Our Utopian Ideas

I actually had a dream about this tonight. The idea is not very well thought through as there are more interesting things for a guy to dream about at night if you know what I mean.

The idea is to build a community project website where everyone could feed bits of knowledge to build a base for an Artificial Intelligence.


First Stage - Object Recognition

First thing AI needs to distinguish are objects. Community members -- or just casual passers by -- could enter names of various objects. AI would periodically run search for these words on images.google.com and gather a an image base. Users could help identify which pictures are relative an which are not.

By applying well established shape recognition algorythms, AI could start gathering object database and, eventually, start recognizing objects from images on its own.


Second Stage - Composite Objects and Action Recognition

When there would be a critical mass of objects that AI recognizes, the project could be taken one level up - to recognize composite objects and relating them together resulting in action recognition.

A number of possible actions could be related to the particular object. Again images could be used to give AI a clue for a visual distinction of the action.

For example:
milk pours
man runs
building collapses

By relating various objects and action info in a composite image AI could actually start recognizing what's going on in the picture.

I.e.:

girl [object] + bottle [object] + white liquid (milk) [bottle property] + glass [object] + bottle at nonvertical angle (pours) [action] + bottle close to a glass [relation]
=
girl pours milk into glass from a bottle


Third Stage - Video Recognition

Introduction of video brings even more simplicity into action recognition. As AI has not one but numerous pictures to analyze (something like 30 pictures a second).

By relating analysis results of each consecutive image (frame), AI could even easier relate what is going on.

Example #1:

Frame #1: tower [object] + angle 90 [property]
Frame #2: tower [object] + angle 89 [property]
Frame #3: tower [object] + angle 88 [property]

Result: Tower is falling.


Example #2:

Frame #1: man [object] + size X [property]
Frame #2: man [object] + size X+1 [property]
Frame #3: man [object] + size X+2 [property]

Result: A man is going towards camera.


Forth Stage - Speech Recognition and Communications

This is the part where I woke up. But I promise to figure that out until someone implements the three previous stages ;)


Promotion

In order for such project to work it should have a critical mass of contributors. Therefore media attention is a must. It should be fairly easy to reach that since many popular sources (like for example hi-tech blog gizmodo.com) tent to feature such crazy projects.

Btw, gizmodo, it's time to feature this site, don't you think? C'mon, you're my third favorite daily read after StuffOnMyCat.com and Pearls Before Swine.

I've always had a fetish about autogenerated content. I also unconditionally love Cyberpunk literature. I think these are the two reasons behind the idea that sick brain of mine generated like eight years ago.

Why can't a computer software be written that generates whole books or at least short fictional stories?

As crazy as it sounds it might prove not that utopian after all. Of course computer cannot generate a readable book if given only a plain word dictionary. Well ... not until at least a proper AI is invented. But then it wouldn't matter 'cause we would be slaves to robots or something.

No - simple dictionary is not enough. A human should feed software with some innitial well structured info, like:

• Fictional business, country and other entities with corresponding descriptions and core values that could be used in determining their relationships with other objects, events, etc.
  
• Non-standard objects, like custom rifles, vehicles, etc. Their relations to business entities.
• Highly configurable plot structural elements - events or connecting points (these could also have many settings, like number of persons participating in it, potential outcomes - death, injury, discovery of a secret, etc.)
• Character names and various stats about them that computer could go crazy randomizing
• English grammar rules, thesaurus, etc.
  
• A database of standard objects, political and business entities, historical events, real people and brands, etc. Each of these DB objects could have a number of config variables that could be used in determining their relationship and relevance to plot and/or other objects.
  

Prior to generating an actual book, an operator -- I don't dare to call him/her a writer -- would have to set numerous configuration parameters to head computer in specific direction, i.e.:

• Violence level
• Plot twist level
• A level of fiction vs real
• Overal freedom for randomness
• Particular characters, objects, etc. to use
• etc.
  

Of course, the output from computer would still require many human editing and final touches. But I like to toy with the idea of the unpredictable twists and crazy text that computer might randomize and provide good base for a real person to finish on.

