One of the earliest predictions about artificial intelligence was that robots would steal jobs from people, especially for menial and menial tasks.
Case in point: Flippy 2, an “autonomous robotic kitchen assistant” developed by Miso Robotics to flip hamburgers as they cook. Restaurant chain CaliBurger is testing Flippy at various locations. But what a robot does best isn’t what it’s designed to do.
“[A CaliBurger franchise owner] ‘Flippy isn’t as good at flipping hamburgers as I’d like, but it’s great at pulling out baskets of French fries after they’re done,’ said Thomas Davenport, a digital fellow at the MIT Initiative. Digital economy. “We’ve known for some time that AI and robots will take over the jobs of burger flippers — but it hasn’t happened yet.”
Davenport’s new book, “Working with AI: Real Stories of Human-Machine Collaboration,“ Examines 29 examples of humans working with AI-enabled systems, showing when and how AI works best and how companies can use AI responsibly. The book was co-authored by Singapore Management University Professor Steven M. Miller.
The common theme is that AI is not fully automating what humans do. That means many jobs are safe, Davenport said in a recent webinar hosted by the MIT Sloan Management Review.
“One of our main conclusions [augmentation] It doesn’t look like there will be a job loss,” Davenport said.
Davenport and Miller explain how companies are deploying AI to best advantage:
AI works best in “normal” situations
Miller defined automation as a process in which a machine completes enough tasks that the human who previously performed them is effectively displaced. Augmentation, on the other hand, is the automation of small tasks, not the entire task.
“Humans are involved in direct enforcement,” Miller said.
This distinction is important, and it helps explain scenarios when organizations are looking to use one or the other.
“AI is great at operating a machine that we think of in normal situations. “In predictable situations, AI is really good at keeping things going,” Davenport said.
An example is searching for keywords in long documents such as legal contracts or health records and providing results in seconds. Another monitors information from thousands or even millions of sources, whether sensors on the manufacturing floor or cybersecurity monitoring tools. These types of processes are laborious for humans but easy for machines.
“Full automation works best when you are in familiar areas and the nature of the differences and the unexpected nature are somewhat known. You can really drive for that efficiency, stability, optimization and productivity,” says Miller.
In fact, the conditions for implementing AI without human intervention are rare. That’s why Flippy needs an experienced cook to know which part of the hot spot grill cooks the burgers the fastest. That’s why a robotic weed-puller developed by agtech company Farmwise still needs a farmer walking behind it to make sure the plant it’s grabbing isn’t a vegetable. (This still beats the historical technique of spraying the entire crop with dangerous chemicals to kill weeds.)
That’s why autonomous vehicles work well in the dry air of Las Vegas or Phoenix, but not in rainy or snowy environments.
“I dread the day they come to Boston in the middle of winter,” Davenport said.
4 Situations When AI Augmentation Makes Sense
The concept of augmentation is not new. Humans started using stone tools thousands of years ago. It’s also familiar to knowledge workers who don’t need calculators on their desks.
“Anyone who’s ever used a spreadsheet understands the power of reinforcement,” Miller says.
However, the growth enabled by AI is often underestimated in today’s businesses.
“There are more opportunities than people realize,” Miller said.
Davenport and Miller provide examples that fit into four broad and overlapping categories:
- When companies want to experiment. Product development is expensive in the pharmaceutical industry, where the median R&D cost for new drugs is now north of $1 billion. Pharma manufacturers are creating AI infrastructure to sift through their data and predict potential use cases for a new drug within two weeks, rather than spending up to a year on a drug only to realize it doesn’t work, Davenport said. This leads not only to frontline workers, but also to new “data generation” roles to handle these AI use cases.
- When there is a lot of pencil pushing. “Every entrepreneur we meet and any company we talk to almost always has the same lament: ‘We have so much to do and we don’t have the numbers to do it.’ At the same time, they complain about how much grunt work they have to do,” says Miller. Under these circumstances, automation frees up employees to do more hands-on work through AI, from product development to predictive analytics.
- When things go fast. Cybersecurity monitoring software can assess threats from millions of sources in real time. Some tools prioritize threats and can respond to them automatically — for example, by shutting down a compromised device. But in “complex and nuanced cases,” Miller said, an experienced security analyst is needed to assess all the data collected and determine whether any one incident is indicative of a pattern of attacks that require mitigation on a larger scale.
- When workers want to recommend. When consumers place their trust in healthcare professionals, growth acts as a decision support. For example, AI looks at available information and makes recommendations, but the final decision rests with the doctor in front of the computer. In countries like China and Indonesia, where hundreds of millions of patients have access to AI-powered diagnostic tools like Ping An Good Doctor, “the doctor actually has to come up with a diagnosis and treatment, otherwise it’s invalid,” Davenport said. “But an AI system can save the doctor a lot of time and effort.”
How to work with AI in responsible ways
Moving forward with AI augmentation means accepting that machines are better than humans at some tasks.
Again, this concept is not new. “
We always have super tools. We have jet planes, we have cars, we have power shovels and cranes,” Miller said. “Now we have supertools that can do non-routine, cognitive-based tasks.”
At the same time, machines cannot bring all aspects of intelligence. Cornell University’s Robert J. Citing adaptive intelligence theory from Sternberg, Miller says the human mind is uniquely suited for creativity, intelligence, and context — not just realizing that something can be done, but that it should be done. In other words, don’t underestimate what humans can do.
The most effective and responsible use cases for AI augmentation are more than basic tasks or job automation. The goal should be to bring humans and machines together to achieve a level of intelligence greater than either a human or a machine can achieve on its own, Miller said. The whole must be greater than the sum of its parts.
By doing this, firms can expect to create new types of products and services as well as new types of work. But this comes with two clear caveats: Workers need the right skills to use AI systems, and they need to see that AI will help them and not replace them.
“We really need to step up and help people adapt to some of these new tools,” Davenport said. “We need to retrain workers and engage them in this new type of work.”
Watch the webinar
Read next: Gary Gensler on how AI is changing finance