Not many ideas this month, too busy writing code, and for the first time an idea sent in by someone else. If you want an idea to appear in next months dump, either anonymously or with your name attached to it feel free to mail me (email@example.com).
Coming up with these is always lots of fun.
BigEye is an idea about a field that I know just about nothing about, so chances are that it’s already been done. The basics are simple enough: use a very large number of low quality sensors to create a telescope with an enormous baseline and very high resolution. A bit like the eye of a fly, tens of thousands, or even millions of webcams pointed up at the sky looking at the stars. A few bright stars could be used to stitch all of the images together, the sheer number of sensors would be used to obviate the need for chilled sensors and to get rid of thermal noise by oversampling.
The technical difficulties in implementing this are likely ‘non-trivial’, I can’t begin to imagine the amount of processing that you’d need to do in order to merge all the images, do accurate detection of cloud cover and other disturbances, but if it is possible to build something like this I really wonder if you could make it work so well that we’d be able to see much further than even our space based telescopes can do at the moment. Just like ‘folding@home’ and ‘seti@home’ you’d be capitalizing on the fact that millions of people already have all the hardware laying around, all they’d need to do is to remember to ‘put the webcam out’ at night.
Mark Bao did a great job on building a website from scratch and selling it within a couple of weeks, and it has inspired countless imitations. Flattering as that is I don’t think imitating something built by a person that has been very open about the whole process including the success and the monetization issues is a nice thing to do. So, rather than to suggest that people go out and imitate threewords.me I’d like to give you an idea that is a variation on the theme but that has not yet been done as far as I know.
One of the bigger issues with the threewords.me concept is to monetize it, and another one is that it will eventually burn out. People giving you three words about other people builds up a nice database of information about those people though, and even though in that context that’s hard to monetize you could get people to give you three words about websites, brands and products.
That would not be quite as viral, and it would require the cooperation of those websites, but the product here (a database of keywords about products, brands and sites) is something that would have immediate value in the SEO game. So you could sell that database to site owners on a subscription basis to spot new trends and to get inspiration about directions they could take their products in. It would essentially be a ‘mini survey’.
Programming and kids has been a subject that has always interested me, and I’m very much impressed by projects like Steve Klabnik’s Hackety Hack.
Logo and Turtles used to go hand in hand, now it seems that Ruby is the way to go if you want kids to get in to programming (or perhaps, Python).
But building a turtle is beyond most dads, so why not make a turtle ‘at cost’, with a standardized controller talking via bluetooth to the computer that controls it.
Such a turtle could be mass produced for probably less than $10, running of two standard AA penlights.
Flooding the world with Turtles, making it possible to buy them in bulk for distribution to schools, basically doing an AOL carpet bombing with the most interesting peripheral that was ever made for kids would be a great thing to do!
If $10 is too much for you, but you do have an internet connection, we’ll give you access to ours. What it would be is an arena the size of a large sandbox with a huge white sheet of paper as the base. Turtles could do their own thing, or they could compete for a prize, maybe be programmed to play football based on the images seen by the cams (that’s pretty advanced stuff) or in other little games.
Periodically someone would have to park the turtles for a little while and replace the sheet of paper, an overhead camera could save the ‘designs’ for posterity after each run and give you a permalink to it.
Alternatives are obstacle courses, turtle races and so on. Companies could sponsor the operation of the turtle arena by having a turtle run around the arena to write a corporate logo in selected spots, marking off the time to the next ‘reset’.
The internet is a stalkers wet dream, plenty of victims, almost perfect anonymity and governments and providers that are pretty lax in their dealings with these negative elements.
Enter ‘stalker stalker’, a service by ex-victims of stalkers that do everything they can to make the lives of the stalkers miserable. Keeping track of them and exposing them where ever they pop up their heads.
The stalker stalkers would use a wiki like system to keep track of the whereabouts of the perps and would alert law enforcement in those cases where it can be proven beyond reasonable doubt that some crime has been committed. That way the law enforcement guys can do what they’re good at and the people that are better at working with the internet can give them a hand.
I’m currently involved in tracking the movements of a stalker and we’ve done exactly that and I think that there is plenty of employ for a similar system outside of this particular case.
After deciding to stop contributing to HN I’ve had lots of time to reflect on the time I spent there and the subject of community. Imagine a community that would never have to worry about the ‘slide’ towards reddit, digg or any other ‘undesirable’ direction. HN is forever pre-occupied with that (as though ‘reddit’ is that bad!), and even though there are far more important problems in this world it is an interesting one to attempt to solve.
HN does this by using the following tools:
banning of ‘undesirables’
killing of ‘off topic’ stories
The stated mission is to attract the top 1,000 smartest hackers and I think that within that narrow mission statement these tools work well enough that HN is an unqualified success. But what if you had different criteria? What if you wanted to be much more inclusive, what if you wanted the 10,000 ‘nicest’ people (by some definition of ‘nice’)? What if you wanted people that are not motivated by the twin pillars of hacking and start-ups?
Those people might frown if someone got booted out just for being unpopular, and those people might not agree with the killing of a story just because community leaders decided that a story was not ‘fit for discussion’.
Downvotes are a way to express a strong dislike about the way a particular opinion is voiced and to send a direct message embodying a message of negative peer-pressure to a person writing a comment, but those same downvotes can be used to gang up on people.
So, how to solve that?
If evaporative cooling / eternal September is the way in which eventually every community comes to an end or some stagnant state maybe the way to go is not to try to fight it but to embrace it and make them cornerstones of the community. So, be ‘digg’ to the ‘diggers’, reddit to the ‘redditors’ and so on.
In one of Neil Stephensons books (Snowcrash) there is a bar in a virtual reality world called ‘the street’. The bar is called “The Black Sun”, and it is ‘the’ place to be. A spot to network, make deals, meet (interesting, or nice) people.
So, that’s where the name for this idea came from. The twin evils of evaporative cooling and the influx of ‘new’ members could be dealt with by creating a system of ‘leveling up’. What this would do is to allow people to migrate from one level of the layercake that forms the society of all the people there to another, if they want to and if a majority (or some other number) of members at that next higher up level would agree that this person is a nice enough contributor at the lower level that he or she would be seen as a ‘net positive’.
Like that evaporative cooling simply becomes ‘levelling up’, if you’ve seen enough of the current level to be either bored or disenchanted with the direction, and if enough people higher up in the cake feel you are welcome there you could take the plunge and level up. You lose contact with the lower level other than to express your opinion on votes further down, possibly even the communication at the higher level would be hidden from view (or maybe just the ability to vote).
The issue of downvotes is a complicated one, I would use downvotes as a partial disconnect, if the same people downvoted others consistently I would simply stop showing the output of the one to the other to avoid further infighting and flame wars. That way serial downvoters would isolate themselves and people structurally misbehaving would be isolated by their peers.
In Neil Stephensons book The Black Sun features an area just outside it where all the people that would like to be on the inside gather, that’s what gave me the idea of the multiple levels.
Anyway, this is a complicated concept, it would take a lot more work to flesh it out, but I can see that there may be a way to ‘hack’ the community issue in such a way that one single site could span the entire continuum from 4chan to HN and present a uniform face with different content to the people at each of the levels.