Pretty simple, huh? :)

I absolutely love to think about the controversy this would create if it ever becomes real. With real writers going all like "This computer-generated stuff will never be as good as ours - it is limited and lacks human touch" all over media. Cool :)

Btw, the reason I mentioned Cyberpunk earlier is that I think it is a perfect genre for this kind of autogenerated content. Cyberpunk plots are fictional enough to allow coputer-generated craziness. Still it's based in pretty real "earthly" settings.

It would be hard to generate something as psychodelic as "2001: Space Odyssey" or as sensual as "Brokenback Mountain", wouldn't it?

The idea is to have an implant somewhere in the blood stream system. It could monitor vital body stats, alcohol level, and provide immediate info and historical log through mobile devices and web servers.

Technology

I'm neither a doctor nor a scientist. But I'm quite positive that there is plenty of energy in human body to power a tiny device.

I certainly not sure if such device is feasable right now. But hey, look at the site name :)


Alcohol level monitoring

The most obvious consumer-grade application of this device is to monitor alcohol level in blood. It could be usefull for all types people. Recovering alcohol addicts could be monitored by their doctors or sponsors. Occasional booze lovers (like me) could get an early warning when they're about to exceeed their "safe limit".

The communicational features of the device would offer many opportunities for drinking self- and enforced control. Several examples:

• Wife gets SMSed when her husband overreaches the limit which they agreed upon
• A blouetooth-enabled mobile phone displays warning when approaching configured "safe limit"
• Device could poll some webservice in order to keep log of your alcohol consumation. You could derive various stats from this data later on.
  
• And even more utopic one: a compatible car reads info from its owners device and refuses to start if alcohol is above allowed limit (just in case someone is wondering - I've never driven while drunk)
  

Besides all the obvious "correct" reasons, there is a lot of potential in using this for fun.


Connectivity

The device should be Bluetooth, WiFi and/or GSM enabled.


Medical uses

Beyond it's primary task - alcohol monitoring - the device potential is limitless in medical scope:

• Emergency teams could read vital stats from victim: drugs he/she recently has taken, etc.
• Doctors could monitor their patients' vital stats (blood pressure, sugar level, etc.) and give early warning

Ethical issues

This device should be by no means an RFID. In fact it should not have any distinctive ID at all. I think that any device to read from the body monitor should at first complete pairing process - maybe even implant-initiated.

I'm sure there are ways to make it a truly anonymous medical device rather than digital dogtag.

Ok, this one's been nagging me for quite a while.

Have you ever been taken to a mall after the stormous night out with your friends just to find yourself tagging along your wife through N-th shoe parlor? Has your fligh ever been delayed trapping you for hours in a foreign airport?

It happens for me all the time.

That's when I remember my several years old busines idea - Cheap Hotel.

The Idea

Everytime I'm in a/m situation I think about nothing but taking a quick nap for several hours. Supposedly I could find a free bench and crash there. Unfortunately my bag would be stolen just before I would be escorted out of the building by security guards as homeless person. Besides it's not too pleasant to be sleeping in the middle of a crowd.

So, I lack security and privacy. That's where my idea comes in - a small room with a bed you can lock in for several hours - Cheap Hotel.


Facilities

The clusters of "rooms" or more like cells - would be located in some heavily crowded places like malls, airports, stations, etc. Never in an outdoor or deserted enviroment where they're would quickly be vandalized.

Large places of attraction solve one more problem - sanitation. Cheap Hotels would not have to have bathrooms or any other water related facilities of their own. Inhabitants could simply use buildings facilities.

Basic cell would hold only bed. Premium - additional services like wall-TV, touchscreen internet-enabled computer, etc.

Double cells could also be available.


Maintenance

The very big advantage for such service would be that the facility would be almost totally selfmaintainable.

The cells would be equipped with authomatic doors, credit card reader and bed.

Bed shoud be made from some very durable material that cleans easily. Perhaps each inhabitant could be issued some kind of disposable pillow by the built in dispenser.

The whole beauty of this idea is that it would take only one two-person team to maintain several clusters of Check Hotels.


Vandals

Of course, there is still a chance that some dorkhead would take a dump on a bed and leave it there rendering the cell unusable until next routine maintenance round.

Some equipment might also be damaged or destroyed. Although that is very unlikely due to the nature of security in such crowded places.

Supposedly cameras could be installed to use as a proof for vandals, but that would be a real deal-breaker for some of us, wouln'd it'?


