-).
I honestly cannot really think of something I do now that is really useful for me and that I was not already able to do 10 years ago (hint: I don't use a copilot to code).
First it was audio, then video, and RAW camera images are almost there with the prosumer digital cameras. And those are compressed files. Software that decompresses the whole thing into arrays of regular samples will need much more RAM...
They say this is happing due to abstraction and saving development costs but I'm not so sure anymore. Windows 11 infamously has large swathes of boilerplate copied around apps that many of their engineers don't even know what it does.
At this point we're far into abstracting abstractions and those abstractions bear their own complexity that may or may not be more complex than what we began with at a fraction of the resource cost...
And most of this stuff isn't even doing anything that special or demanding, which is the sad part.
They get paid to be productive from the point of view of the company paying them. And by slapping layers of stuff together until it barely works, apparently they make profit for their company.
They are not paid to write optimized (or "good" or "useful for society") code.
Which is sad, no question asked. But it makes perfect sense given the system we live in.
Which I believe is not taken into account in the performance evaluations. How could you blame an employee for consequences that will happen in 2 years, when the employee has already been promoted for their good short-term results?