First they said machines would never beat human champions at chess. But on Feb. 10, 1996, IBM's (IBM -0.56%) Deep Blue defeated Russian chess grandmaster Gary Kasparov. Then they said machines couldn't possibly outplay humans at the more complex game of Go. But in 2016, Alphabet's (GOOG -1.80%) (GOOGL -1.82%) AlphaGo took down one of the world's top Go players, Lee Sedol.
One more first for artificial intelligence (AI) took place a few days ago. Two different AI systems beat humans at reading. This achievement could have even more far-reaching implications than those other milestones. And it provided yet another reason why you should invest in AI sooner rather than later.
Image source: Getty Images.
Wikibots
It's not a new thing for AI systems to read material and learn from it. But with the reading capabilities of AI advancing significantly, Stanford University developed a test designed to see just how good the computers have become. Stanford AI experts created over 100,000 questions based on 500 Wikipedia articles. The test is considered to be one of the best barometers of machine reading.
Until this week, no AI system had ever beaten humans. However, Alibaba (BABA -0.94%) announced on Jan. 22 that its deep neural network had scored 82.44, higher than the human score of 82.304.
That all-time high performance by AI at reading comprehension didn't last very long, though. An AI system developed by Microsoft beat humans on the Stanford test the next day, achieving a score of 82.65.
How did Alibaba and Microsoft do it? They used natural language processing (NLP). It's a branch of AI that uses machine-learning algorithms to learn language syntax and semantics. If you have used Siri on an iPhone or Amazon's Alexa, you have already seen NLP in action. NLP was also behind the historic victory of IBM's Watson AI system in 2011 over human champions of the TV game show Jeopardy!.
Reading between the lines
Microsoft spokesman Andrew Pickup told CNN that "these kinds of tests are certainly useful benchmarks for how far along the AI journey we may be." He added, "However, the real benefit of AI is when it is used in harmony with humans."
Contrast that with the words of Luo Si, chief scientist for NLP at Alibaba's Institute of Data Science of Technologies. In a statement released by Alibaba, Si said, "The technology underneath can be gradually applied to numerous applications such as customer service, museum tutorials and online responses to medical inquiries from patients, decreasing the need for human input in an unprecedented way."
Image source: Getty Images.
That doesn't sound like harmony with humans. However, I think Si was right. AI will radically impact all of the areas he listed and a lot more. Predictions for how many jobs could be displayed by AI systems vary greatly. Professional services firm PricewaterhouseCoopers (PwC) projects that 38% of American jobs could be at risk of replacement by automation within the next 15 years. Other estimates aren't nearly that high.
On the other hand, Pickup's view has merit also. For example, IBM says that its focus is "on building practical AI applications that assist people with well-defined tasks," rather than replacing humans. This approach to AI is frequently referred to as intelligence augmentation (IA). The purpose of IA is to increase the productivity of workers and eliminate the need to perform boring and repetitive tasks that most people don't like to do.
Perhaps the most likely scenario is that IA applications are more prevalent over the next few years, with AI helping humans in their jobs more than replacing them. However, as AI gets increasingly smarter, it seems a foregone conclusion that many jobs currently performed by humans will instead be done by machines.
Investing in the future
Fighting change has never worked as well as embracing it. Instead of fretting about the future implications of AI, the better course of action is to invest in it.
All of the companies mentioned earlier are leaders in AI. And there are plenty of other great stocks to consider buying, including some that could be overlooked in conversations about AI. I have personally invested heavily in two AI stocks that I think have enormous potential -- Alphabet and NVIDIA (NVDA -2.48%).
Alphabet uses AI to power its driverless cars. It uses AI in its Google search engine algorithms. It has acquired over 20 smaller AI companies in the past few years. And Alphabet has access to massive amounts of data, which serve as the fuel for AI applications.
NVIDIA's graphics processing units (GPUs) are in high demand for AI applications. In fact, NVIDIA's customer list contains many AI leaders, including Alphabet and IBM. Deep learning requires intensive processing -- and NVIDIA's GPUs are the gold standard for handling this processing.
One thing seems clear: AI will continue to become better at humans in more areas than just games and reading comprehension. Some will profit from the trend. Why not you?
