Artificial intelligence, human choices

“There are no bad students, only bad teachers”. A saying that’s perhaps not always true, but perfect for understanding the workings of Artificial Intelligence, something that for the public at large is still a mysterious subject and (for some) a bit worrisome. «Fearing AI and machine learning would be a mistakesays Fabio Moioli, director of the Enterprise Services Division of Microsoft Italia –. If artificial intelligence sometimes makes huge mistakes, for example when it analyses CVs using non-inclusive criteria, it’s not the fault of AI, but that of programmers who were responsible for training and probably didn’t consider certain variables. A human error, therefore, just like that which occurs in many other fundamental sectors».

So, not a racist AI, nor one that has a subconscious or will of its own, but a tool that instead should be used, taking into account its immense potential. «For this reason, it’s good – explains Moioli – to ask ourselves about the possible risks tied to the improper use of AI, as does Elon Musk, for example. Think of the impact atomic fission or gunpowder had on the world: there are aspects to which we must pay the maximum attention, like privacy, transparency, security, inclusiveness».

Companies know that most people, when they think of AI, often rely on imaginative, almost “apocalyptic” notions. Partially to contrast this trend, many companies directly involved in the AI sector, starting with Microsoft, «have internally set up ethical committees, disconnected from any evaluation of profit or marketing, which critically analyse and in many cases reject projects considered to be at risk. A theme that also interests our client companies a lot», adds Moioli.
And companies themselves are directly called on to evaluate and tap into the potential offered by AI: «It’s a  pervasive technology, that I’d call general purpose, like electricity – explains Moioli –. It can be used in any process: in interaction with clients, in the personalization of services, in the processing of products. But it can also revolutionise productive strategies, helping people to work better. Advantages that apply both to the blue-collar worker and to the engineer».

Advantages, above all, that thanks to recent developments can also be exploited by small and medium enterprises. Usually these firms can’t afford a team of data scientists, nor do they have the vast amounts of information available to larger companies. However, the situation is changing rapidly: «The fastest growing trend is that of AI that can learn more, but with the use of less data. Furthermore, increasingly popular are AI-based preconfigured cognitive service libraries, ready-to-use services (some examples: automatic translations, facial recognition, chatbots) that are highly customizable based on the needs of each individual entrepreneur. The other advantage is that in this case we don’t need real data science experts to programme everything from scratch, but easier to find and less specialized professionals are enough, such as software developers».

There’s no doubt, then, that technicians are needed. But Moioli also offers a suggestion to those who work or will work in corporate functions apparently not involved in this process of change: «Whether dealing with marketing, human resources or something else, a manager must always be aware of the potential offered by AI. He or she must know that certain tools exist and that they can improve his or her work». Thus, a widespread awareness that can’t be separated from training and education, at all levels. «In Italy there are a number of examples of sectoral excellence. FLEXA, the personalised and continuous learning digital platform of Politecnico di Milano’s School of Management, itself was designated by Microsoft as one of the most innovative projects in the world. Now, however, a lot of work also needs to be done in primary and secondary schools: in the future we’ll need professionals who speak the language of AI with increasing familiarity. This is the real priority».





What Jobs Will Survive Artificial Intelligence?

“We want to bring intelligence to everything, to everywhere, and for everyone”. The person who said it is Satya Nadella, CEO of Microsoft, the company that recently launched the “Zo” chatbot that can construct sophisticated human-machine dialogues. And it’s through Microsoft artificial intelligence (AI) instruments that the School of Management of the Politecnico di Milano developed FLEXA, an innovative and revolutionary digital platform of personalized and continuous learning, a digital mentor able to identify and select specific contents, useful for the path of study of each user.

