Sometimes (even very smart) people get carried away with the idea of AI being an existential threat. I believe those fears are overblown. As artificial intelligence evolves, there are plenty of opportunities for AI to enhance and optimize the work we do—without threatening our livelihoods.
However, there are risks inherent in how we use AI. These risks are more mundane than sci-fi horror stories, but they’re real for businesses and consumers.
Keep reading for an overview of the risks of AI and how humans and machines can work together harmoniously, or click here to watch a video version.
Last year, I took my sons to go spend the night with their grandfather. When we got there, my father-in-law was moping around his home. You see, my father has a dog that is getting older and has started having accidents in the house. This time, the dog had done his business right in the middle of the living room (and we’re talking a number two here). The problem is, he had a Roomba vacuum cleaner... and you can already see where this is going.
The device woke up and started its normal cleaning routine. When it encountered the pile of, well, you know, it proceeded to push it around the whole house. What started as a small localized mess turned into a whole afternoon spent with bleach in hand and a sulking father-in-law.
This, to me, is an example of the real risk of AI.
You often hear people talking about narrow AI. The Roomba vacuum is an AI system that has a very narrow goal: to learn the floor plan of your house and clean it everywhere.
This device encountered a situation that was outside of the norm, and as a result, it didn’t know what to do. This is the difference between people—who have a high level of reasoning—and narrow AI systems.
In this instance, the AI vacuum made a bad situation worse. But, what if these AI systems are used in financial markets or autonomous weapons and they encounter something that’s outside of their training data? That is one of the real risks of AI—and it’s happening now.
What do you do about it? Firstly, we need to recognize and understand the limitations of AI’s that are deployed in the real world. Secondly, there needs to be oversight. You can’t give these machines complete autonomy in high-risk environments. People are a key part of the loop, and that will never change.
This leads into the myth that AI is coming for your jobs.
There are certain AI technologies that are actually designed to replace a human. Think about a robot on a manufacturing line.
Although, a lot of the most interesting AI technology is designed to augment people’s abilities and empower them to do their job more efficiently.
The future of artificial intelligence is coming. Although some jobs will be lost, new jobs will be created and opportunities will be available like never before.