Interesting problems this could solve: over-investment and under-use of transportation (lease your bike to a courier from 9, when you arrive at work, until 5, when you leave -- charging a different leasing rate depending on the reputation of the courier, of course); estimating the effects of changes in excise taxes; having a better idea of when it's worth it to spend X minutes for a Y% chance to save $Z; etc.
I'd focus on electronica or something without vocals initially. Each "subreddit" would be focused around the evolved algorithms for a different theme (trance, house, etc.)
Maybe there's something like this out there, but I'd play with it anyway...
By the way, how much of it is evolved versus designed? Do you still seed new patterns? Are there any patterns that don't seem to evolve and have to be designed?
The site's Vying Games (vying.org) if you want to check it out. : ))
I develop a Go web app myself in my spare time: http://eidogo.com/. It's more for studying and bot-playing than multi-player, though.
I plan to add games like Go and Chess eventually. I've tried to focus on mostly lesser known games so far. The Go and Chess communities are very well served, and I'm not sure I could add much for those communities.
Some of my site is open source, btw. The ruby libraries that implement the game rules and bots is available at:
The server itself is closed source, but I plan on adding a well defined http api for outside apps to play there this coming week.
Bittorrent for the masses.
Freenet not only guarantees load balancing, but more importantly untracability of file storage and originator.
1. Technological improvements for human vision. I have a very special friend who lost vision in one eye because of a detached retina and the failure of ophthalmology to repair it and rewire it to the optic nerve. I want to help her, and other people in the same predicament.
2. Energy. Clean, renewable, highly efficient energy is the most important problem humans have to solve today (IMO). The usual suspects of solar, wind, waves, and "clean" biofuels only make economic sense because of government subsidies, and that situation stinks. We need radically new ideas.
It takes a good amount of capital and respect to start a school and have people trust you to apply an uncommon (though potentially very beneficial) approach to schooling on their kids, so in the mean time I'm trying to earn that respect/capital.
I would want to work on life/work/web integration. Like cross referencing gallons of gas in your car, with the number of gallons expected to burn on the way home, and anticipate the next time you will need a fill up, while checking your bank account balance, and having a web interface to do ALL of this.
Making things like a webcam you pin on to your kid that is also a gps tracker that will let you see what your kid see's at pre-school.
Stuff like that.
Note: I would have to learn a lot of stuff to make these things happen.
- no ads, period
- all personal info is held under the watchful eye of a seperate, non-profit org that ensures no personal info is ever shared with third parties
- you can delete your own account whenever you want, and you won't get hassled. An actual deletion of your data, too
- all source code is open source, and people can contribute whatever features they want
- It'd be nice to be able to plug my favorite artist/author/etc. So I'd vote for whatever ads the profile owner wants to put on his/her profile.
- Why not have each person's data held under the watchful eye of the hosting provider they're already with?
- Last I checked you can delete your 1and1/dreamhost/whatever account any time you want by no longer paying them.
- And it could be a GPLed project.
The issue is that facebook is great because so many people are on it. So for the dream project to work, it has to be really easy to set up. I am guessing that the average user does not want to grab a dreamhost account, install software, etc.
$USERNAME @ brevity dot org.
Then I thought of having an opaque jar, labeled "Swearing Jar." When you put money in, it spits it out and says "fuck off," or some other randomly chosen insult.
I also wouldn't exclude anything and would allow articles of any length and detail.
I imagine the end result would be something very much like wikipedia as the community would vote wikipedia-like articles to the top.
But vandalism would be impossible and every encyclopedia style article could link into a much longer article on the same subject. Known authors could sign their articles. So Linus could write something about Git, but the community could still vote some git's article higher then Linus.
Alternately, one could sign a modification of an article rather than the article itself, allowing for one of the greatest benefits of Wikipedia: the three thousand pedants correcting everyone else's spelling, grammar, punctuation and word usage.