«This project, but the idea applies to Artificial Intelligence in general, began from an awareness: we had identified certain needs and technology could help us to satisfy them – says Federico Frattini, Associate Dean of Digital Transformation at MIP Politecnico di Milano and creator of FLEXA –. Specifically, our Masters students wanted to know each other better, also from a comparative point of view, to then pursue ad hoc training programmes, while our alumni, our former students, asked us for effective continuous learning solutions. We reasoned on the basis of these inputs and the result was FLEXA: on this platform it’s possible to carry out an assessment of one’s soft, hard and digital skills and declare one’s career ambitions; once all this information is processed, FLEXA provides all the indications to fill these skill gaps through events, courses and training programmes on the basis of the needs indicated. And it doesn’t only provide contents from Politecnico di Milano: we have agreements with Gartner, New York Times, Financial Times, MIT and many other prestigious groups. With FLEXA, it will also be possible to have a mentor recommended, create a matching system with startups and companies, create new contents».

One of the points on which the debate is understandably most heated involves the impact artificial intelligence will have on employment. In addition to more mechanical and repetitive jobs, such as those carried out by assembly line workers and, as already mentioned, driving cars and some activities in restaurants and supermarkets, automation is also entering the field of services. According to some studies, for example, by 2030 there will no longer be “human” call centres, while in Japan many robots are already operational in assisting the elderly.

On the other hand, artificial intelligence has limits that in many cases prevent it from substituting human jobs. At the same time, the role technology can play in supporting and strengthening humans in carrying out some higher-skill activities may be underrated. The renowned Chinese artificial expert Kai-Fu Lee, a businessman involved in this very sector and the author of the recent book AI Superpowers: China, Silicon Valley, and the New World Order, identifies four weaknesses of AI for work performance:

  1. AI cannot create, conceptualize or manage complex strategic planning;
  2. AI cannot accomplish complex work that requires precise hand-eye coordination;
  3. AI cannot deal with unknown and unstructured spaces, especially ones that it hasn’t observed;
  4. AI cannot, unlike humans, feel or interact with empathy and compassion; therefore, it is unlikely that humans would opt for interacting with an apathetic robot for traditional communication services.

Given this premise, Kai-Fu Lee draws up a list of ten professions that will be immune from the robotic invasion, at least in the next 15 years: psychiatry, physical therapy, medicine, AI-related research and engineering, fiction writing, teaching, law, computer science and engineering, science, and management. In all these professions AI can be of help, but only in a collaborative sense for the management of certain technical details.

 “There’s no doubt that the AI revolution will require readjustments and a great deal of sacrifice” affirms Kai-Fu Lee, “but despairing rather than preparing for what’s come is unproductive and, perhaps, even reckless”.

And then he concludes: “We must remember that our human knack for compassion and empathy is going to be a valuable asset in the future workforce, and that jobs hinged on care, creativity, and education, will remain vital to our society”.

«I believe that the best way to approach artificial intelligence is to tie it in with theories explaining innovation and entrepreneurship – says Federico Frattini –. We can define it as an accelerator of processes of creative destruction determined by digital innovation, taking inspiration from what the Austrian economist Joseph Schumpeter said about the big changes that had an impact on the economy and on society: new opportunities are created, new companies and new professions are created, others evolve and still others, inevitably, disappear. We certainly can’t oppose the creative forces that have changed society over the centuries».


The future of advanced learning will be FLEXAble with the help of AI

Another digital revolution is under way at Politecnico di Milano Graduate School of Business (MIP) with the arrival of the first release of FLEXA, an innovative, personalized continuous-learning platform and a digital mentor for students. MIP is using Microsoft AI services to power this platform that delivers personalized education recommendations to the next generation of executives and business decision makers.
Imagine a world in which learning was invented, for the first time, within today’s digital media environment. There would be no history of print influencing how courses were planned or textbooks were sold, no centuries of structure shaping how universities were designed. In such a thought experiment, we might reimagine how classes look, how long they last, and how they structure the information and interactions shared by professor and students. Instead, we might see the standard quarter or semester structure replaced by longer, more adaptive learning strategies that happen daily, weekly, or monthly, as determined by the needs of the student and the requirements of their chosen career.

A new learning model would also need to leverage a powerful content engine and artificial intelligence (AI) capable of adapting and assisting each individual student. Such is the dream of FLEXA, an approach to learning being pioneered at Politecnico di Milano Graduate School of Business (MIP) in Milan, Italy.

