When people think about artificial intelligence, they often think about what intelligence machines can do on their own, without human interaction. However, some of the most powerful and useful AI systems are those that help or augment human capability. The idea of AI systems that augment human ability is known as augmented intelligence. On a recent AI Today Podcast, Professor Tom Davenport from Babson College, senior adviser to Deloitte’s Analytics and AI practices, and fellow of MIT initiative for digital economy shared his perspectives on the power of AI as applied to augmented intelligence.
For over twenty years, Professor Davenport has been a professor, consultant and researcher on various topics, mainly analytics and big data. In addition, Mr. Davenport has written multiple books focusing on analytics, including co-author of “Only Humans Need Apply: Winners and Losers in the Age of Smart Machines”, and “The AI Advantage”. As part of writing these books, Davenport interviewed many people who were already using augmented intelligence.
One interesting insight comes not from AI, but rather from another technology that aimed to replace human activity – the Automated Teller Machine (ATM). When ATM machines were first put into place in the 1980s, there was widespread concern that it would eliminate the jobs of ordinary bank tellers and bank operations. However, according to Davenport, “One of my favorite statistics is that there are roughly the same number of bank tellers now, as there was in 1980 despite all the ATMs, internet banking, and other such changes.” From this perspective he sees AI too not having the same sort of disruptive effects on employment as many might at first assume.
From Davenport’s point of view, introducing technology that automates and performs tasks previously accomplished by humans actually creates more jobs for people who take time to learn about how they work. For example, these new machines create opportunities for technicians and programmers and whole new industries that are enabled by new technology. However, he did state “there are some concerns and reasons to not be complacent, reassuring ourselves jobs won’t go away and available for everybody may not be true, however, I think most situations will have people working along the machines, if they are willing to learn.”
Perspective on Artificial Intelligence
AI and augmented intelligence to be a way to help humans perform better at their tasks, and be more efficient. Therefore, helping people complete a task easier, quicker, and using less energy that can otherwise be spent doing another activity. Professor Davenport sees many ways that AI is helping us to be better at completing tasks or accomplishing goals that are not easy to accomplish otherwise. He states in the inverview, “Surgical robots are designed to help surgeons with tasks, while surgeons maintain their role.” However, he also stated, “I believe at times it will be our responsibility to make the machine work better.” Therefore, depending on the situation humans may be dependent upon the machines to help make tasks more efficient. Meanwhile, the machines will need humans for repairs and ensuring they are operating properly.
In order for artificial intelligence to work properly, it needs good clean data. As such, Professor Davenport believes the role of data scientists are increasingly needed. He wrote an article on the topic with current Chief Data Scientist of the U.S., DJ Patel, called “Data Scientist Sexiest Job of the 21st Century”. He interviewed multiple data scientists for the article, stating, “there was not much analytic work taking place, rather it was finding data and cleaning data as that is a major issue for nearly all types of AI.” Additionally, Davenport pointed out that “we do not currently have a title for those who specifically work with AI. The role of data scientists was originally to work with data and analytics, and there needs to be a new term for the role of working with AI.” Indeed, people are starting to think more about how important data is, and the role of data that is required for AI to work.
In addition, Davenport has opinions on the role and increasing need for regulation in the context of AI. He believes that there is a role for the federal government to provide in the United States for overall legislation and regulation, but as with most technology-related matters, it really is up to the local states and jurisdictions to put together practical laws and regulations based on how AI will be actually put to use.
In general, Professor Davenport is mostly bullish on the future of AI. He speaks to many executives on the topic, and believes that it is primarily augmented intelligence applications of the narrow AI type that will dominate the sort of applications put into practice. He stats that even Jeff Bezos of Amazon noted in his 2017 shareholder that the majority of machine learning was quietly improving their processes, rather than high-visibility, glamorous uses of AI. According to Davenport, this is the way to move towards major tasks, minor tasks, and those that just help us live our lives better and more productively.