I'd rather make an impact in space exploration, but I'm not sure how an individual would do that, yet. But I'm watching Cringely.
I don't think I could be happy working for a games company (I have a few friends who do, and they confirm what I fear - it sucks all the joy out of it, especially when you have to start from the bottom), and if I worked on it on my own it would likely be too ambitious for me to finish in any reasonable amount of time, so I would never expect to pay the bills that way.
It also wouldn't be "things people want", but just "things I want".
What about folders that represent tags? I could tag a file (say a picture) with Danny (my name), Vacation, and 2007. The system creates 4 "folders" as we understand them, each one labeled with a tag, and the final one labeled as "images" or something similar (whatever the file type is). You could open the "images" folder and see the image, along with all other images. You could drill down further into images>vacation and see all files tagged with both images and vacation. You will also see additional folders for other items that have additional tags and fit in the images>vacation tag.
You could also get to the same file by going into the "Danny" folder, to see all files tagged with "Danny,"
It's a little difficult to explain, and I'm sure that what I'm trying to express is already being developed by someone, but either way, it is a more efficient system than what we already have. No more annoying shortcuts and restrictive folders.
One can imagine it's Time Machine'sque metaphor being extended to archive your online life (twitter, Facebook) as well (assuming the data could be retrieved)
The only thing I haven't been able to figure out is how to represend these tags as directories. You shouldn't see them all at the top level, that'd be pretty crowded.
I'm sure Windows has something similar, but on MacOS X.5 (a feature borrowed from BeOS, IIRC), you can create a Smart Folder that's basically a canned query for files matching a very large set of criteria. I'm not sure what the API is for extending the criteria, but presumably adding tags wouldn't be all that difficult.
You are in Foocorp documents. East is Bloggins&Smith, south is Vexcorp.
You can see Foocorp.xls
Bob is here. Jane is here.
Bob says "Jim, can we go over the numbers for Foo?"
> Use Foocorp.xls with Excel
seriously. i think i would own a bakery.
It looks like he's up to 4 bakeries (soon 5), a cafe, a book, and a TV show.
The candy store would be really cool, IMO. Has anyone ever read Chitty Chitty Bang Bang? Of course you've read Charlie and the Chocolate Factory. If I ever did a startup, the whole point would be to earn enough money to open something like that, or maybe a toy company. I'd want enough money that I wouldn't have to bow to market forces and could make candy and toys that are actually good (Lego RIP).
Here's a guy who's doing good things with pizza: http://www.billyreisinger.com/pizza.php
Some inspiration: http://www.billyreisinger.com/pizza_log.html
Poetry and monetize don't go in the same sentence.
The Nintendo orchestra simulator is a taste, but my dreams were/are far bolder.
Something nice - there is huge crossover in the latin used through much of the choral repertoire. Thus, it's a nicely defined group of samples you would have to record for each singer - would probably only take a couple of hours per singer to get to a point where their voice would slot in and be in effect generic. Perhaps you could break down constonants and 'cambridge vowells' and come up with an even simpler and more flexible sample set.
If money were no object I'd also buy a second-hand stone church, build a house on the back, take up serious lessons and se if I could find a way to get an apprenciceship as an organ builder. Maybe get into choral conducting and try and bring music to people who have talent but haven't had the sort of childhood full of exposure. I'd read a lot of economics. I'd like to think I'd get more exercise.
Oh - I'd also like to work on an interaction engine at the same level as HTML but far more advanced. I think Silverlight and an Adobe project are trying to address this at the moment, but we really need something that's free software and has a lot of the simplicity of HTML without its suckiness (and with better asynchronous event support).
The great thing is that there's so much money washing around the world today that I'm working full time and still getting plenty of time to explore ambitions that are not too far removed from what I've listed above - and what I have written there is my ideal. Computers are stupidly powerful; demand for geeks is high; there are lots of people doing neat stuff with music.
A "keyboard" that changes is form and texture to be different interfaces.