Customized and personal

Planning for FLEXA began in 2016 when Federico Frattini, professor at MIP and director of the MBA and Executive MBA programs, began thinking about the next wave of digital technologies that could advance executive learning. In doing so, he realized that some unanswered questions remained about how to most effectively use digital technology in education. First, Frattini notes, there is an “opportunity to personalize the journey for our students undergoing long programs like an MBA, Executive MBA, or a specialist Master’s degree. Beyond in-program personalization, digital technology also offers the possibility for providing continuous learning over time for our alumni community, and that’s something shared by many other schools.” FLEXA offers students many features, including networking opportunities and exposure to companies registered on the platform that can promote talent acquisition initiatives. “FLEXA is a great tool for becoming more cognizant with the needs of our students,” says Frattini.

The crux of FLEXA begins, Frattini explains, with “a diagnostic tool that evaluates the student’s hard, soft, and digital skills, then uses Microsoft AI services to create personalized learning pathways that can be very short one-day paths or longer six-month paths. It can suggest physical and digital content customized to the experiences of each student, alumnus, or prospective students, designed to close their skills gaps.”

It works like this: Students sign in to the system using their MIP accounts, whereupon they take the assessment of their various skills—hard, soft, and digital. “The hard skills section,” Frattini says, “is a self-assessment based on the experience they have gained in their career or studies, covering a number of dimensions of competencies that are relevant for tomorrow’s jobs. Then they will take tests for the soft skills and digital skills. They will be asked to indicate where they expect to be in their career in three years’ time. So, in terms of financial expectations, they will be working on the kind of role they would like to take on.”

From there, students can access their dashboard where they can find their skills profile and see their strengths and their gaps, all depending on the career aspirations they have previously identified. Students determine how much time they’d like to spend to deepen their knowledge surrounding those areas, and FLEXA leads them along the path to filling their skills gap efficiently. Frattini explains, “Imagine having a personal mentor supporting you day-by-day along your career path. Imagine that this mentor were to recommend the right content in order for you to keep up-to-date and be able to close the gap between your current skills and those needed to achieve your career goals. This is FLEXA.”

Working with Microsoft

This is where Microsoft AI comes in. For any given combination of hard, soft, and digital skills, and for the specific aspirations of any individual student, there is a wealth of content that could be useful. Some of these items can be read in 10 minutes, some in 20, and some require more time. The AI can recognize and categorize the content appropriately and make recommendations accordingly. FLEXA might recommend attending an advanced marketing course at MIP, then follow with recommendations for three specific book chapters because they touch on a particular skills gap. These recommendations evolve over time. Frattini notes that, “FLEXA takes into account the feedback of other users on the platform that are using the same content. It also considers how relevant the content is to a student based on other profiles similar to that student’s in terms of overall career aspirations.”

For such an ambitious goal, the cooperation and technology of Microsoft were critical. Frattini is clear that, “there was no other partner with whom we could have done a project like this. The vision, the passion, the support of Microsoft Italy—the relationship was very positive, and fun.”

The possibilities with such a system are incredibly exciting. It can be integrated into curriculum goals. Frattini offers an example: “Say I’m a professor interested in the applications of blockchain. I can launch a challenge on FLEXA, whereby I ask three teams of five people to prepare a video sharing their experience and knowledge about how to use blockchain in the food sector. The three teams are created on FLEXA. They take part in the game and the team producing the best video will be selected by the faculty. We have the video included in FLEXA’s datalink and it can be sent to other users.”

At the same time, given the assessment tool, FLEXA provides a granularity of information regarding the actual needs of students, and those needs can be communicated to faculty in order for them to optimize course and lesson plans to best benefit their students.

Into the future

The FLEXA pilot launched in the last quarter of 2018, and already the team at MIP has ideas about how to evolve its operations and develop future iterations of the platform. Reimagining how to serve students in a digital age takes time, after all, and a willingness to innovate and take risks, and to use new technology to rethink old methods. The ingenuity on display by the MIP team and the power of Microsoft AI makes FLEXA a project to watch, as these sorts of experiments may very well define the next decade of curriculum planning and educational excellence.