Artificial intelligence, also known as AI, is popping up often in conversation at J.L. Grove College of Business at Shippensburg University. The term is also regularly mentioned in media headlines. Some people are excited; many are worried.
AI is such a wide umbrella term that it is often necessary to be more specific about the type of AI we are speaking about. Here, I hope to clarify and highlight the positive opportunities AI is creating, and to map the AI field into two parts.
Repeatedly, technological advances have left us uneasy about the role of machines in our society. Inventions, such as the steam engine more than two centuries ago, or the automobile or airplane more than a century ago, were mistakenly seen as dehumanizing culture and threatening national fabric.
The same worries are surfacing now as the AI wave is gliding over our job landscape. AI has created worries among the greatest of our intellectuals, and thought leaders; among them are Stephen Hawking, Bill Gates, Elon Musk and Henry Kissinger. Are those worries legitimate, and about which type of AI are they concerned?
One way to see the broad development stages of AI is to define applications available today as the stage of artificial narrow intelligence (ANI). This refers to applications that are severely bounded into their domain, such as playing chess, recognizing speech, identifying diseases on microscopic video, acting as digital assistant, trading securities with neural network models, trouble-shooting complex machines, etc.
In this regard we have gone quite far, and machines are routinely outperforming humans on many specific cognitive tasks.
In my view, this is no less natural than an ax outperforming a human hand in wood splitting, or a steamship outperforming a rowboat. We should be glad for the improved productivity AI has enabled, the same way as we cheer many other inventions that improve our daily lives.
The second stage of AI is sometimes called artificial general intelligence (AGI), and when technology approaches this stage, the machines will be more generalists. Once we can move from a machine winning a chess game to a machine learning strategies to win any board game, we have taken one step in that direction.
This stage could lead to new kinds of applications, such as a robot that could participate in the financial market with aim to make profit, or a machine to improve the cost-service level ratio in a supply chain. In the extreme, AGI could reach such a general format that it is setting its own objectives.
The journey from ANI to AGI will certainly be gradual and twisted, but this is the direction where technological development is pointing. Like the path of any creative work, the path of AI is unpredictable. So far we have seen a decades-long acceleration in development, and the future will show how long we can follow this path before we experience diminishing marginal development rates.
We should focus on the positives: AI is enabling automation and improving efficiency. That is a positive challenge, as we are going to see possibilities to revise and redistribute the work in our society. Humans are clever generalists and they are widely benefitting from more efficient tools that ANI is providing.
However, the risks of making the masses replaceable might exist once AGI advances far enough. I don’t see a feasible option to stop the development of AI. The results of AI development so far are positive and provide tangible benefits. So let us use them in work and pleasure.
In this issue, resistance is futile.
Otso Massala is associate professor and director of Charles H. Diller Jr. Center for Entrepreneurial Leadership and Innovation at the John L. Grove College of Business of Shippensburg University in Shippensburg, Pa. Email him at OAMassala@ship.edu.