A language translation system like Reasoning, Inc's "Refine", but intended to swallow code from ANY language,
libraries, factor and globally optimize, and emit code in
any other language.
With some tool that points out "routine A" is a not quite complete (errors or
omissions) implementation of "routine B".
Eventually, using the accumulated library of code, produce a goal (constraint) based
language that asks interactive questions, to produce new programs.
Actually I need to re-build it because some crackhead stole the first one on the way back from the Maker Faire last year.
There's interesting work going on in the problem space, (CiviCRM being the 500lb. gorilla) but no free turnkey solutions I know of. Some organizations literally have to beg to scrape up the money for a basic web hosting account, so monetization of the platform would be tough.
Not sure if it does all they need, but ironic timing.
You might be able to make money at this by offering a no-hassle hosted version for the lazy, and give away the API and GPLed reference implementation.
I'm interested in EAs like ericb, and want to see, if combined with human creativity, they can produce interesting creatures.
Also, once I have time, I've thought of implementing a MUD in scheme so that it is reprogrammable by the players. I'd use kawa so they also have the power of the java libraries.
2) Writing fiction.
3) Create a programming language.
4) Audio "visualization" of complex data. It's an idea I've had for a long time: turn multi-dimensional data into sound (not necessarily musical) using pitch, rhythm, overtones, beats, and many other parameters. I figure our brains might be able to parse sounds easier than images for certain kinds of data.
ALL kinds of risks there with insurance and what not, but it's mostly just a crazy out of the box, but a bit too optimistic fantasy.
so... 2nd option.
A better web.py web site.
I'd also like to build a missile pack to shoot missile plumes from, like on the MechWarrior Vulture or Gundam.
It is impossible for, say, the New York Times to change or obliterate what it printed on some day in 1997. For NYT.com it's a simple as a mouseclick and a letter. That means there is no such thing as "public record" on the internet. Public record is the basis of a literate society.
"LOCKSS (Lots of Copies Keep Stuff Safe) is an international non-profit community initiative that provides tools and support so libraries can easily and cost-effectively preserve today’s web-published materials for tomorrow’s readers."
Currently OpenBSD-based, but I think that I heard that they are looking to migrate to FreeBSD to take advantage of ZFS.
Phase 2 would obviously character interaction, but I never got to that point.
People walk in wearing t-shirts and flip-flops, I make them a sandwich, and they go away happy. Immediate gratification for both of us, plus I get to invent cool sandwiches. And fresh bread too!
The thing I'd build if I really had nothing else to do would be a shift-swap webapp for my food coop ( http://foodcoop.com/ ).
There's some synergy in that every programming language needs some system to script if it wants to gain massive adoption, so you might as well use the programming language you just invented to write the cool app you're about to invent. Plus, there're some really interesting possibilities if you could replace all these social networking websites with a gigantic distributed P2P network. Social software partitions very well - even if a particular service has millions of users, it's unlikely that any given user exchanges data with more than about 150 other users. You'd eliminate the need for Facebook, at the least, and probably the need for most web startups.
One other thing I really wanted to investigate: using statistical uptime records to distribute data across peers and get high-availability. One problem with P2P networks is that infrequently-accessed material tends to become unavailable - this is even an explicit design goal of FreeNet. It's fine when you're downloading the latest TV show, but it's a real problem if you want to distribute micro-social-network-apps over P2P. Yet most people keep their computers on at fairly predictable times of the day. So you could use that statistical uptime data, from their peers, to figure out where to spread data to make sure that at least one copy is always online at all times.
Then you could merge that with access-logs so that data that's frequently accessed migrates to computers near the ones that want to access it. Sorta like a P2P-Akamai. I've noticed that my flists on FaceBook and LiveJournal are very cliquey, and it's always the same folks posting. It's not outside the realm of computational feasibility to connect together your computer with all your friends' computers, P2P, and only push updates to people who are interested in them.