Artificial Intelligence (AI) has its roots in ancient myth and the study of logic and formal reasoning. In the 20th century, AI first became plausible, and scientists as well as science fiction writers began to explore the possibility that machines could one day think as we do.
In 1950, Alan Turing wrote a paper in which he devised his famous test for detecting whether a machine could think. He proposed that if a machine could carry on a conversation that was indistinguishable from a conversation with a human, the machine could be said to be “thinking.”
In 2014, a computer passed the Turing test for the very first time. Recently, a computer was able to beat one of the world’s best players at the game of Go, a 2,500 year old abstract strategy game, three times in a row.
Today, AI is controlling cars, learning and adapting our web search results to our personalities and interests, assisting with investing, and performing countless other tasks that would have been thought of as impossible as recently as a decade ago.
However, we’re still not yet at the point where computers are anywhere near as good at solving “general intelligence” problems as humans are. Programming a computer to read and understand a children’s picture book or to be able to correctly identify a cat in a series of pictures of a cat is still extremely difficult. The best estimates of computer intelligence at this point estimate that the most sophisticated artificial intelligence may have the IQ of a 4-year-old.
But, AI systems, such as those being developed by Google and government agencies, are gathering information at a pace no human could ever hope to match. With more information, computers continually make better judgments — whether in stock trading or human resources.
Estimates for when computers will exceed human intelligence are currently measured in decades, not centuries. Even the most pessimistic artificial intelligence researchers believe that it will happen in the 21st century. What this means, is that many of us may see super-intelligent computers within our lifetimes.
Elon Musk, Steven Hawking, and many other respected scientists and visionaries agree that super-intelligent AI will happen, and they’re nervous about it. Musk calls AI the “biggest existential threat” to humanity. Would computers that are as far removed from us intellectually as we are to ants see us as their parents or as a nuisance?
At the same time, these same scientists see tremendous positive benefits that could be realized from Artificial Super Intelligence. For example, curing diseases, solving the threats of climate change, and finding a way to give access to food to every person on the planet are just a few of the problems that humanity is struggling to deal with, but that could be solved by a network of super-intelligent machines that don’t need to rest, or to even slow down.
Should we worry? Ray Kurzweil, who has a pretty good track record with predicting the future of technology, has predicted the AI will make it possible for people to become immortal by 2040 — so, if you are inclined to be worried, you’ll certainly have a lot of time in which to do it.
Update: I did a talk at the 2017 We Are Developers conference in Vienna, Austria, which was loosely based on this article. You can view it below!
If you want to read about how this talk came to be and what it was like to prepare for and give this talk, check out my blog post: From Writing to Public Speaking