By Samuel Irungu
Google finds answers for your queries, Amazon know your preferences, Facebook not only knows your friends but also can help you find the perfect partner. These platforms seems to know what we are thinking almost before we do. Our world has taken on a digital smartness through AI, data, natural language processing, automation, and robots that, although nearly invisible, impact much of what we do.
This digital smartness is projected to have a massive influence on the world economy, adding $15.7 trillion to global GDP by 2030. It will increase productivity and wages, allowing individuals to purchase more and/ or better products. Automation driven by AI and robotics, is estimated to require the reskilling of the work force.
If smart machines are having such impact on the economy and our profession, what will they mean to higher education? For example, could a chat box be your next Teaching Assistant?
At Beckett University in the United Kingdom, chatbots help prospective students find available courses for study. Georgia State Universiry (GSU) uses an AI chatbots to respond to questions on enrollment and financial aid, handling peak volumes of as many as 2000 calls per day, with 200,000 questions answered. In a situation where the system is less than 95% confident of an answer, the query is passed on to a staff member. The impact goes beyond handling call volumes - GSU estimates that the timely responses to questions helped reduce “summer melt” (i.e., the loss of students who are admitted but not yet registered) by 20%.
Deakin University in Australia has created a platform - Genie - that combine chatbots, AI, voice recognition, and a productive analytics engine to create an intelligent virtual assistant that provides students with advice. They are being tested as English tutors.
As the world around us is getting smarter, what does it mean to be a professional?
The smart machines around us
These increasingly capable systems not only retrieve and present information more quickly and accurately but also solve problems and offer advice. Machine learning allow computers to consume information such as medical records, financial data, purchases and social media and then develop predictions or recommendations. Today’s AI uses brute force computing, enabled by massive amount of data, memory and processing power. Beyond processing instructions at incredible speed, these machines can create their own guidelines and discover patterns invisible to humans.
In healthcare, AI allows IBMs Watson, to aggregate clinical guidelines, medical literature and patient data to help physicians diagnose and treat cancer. AI and imaging software can speed up the diagnosis and treat treatment strokes. Robots allow surgeons to perform precision surgery. Robotic prosthetics and exoskeletons help amputees and those with impaired mobility, self-driving cars and trucks promise to make transportation more efficient.
And that’s not all, today’s robots interacts with the physical world. Robotic sensing gives machines the ability to ‘hear’ through signal processing, ‘see’ through image processing, and ‘touch’ through pressure and pattern processing. This generation of robots can detect and express emotions. Social companion technology, in which a machine displays empathy, is being explored for the elderly to help to combat loneliness as well as monitor wellness. These part robot, part – AI system use animatronic gestures and ‘speak’, providing information, reminders and support as they adapt to and learn from their human companions.
If smart machines can take on all these human tasks, what does that mean for people? Will we need to know or do less –or more? And with this large impact on the economy and the work force, what will they mean for higher education? Rather than replacing people, smart machines augment human capabilities, meaning that we need to learn to work with machines as partners. Changes in our professions are becoming more rapid, suggesting that the way we develop and find expertise will change as well.
Augmenting Human Expertise
AI and robotics have catalyzed a wave of automation – based on artificial cognition, cheap sensors, machine learning, and distributed smarts – that will touch virtually all jobs, from manual labor to knowledge work. However, automation may be a less apt term than augmentation. As Garry Kasparov, former world chess champion, has observed: “Humans are not being replaced by AI, we are being promoted. Machine-generated insights add to ours, extending our intelligence in the way a telescope extends our vision. Think of AI as augmented intelligence. Our increasingly intelligent machines are making us smarter.”
As machines can do more, professional roles shift. New tasks take the place for the ones that were automated. Historically, new technologies have spurred the creation of more jobs than they have destroyed.
Whether it is AI, robotics, or another technology, today’s machines can work alongside professionals as partners, amplifying human performance and augmenting human intelligence.
