If Hallmark carried a category for IT industry milestones, there would have been a lot of birthday cards passed around last year.
IBM’s PC turned 20. E-mail turned 10; so did Linux. We’ve also celebrated anniversaries in the microprocessor field and earlier this week, I discussed the five-year-old
DVD. Inevitably, a kind of nostalgia has set in as we look back at a sort of golden age in which the field’s great discoveries were made. These are well worth celebrating, but it’s also important to keep our eyes open to the new pioneers in IT. I’m not talking about the dot-commers, who merely tried to use a new medium to create delivery channels for traditional products and services. I’m talking about the real inventors who are pushing the boundaries of how technology is created.
In other words, I’m talking about Leonard Aldeman.
Aldeman first got the world’s attention in 1994, when he managed one of the first successful experiments with a computer made from deoxyribonucleic acid (DNA). A computer science professor at the University of Southern California, Aldeman began with one of the most basic brainteasers: how to get from A to B. Specifically, he tried to get his DNA-based computer to figure out the shortest route among seven cities.
His method seems almost childishly simple. Each possible route was represented by a stand of DNA. The problem (which is the best route?) was represented by another strand. These DNA strands then mix together, but only one of the possible “”solution”” strands will successfully bind with the “”problem”” strand. If you remember anything from your high-school biology class, you could take it a step further and think of it as being similar to the way chromosomes mix and match genes, leaving some dominant traits (like eye colour) and some recessive ones that don’t make it into another generation.
It may have been revolutionary, but it wasn’t fast. Almost anyone with some basic math skills could have done as well by hand. Earlier this month, however, Aldeman and his researchers announced a breakthrough of sorts: an experiment where a DNA computer had solved a problem with millions, as opposed to seven, possible solutions. Scientists in this field always have a hard time trying to dumb it down so we can understand their work, but they’ve done it this time. They came up with the analogy of a customer coming into a car lot requesting a vehicle with a long list of must-have requirements (in our industry, think a particularly extensive request for proposal for an enterprise IT project).
So far, DNA is too error-prone for commercial applications, but the technique it uses offers profound hope for the high-end market. Even with the most sophisticated processors running in parallel, today’s computers wade through each possible solution one by one. DNA is pure parallelism — its process of elimination allows the computer to assess (in this case) hundreds of thousands of possible solutions at once. Think of compute-intensive environments like life sciences. Think big money.
Bringing DNA computing into the mainstream is like trying to cure AIDS — no one knows if it’s centuries away or just around the corner. I may be naive, but I’m counting on the latter. In only eight years, Aldeman has already made considerable progress. He is without a doubt our next Andrew Grove (though hopefully not our next Bill Gates). Pay attention to his work now, because it will lay the foundation for the next stage of IT development. There is comfort in the great accomplishments of the industry’s past, but let’s not ignore the future that is unfolding like the twisting molecular vines that make up the very essence of life.