Unlike, say, particle accelerators—large
Posted: Mon Feb 03, 2025 4:00 am
Even simple interactions can lead to trouble. In 2011, a biologist named Michael Eisen learned from one of his students that the least expensive copy of an unremarkable, used book, The Making of a Fly: The Genetics of Animal Design, was available on Amazon for $1.7 million, plus $3.99 shipping. The second-cheapest copy was selling for $2.1 million. The sellers involved were reputable, with thousands of positive reviews. Eisen visited the book’s Amazon page for several days in a row and found that the prices kept increasing, and doing so on a regular basis. Seller A’s price was 99.83% of Seller B’s price; Seller B’s price kept dropping every day to 127.059% of Seller A’s. Eisen surmised that Seller A had a copy of the book and was trying to lower the price to the next level. Seller B, meanwhile, did not have a copy, and so the book’s price was higher. If someone bought the book, seller B could order it on behalf of that customer from seller A.
Each trader’s intended strategy was rational. But the interactions of their algorithms produced irrational results. The interactions of thousands of machine-learning systems in a natural ecosystem promise to uae number data be far more unpredictable. Financial markets, where advanced machine-learning systems are already being deployed, provide an obvious breeding ground for this kind of problem. In 2010, in the space of an excruciating thirty-six minutes, a “flash crash” caused by algorithmic trading wiped more than a trillion dollars from the major U.S. indexes. Last fall, JP Morgan analyst Marko Kolanovic argued that such a crash could easily happen again, since so much trading relies on automated systems. Intellectual debt can accumulate where systems collide with one another, even if they are not formally connected. Without anything resembling a balance sheet, it is impossible to know, either in advance or retrospect, whether any particular amount of intellectual debt is worth it.
The rise of intellectual debt may also involve a shift in our thinking, from basic science to applied technology. capital projects backed by consortiums of wealthy governments and run by academic research institutions—machine-learning tools are as readily embraced by private industry as they are by academia. In fact, the kinds of data that yield useful predictions may be more readily available to Google and Facebook than to any computer science or statistics department. Businessmen may be perfectly content with such unexplained knowledge, but the intellectual debt will still grow. It will belong to corporations far removed from the academic researchers most interested in paying it off.
Each trader’s intended strategy was rational. But the interactions of their algorithms produced irrational results. The interactions of thousands of machine-learning systems in a natural ecosystem promise to uae number data be far more unpredictable. Financial markets, where advanced machine-learning systems are already being deployed, provide an obvious breeding ground for this kind of problem. In 2010, in the space of an excruciating thirty-six minutes, a “flash crash” caused by algorithmic trading wiped more than a trillion dollars from the major U.S. indexes. Last fall, JP Morgan analyst Marko Kolanovic argued that such a crash could easily happen again, since so much trading relies on automated systems. Intellectual debt can accumulate where systems collide with one another, even if they are not formally connected. Without anything resembling a balance sheet, it is impossible to know, either in advance or retrospect, whether any particular amount of intellectual debt is worth it.
The rise of intellectual debt may also involve a shift in our thinking, from basic science to applied technology. capital projects backed by consortiums of wealthy governments and run by academic research institutions—machine-learning tools are as readily embraced by private industry as they are by academia. In fact, the kinds of data that yield useful predictions may be more readily available to Google and Facebook than to any computer science or statistics department. Businessmen may be perfectly content with such unexplained knowledge, but the intellectual debt will still grow. It will belong to corporations far removed from the academic researchers most interested in paying it off.