Another Round

It has been a while - so here are a couple of ideas:

1) A couple of years ago I told one of my friends how it would be interesting to correlate the news.google.com news articles from the past couple of days to today's stock prices. The algorithm I envisioned would "fish" for correlations between somewhat predominant word combinations in the articles and changes in stock prices. Imagine how excited I was when I saw that the new finance.google.com service lists news articles with the company name next to the stock price graph and points out with arrows when each news item came out directly on the price graph. A correlation is only a step away.

2) On a related note, online trading services such as etrade already provide a way for their users to place a stock on a "hot stocks" list (or a "watched stocks" list). Now if the progression of stocks from "unnoticed" to "watched" to "purchased" to "sold at profit" is analyzed across the user base in real time, not only the most savvy investors can be singled out, but also the service can (for a fee) offer the results of such analysis in real time back to its users, displaying stocks which recently became most watched (by the investors whose predictions are above average at bringing profit). Such a service would only display aggregate information to respect the individual users' privacy.

3) If one was to take a two hour home video of a birthday party and subtract from it, frame by frame, pixel by pixel, the two hour star wars movie (given the same resolution), the result would be a two hour nonsensical "difference" movie. Now if the "difference" movie and the home video are posted on a service such as google video, which (for now) allows posting movies of unlimited length for free, we would have two movies neither of which violates copyright, freely available for download. It is only the users with their knowledge of the two movies' relationship and a simple client side software which downloads and reconstructs the film which could be accused of copyright infringement. Of course, the "difference" movie and the home video do not have to be separate files - they may be interleaved frame by frame or interlaced. This way the copyrighted work is not stored by google and yet freely available. I suspect I should read more about steganography...

That is it for now.