ed AlphaGo shock
It’s time to narrow gap in AI technology
How far can the alarming progress of artificial intelligence go? The first and second go matches on Wednesday and Thursday between grandmaster Lee Se-dol and AlphaGo showed that prediction is all but impossible.
The moment the first of the five-game tournament in Seoul ended up with Lee’s shocking defeat, people around the world must have felt that AI is evolving far more rapidly than many expected.
In fact, go, called baduk in Korea, has been seen as a game that wouldn’t allow challenges from computers so easily. Aside from the fact that it involves so many probabilities, go requires a high degree of intuition, imagination and creativity.
So it has been generally believed that top professional players would continue to gain the upper hand over machines. In the run-up to Wednesday’s first game, many experts predicted a clean sweep for Lee.
But AlphaGo, an artificial intelligence system developed by Google, was much stronger than previously thought, and this raises the possibility that AI might change human society completely.
The rapid development of artificial engineering technology offers both a boon and a bane.
Google is confident that humans will use AI to enrich their lives in various areas, including in climate change responses and healthcare. IBM’s AI supercomputer Watson, for example, helps diagnose cancer in patients at Memorial Sloan Kettering Cancer Center in New York. Watson’s diagnoses are reportedly 90 to 100 percent accurate.
Even so, there are concerns that computers might rule humans someday as depicted in some sci-fi films, given that the progress of AI has reached a point where machines can think, study and infer like humans, as was seen in the example of AlphaGo.
As AI develops further, robots could replace humans in professional jobs such as translation and legal services as well as in simple repetitive work. Then ethical problems would emerge as a result of corporate pursuit of profits using AI technology. That is why we expect a global ethics committee to be created sometime in the not-so-distant future so that AI technology can be applied ethically.
But there is no need to be blindly skeptical about AI’s rapid progress. After all, humans in fact invented AlphaGo. It will be better to pool wisdom to get AI to contribute to the development of human civilization rather than worrying if it will take over.
What’s even more worrisome is that Korea, which prides itself on being an IT powerhouse, lags far behind other industrialized countries in AI.
If America’s level of AI development is 100, Korea stands at only 73.1, according to the Korea Evaluation Institute of Industrial Technology. This is compared with Europe’s 85.7 and Japan’s 83.7.
All this might be the consequence of both the government and companies having chased after short-term projects that offer immediate results rather than conduct basic research programs.
It’s a bit late, but Korea should hurry to jump on the bandwagon of the “AI revolution.’’ Policymakers need to lift relevant regulations across the board and devise support measures. Companies have to increase their research and development expenditures with a long-term perspective.