Do you have a resource for us to do this for ourselves? Seems like a pretty good thing to do

(from my notes) Disabling swap: - Disable SIP (`csrutil disable` from recovery environment terminal) - `sudo nvram boot-args="vm_compressor=2"` in macos terminal - reboot, confirm with `sysctl -a vm.compressor_mode`

`browser.cache.disk.enable` = false in Firefox `about:config` (I think there's more to this, but I can't recall, more research required)

I disabled spotlight altogether among some other services, for some reason I didn't finish my notes. It's a start though!.

+1. When there were still spinning HDDs, some of this fuckery was obvious, like spotlight indexing all the time like crazy. This might still happen, or happen again, but we sheeple would'nt know because SDDs are much quieter than HDDs.

Have you used this? Looks fairly interesting.

I've been using Rewind for a while and find quite a bit of value for it. Because it uses GPT-4 for Ask Rewind it's not local-only for that feature, but I find myself just using the local search most of the time and it works fairly well (and have it firewalled off, so I know they're at least not lying about the data staying local except for updates, analytics [if you opt in], and Ask Rewind usage).

To get even less "structured" backgrounds you should try replacing the support backgrounds and the far backgrounds with a light bit of depth noise.

For people interested in this def look into autojump before doing this

We saw this happen with Stable Diffusion and it's not surprising we see this happening here. There is a lot of interest in taking these models that are in striking distance (single order of magnitude) from running inference and training on consumer level hardware and as such a lot of energy is going into making the optimizations that can get us there.

Generally speaking, research is not usually done with consumer usage in mind, so what this is, and Dreambooth etc. for Stable Diffusion was, is that gap between researcher software and accessible software being bridged.

TouchDesigner!

Also Natron!

https://natrongithub.github.io/

https://derivative.ca/UserGuide/TouchDesigner

Also Notch!

https://www.notch.one/

So many parts across the stack need to work well for this to go well. Early support for popular software is a good example. This goes from partnerships all the way down to hardware designers.

I'd argue it's not about engineering more than it is about good organizational structure.

And having execs who design the organizational structure around those goals is part of what makes good engineering :)

Pinetime can't hold a candle to the bangle.js: https://banglejs.com/

The bangle.js: - ships significantly faster - has an always-on display with similar 4 week maximum battery life - can be updated without flashing - has a thriving app ecosystem

I have a bangle.js. It's... alright. Does the job. Feels cheap. OS is nightmarishly slow. Only one physical button, so annoying and fiddly to set with a touchscreen (interacts badly with the slow OS). Has GPS, but doesn't really work. Has heart monitor, but doesn't really work. Not terribly stable.

My Pebble Time Steel was vastly superior.

I wonder: could one possibly run a custom ROM like bangleJS on a Garmin?

Obviously government subsidy..

Subsidies.... to the landlords who own the buildings? Or are you saying this will be owned and operated by the government?

Meaning taxes

Or, good old QE.

Who wants to live in a city if you can work from home? If you are poor and need housing subsidies and there are no jobs for you, there is no reason to live in the city to begin with. People go live where the jobs are. Without "white collar" jobs producing demand for services, there is much less demand for "blue collar" jobs.

If it's just cooping people up, there are cheaper ways to house people than maintaining expensive buildings that need financing and upkeep.