I've run it as a hobby for 20 years over four careers. You might have seen my Hexayurts at Burning Man - they were the starting point of the project.
I knew that was the only way to keep it alive for the three decades I expected to take it to work.
Funnelling refugees into makeshift camps has been a horrific failure (see: https://en.m.wikipedia.org/wiki/Calais_Jungle) enabling this at scale is to me basically building district 9.
It would be awesome if there could be many of these, in smaller and smaller cells around the world as time went on, such that any disruption could be handled, to prevent an 1177 BC style civilizational collapse.
At the end of a tree's life, it is decomposed and turned into fertile soil that will feed future generations of the forest. This is a sustainable cycle.
I think we need to develop a plastic alternative, say X, that can be used to make products such that at the end of the products' life, the product can be converted back into X. Then, X can be reused to make new products creating a sustainable cycle.
Or maybe teaching tech to youths.
I love this stuff but the cost of setting it up is in the millions.
I'd love to read more.
The last 2 decades have seen a fall in large scale infrastructure projects as the will to “kick the can down the road” has only increased. This makes the environment for longer term building of things like nuclear plants, much less appealing.
The lack of will to do things on a medium term timeline is going to reduce how long our long term timeline will be.
e.g. laptops with matte displays and columnar layout keyboards. Think different.
And by novels I mean a "shitty first draft," not a publishable work.
I've made a start on it, but I get side tracked by having to make money.
Less flippantly, I'd probably spend my time writing free speech therapy apps for kids with autism. I might even become a speech therapist myself.
Naturally this costs quite a lot of start up money and doesn’t tend to generate much in the way of cash flow, so I think it will just be a dream, unless money were no object!
This coincidentally is basically the underpinnings of every startup plan I've ever seen, at least the good ones anyway.
Would never put any effort into anything meaningful again.
For example, I'd love to see humans write an outline of a book then feed it to AI who writes the book in full, using a consistent writing style.
Or say, you have a business idea. You can brainstorm a (domain) name with the AI, brainstorm a pitch, the pitch deck, the company colors, and so on.
We need to have the cost of pollution properly factored in to the price of things, to the taxes individuals pay, etc... so that they are offset.
I'm fairly certain it's currently impossible with Linux, Windows, etc. I'd like to get us back up to at least the safety we used to have with an IBM PC XT, 2 Floppy drives, running MS-DOS, with stacks of write protectable floppy disks
https://external-content.duckduckgo.com/iu/?u=https%3A%2F%2F...
I guess one could get a Raspberry Pi with camera, and 3D print such a box (or order from AliExpress). Install a QR code reader software that can talk to Spotify, and print some cards, ideally with QR codes on both sides.
The end user would just have to place the card on the glass to get music to play.
2. Some game in a Lisp
3. Working out regularly and getting in good shape, and having a rock solid sleep schedule
4. Drums
5. Learning the math for general relativity and general relativity
Maximizing human-computer bandwidth/throughput
Education
Mathematics
2. Tabletop roleplaying games
3. Various environmental and social causes
2. Figuring out reverse engineering of malware somehow
3. Data engineering for some start-ups
Nope, not as a hobby but competitively against MacOS and Windows.
There are a ton of great challenges trying to compete and differentiators (by features) alone are probably not enough but key aspects I am interested in tackling are…
1. A hybrid cloud OS
2. Remote access
3. Remote data processing
There are some other ideas but I think there’s a lot of innovation that can still be done at the OS level
The use of capabilities offers a way to handle authentication without a lot of grief for accessing remote resources... the capability IS permission, so there's no need for usernames/passwords, etc.
Imagine you give a task a list of URLs (to make it simple) at startup.. that list is the ONLY set of resources it is allowed to access. No matter how confused or rogue the task gets, it can't corrupt anything else, or store trojans anywhere.