Payments

This is the good part. The whole charging would be done by using credit cards.

Say, a tired person wants to take nap. He comes up to a Cheap Hotel. Finds unoccupied cell. Inserts his credit card. The door unclocks. The guy comes in and crashes. In 2 hours his wife calls on his cell to tell shes already met her quota for shoes for today so she's ready to go home.

The guy goes out of the cell. The booth charges his card and spits it out, prints a receipt, says THANK YOU on a tiny LCD display.

The customer is charged for the duration of his "stay" on a per-hour basis plus usage of premiuom facilities (TV, Internet, etc.)


The Cost Of Business

The more I think about this the more I see a success story here. The only thing that concerns me is development of the automated booths.

Other costs should be very small. Due to small footprint of one cell (2-3 square meters) it shouldn't be too expensive to rent an unused space in a mall or airport.


Additional revenue opportunities

• Setting a wake-up timer
• Ads of the nearby services/shops

Updates

06/12/2006

One commenter pointed me to a simmilar invention - a Capsule Hotel in Japan. Thanks!

However this is not the same for a number of reasons:

1. Capsule Hotel is claustrophobic - I would NEVER agree to sleep in such coffin.
2. They're still positioned as a hotel - i.e. they charge for a night, whilst my idea is renting a secure and private location to crash in WHENEVER you want or feel like to and get charged an hourly fee.
3. Capsule Hotels still require attending personnel as well as buildings to maintain. My idea is to utilize spare space (one that might not be attractive to stores) in crowded places. That's why I thing my idea is more cost-effective from the business point of view

There are more things that make Capsule Hotel unattractive in my european eyes, such as you have to leave your luggage in a separate locker (and propably charged for it extra), you have to leave your shoes and wear special slippers, you OBVIOUSLY cannot make sex in there :)

I guess some people made big money from selling domain names of small countries (usually islands), such as .tv (Tuvalu island), .cc (cocos islands), .fm (Federated States of Micronesia), .tk (Tokelau), .ac (Ascension Island) and some other. Well, I wouldn't buy .ac or .tk domain, but .tv, .fm are really good ones!



So, I thought that it should be possible to create a new country (or some kind of half-independent island) which could get official country code and top level domain and then sell domain names. Of course, the creators should pick a name carefully. Here are some unused top level domains which could be popular:

* .go (Maybe Sweden should let Gotland be free?);
* .aa
* .zz
* .ww
* .oo
* and other unused domains with the same letters.

Here is the full list, so you coud think about other possibilities.

The Problem

Not exactly a business idea. It's more of product I'd like to use myself.

My living room is the most computerized place at home. With 3 PCs, an XBOX, TV and stereo set the room is meant to be a place a guy would be willing to spend 110% of his spare time in.

The trouble is that sometimes I kind of want to use all the a/m equipment at once. I.e. chatting with a friend over ICQ while trying to nail a Silent Assasin rating in just another level of Hitman: Blood Money on the XBOX. Supposedly I could run back and forth from one piece of equiment to another but that's not very efficient use of your time. (Married ones will understand me - in those rare cases when your family is away you just HAVE TO make a good use of your time)

The Solution

I would certainly appreciate some kind of device -- a scoreboard -- that would communicate with PC and display various configurable stats, like:

• Number of unanswered IM messages
• Number of new mail messages
• Todays AdSense revenue
• etc.

The device does not have to be built on OLEDs - just a simple LEDs- so it's as cheap as possible. The LEDs have to be bright enough, so you can see the numbers in the dark (you know - to put one in your sleeping room, too, so you can check your "stats" at 3am if you wake up suddenly).

Device can be equipped with Bluetooth or WiFi connection for easy communication with PC. It should be able to accept info from multiple PCs.

Ability to produce sound signals would be a also appreciated. (you know - those ICQ "uh-oh's" have to be heard in bathroom, too)

And yes - the device has to be under like $60.

Anyone knows such device? Wrap up a couple for me.


API

The device drivers and software should have an open API. This way we would see a lot of independent plugins popping out.



It could also be able to read specially formatted RSS feed off the Web. Some people would create interesting services like live match stats, weather reports, news tickers, jokes, you name it.


Updates

06/19/2006

I came across this device. While the screen is abviously too small and there are no wireless connection options, this device is a step in the right direction.