Search Vendors Under Siege: The AI Aiyaiiii Revolution
September 26, 2014
Science is a marvelous manifestation of human curiosity. I read “Five Ways the Superintelligence Revolution Might Happen.” Note the “could” as in “woulda, coulda, shoulda.” These terms shingle protectively many backsides.
The write up lists the options for “superintelligence”, which is a variant of artificial intelligence with a dash up the Googler’s singularity stirred in for good measure. These scientists remember some of the basic methods of chemistry. List the constituents and mix ‘em up. See what happens.
Here are the five ways technological nirvana will arrive:
- Make intelligence out of software. Essential humans will code a brain. Great idea for a research project. Probably won’t work particularly well for a while.
- Just use math. This is the approach my uncle and his both would have favored. You remember Kolmogorov and Arnold, don’t you? Also, Googlers and Xooglers are into this approach.
- Brute force. This is the technological equivalent of using surplus military equipment to check out a noisy fraternity party. In terms of smart software, the article is into using von Neumann systems to crack P=NP problems. Yep, that’s a good idea.
- Plagiarizing nature. This idea is that one emulates in software biological processes. Think how ants find a ham sandwich at your picnic. I was on the board of a company that puttered around organic algorithms in 2001. For some problems attacked via well managed methods, the results are interesting. Maybe IBM’s new chip or quantum computing will help out, but it is a subset of the math and brute force approach. Messy categorization I conclude.
- Use humans plus any other methods that seems to work. There you go. This is the state of the art. I have discussed the approach in my analysis of Google’s nowcasting model here. Only hitch? Well, it is not right when it counts: Ebola threat, ISIS/ISIL, horse racing.
Net net: Not much in the breakthrough arena. Looks like we’re into incremental improvements. This method will work, but it won’t arrive quickly enough to keep some of the venture firms funding the wild and crazy AI Aiyaiii world.
The outfits that will be directly affected by this AI craziness are the search and content processing vendors. Many of the companies in this sector will assert that their systems are “intelligent,” “able to comprehend human utterance,” and “predict” user needs. The problem is that delivering on inflated expecations is more difficult than doing a PowerPoint about the magic of information access.
Stephen E Arnold, September 26, 2014
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
