Getting Real About Artificial Intelligencepost was originally published on this site
What’s the most exciting technology on the horizon? Is it self-driving cars? Voice-activated computing? Computer-assisted medicine? While each of these technologies is exciting and has the potential to revolutionize our lives, they’re all based on a single technology: artificial intelligence.
The potential impact of AI is incredible. It’s difficult to overstate the importance of this technology. According to a study by the McKinsey Global Institute, AI could deliver $13 trillion in annual global productivity gains by 2030.
Shaker/Montage is a leading entity in the field of AI and its applications for business. I recently had a conversation about the current state and future of AI with Shaker’s Eric Sydell and Isaac Thompson. Sydell is executive vice president of research and innovation, and Thompson is principal data scientist. Here’s some of what we discussed.
What Do We Mean By ‘Artificial Intelligence’?
AI is loosely defined as computers that can learn and adapt to input and feedback. Many applications of AI today are based on the concept of “deep learning.” This technique of using computers to analyze enormous sets of data for patterns and insights is behind everything from computerized chess programs that can defeat the best players in the world to radiology systems that can diagnose cancer earlier and with more accuracy than can be done by humans.
Are There Any Drawbacks to AI?
Just like any other powerful tool, AI can be used in good ways and bad ways. It falls on us to use AI in a way that is fair and equitable, and that leads to good outcomes.
Is There an Inherent Bias in AI
Humans have innate biases, while AI evaluates the data it is given without any built-in biases. If the data we feed an AI system is itself flawed, limited, or biased, then the resulting prediction will mirror those issues. Programmers need to ensure that AI is developed using good data, and must be careful to avoid unintended consequences like bias.
Is AI a One-Size-Fits-All Solution?
Different companies will use AI in different ways. One of the biggest problems with AI today is that many companies are rushing solutions to market without fully understanding how they work, let alone how they’ll perform in the real world. Developing beneficial AI solutions that solve real-world problems requires a great deal of research and revision to get them right.
What Should We Expect in the Future of AI?
Our field and world are in the early stages of exploring the power of AI. We expect the benefits above to be developed over time, and AI to gradually learn to make better decisions than people do. This process requires a lot of data and a lot of research, so it won’t happen overnight. There are also hurdles and pitfalls of which we must be aware.
Artificial Intelligence is one of the most powerful technologies humanity has ever developed. It has the capability to revolutionize our world. And yet, even with the power of this technology in our hands, we must choose how to use it to make our world a better place.