“Knowledge processing” – something much more sophisticated than information retrieval – is an example of a new approach to professional work. Today’s systems can capture and reuse massive amount of information allowing a computer to compare a patient’s symptoms against a database of millions of past patients.
In law, intelligent search systems can outperform junior lawyers and paralegals in reviewing large sets of documents. Court decisions can be predicted by tapping databases of hundreds of thousands of past cases. Machines can consume vast quantities of information, discern patterns, and predictions that allow professionals to work in different ways.
Scientific research is an example of how the work of a higher education professional can change. Data intensive science and computational science have augmented the traditions of theoretical or experimental research. Today, AI and automated hypothesis generation platforms are used to mine scientific literature and formulate hypotheses to help researchers focus their laboratory resources in areas that are most promising. For example, Baylor College of Medicine, used IBM’s Watson to design a Knowledge Integration Toolkit (KnIT). One test KnIT focused on the functional properties of p53, a protein that is important in tumor suppression. At the time of the test, there were about 70,000 scientific articles involving p53. Humans can consume about 1 to 5 scientific articles a day, so it would have taken a researcher approximately 38 years at best (assuming 5 articles consumed in every single day) to assimilate the existing research. In one month, KnIT successfully helped researchers identify 6 proteins kinases that phosphorylate had been found in the prior 30 years. Following the many steps in the R&D process –observation, hypothesis generation, experimental design, and result analysis, AI can provide insights that augment human capability, increase efficiency, and improve outcomes.
More Accessible Expertise
Smart machines can perform faster and more accurately than humans, but they don’t necessarily use the same processes. For example, when there is an unresolvable dispute between two parties, they can go to court which is too time consuming and expensive to be viable for low-level claims such as are common in online commerce. Rather than sending dispute through the courts, ebay resolves an estimated 60 million disputes per year using Online Dispute Resolution (ODR). One approach to ODR involves a three-round blind bidding system that matches plaintiffs’ demands with offers from defendants. If the offers are close, the system splits the difference between the bids and declare a settlement. Many disputes are resolved in the first round. In New York City, ODR has been used for personal injury claims, with 66 per cent of the claims settled within thirty days, saving a lot in litigation on claims. The U.K. has explored ODR and Internet-based court services as future options because the current judicial system is too slow, costly, and complicated, making it inaccessible and unaffordable for many people.
The importance is not that machines can do things differently – it is that people can benefit from the outcomes. Society profits from sharing of expertise, not just from the legal fields but also in areas such as health care, education, business, architecture, agriculture, and engineering. Globally, there is a huge unmet demand for this experise.
We tend to think of professional work as being conducted by experts – people who hold degrees certifying their expertise and whose practices are defined by their profession. However, large numbers of people making small contributions have the power to impact scientific advances, social movements, product innovation, fund-raising and more.
Online innovation platforms capitalize a lot on ‘collective intelligence’ encouraging more people, enthusiastic volunteers to become involved in solving problems. The ideas and insights gained through increased cognitive diversity can spark new ideas.
“Communities of experience” which tap the experiences of laypeople to advance a profession, are another form of collective intelligence. A good example is PatientsLikeMe - a social networking site for patients who suffer from rare and chronic diseases. The platform does much more than provide moral support for patients and their families. Over 600,000 people report on their experiences with 2,800 conditions. The platform aggregates and organizes more than 43 million member data points with clinicians, pharmaceutical companies and other institutions enabling research and innovation. Using a give-data, get-data philosophy, patients are helped to find new treatments and connect with others.
Challenges for Higher Education
AI and other technologies will find their place in higher education but greater challenge is to anticipate what it means to be a knowledge worker in a world of smart machines. Changes brought about by AI and robots are taking place in the professions faster than they are in higher education. Without a close connection to business and industry, higher education will be challenged to anticipate the changes in our disciplines and professions.
END OF PART I.