The Algorithm - artificial Intelligence

In this week’s issue of The Algorithm, I’d like to share some thoughts I have about artificial intelligence, the hype that has accompanied it over the last several years, and how I think we can make the most practical use of it. 

The term artificial intelligence is most often used to describe machines or computers that mimic “intelligent" functions associated with the human mind, such as learning and problem-solving. 

For years now, we’ve seen news article after alarmist news article predicting that artificial intelligence is going to replace human labor in numerous industries. The underlying premise is that machines are going to learn how to perform intellectual tasks better, faster, and cheaper than humans, so of course it makes sense for businesses to replace human workers with intelligent machines. 

The reality is that while an increasing number of these articles are being published, machine learning practitioners and researchers are having a difficult time trying to figure out how to train their algorithms to accurately predict what’s coming around the corner, correctly group similar things together, and find the best ways to arrive at solutions that are important in their fields. Ask anyone who actually does this type of work for a living and they will tell you that getting models to consistently perform well is actually extremely difficult.

In my opinion, it’s not the methods that are misguided but what we are supposedly working toward. Getting machines to possess “human-like intelligence” doesn’t seem like the best use of resources because humans and machines are fundamentally good at different things. 

• Machines are great at performing complex calculations, storing and recalling large amounts of information, and executing digital tasks very quickly.

• Humans are great at tasks that require flexibility and creativity, using context and life experience to understand complex situations, and social tasks that involve communicating and collaborating with other humans.

I believe the best way to incorporate artificial intelligence into society is to create programs that augment, enhance, and extend human capabilities. In other words, let’s take the things we don’t do well and have machines help us do them better. This is known as human-machine collaboration, and it’s an area of artificial intelligence that I am increasingly interested in and intend to research and write more about in the near future. 

Until next week,

Tony Ojeda 
Founder
District Data Labs