On the fourth day of IMD’s Orchestrating Winning Performance program in Singapore, faculty and participants focused on technology and sustainable transformation.
In the morning, Amit Joshi, Professor of Digital Marketing and Strategy, in his session “Fastest introduction to the world of AI” made the point that artificial intelligence (AI) and machine learning are not interchangeable.
“They don’t mean the same thing,” he said.
“AI is the idea that machines can think like humans – that machines can do everything human beings can do,” he said.
He explained that it remained a fantasy. “AI can play you at chess or balance your retirement portfolio, but it cannot go and make a cup of coffee or open that door,” he said.
The focus of much research at the moment is on machine learning.
“The genius of machine learning is that instead of telling a computer how to do something, instead it is given a lot of examples,” he said.
Three different types of machine learning dominate the business landscape and research at the moment.
The first is “supervised learning” which is used to make predictions.
This requires a great deal of labelled data such as that identified in the challenge–response Captcha picture test used to determine whether a user is human.
The traffic lights or bridges that users identify are often used as data by geographic location companies.
“The bottom line is that you will need labelling for supervised learning,” he said.
The second type of learning is “unsupervised learning” for classification or categorization.
“Let’s say that I am working in the medical field and I have lots and lots of scans. I want to know if this particular scan has cancer or not,” is how Joshi explains it.
These two types of machine learning make up 90%-95% of all applications at the moment but, he points out, there are two issues with them.
The first is that a million data points are not always available, and the second is that machines are being trained based on past mistakes.
The third type of learning that is beginning to emerge is “reinforcement learning”.
“If a machine does something good, I give it a positive score, if it does something bad, it gets a negative score. Reinforcement learning is when the machine then figures out what the right thing to do is,” he said.
Brush your teeth
In his session, “Platforms and ecosystems: How to build and profit from them”, Mark Greeven, Professor of Innovation and Strategy and Chief Executive of IMD China, used toothbrushes to explain business ecosystems.
Moving from a traditional toothbrush, he showed how the technology had evolved first to the electric toothbrush and now to a toothbrush which connects to a piece of software on your mobile telephone and which assesses and manages an individual’s brushing capabilities.
This application, he said, also connects to businesses like dental insurance companies.
“The better you brush your teeth the lower your dental premium,” he said.
Greeven explained why it was that toothbrush manufacturers do not provide dental insurance citing consumer trust and the fact that it isn’t a core part of their businesses.
“You are selling a product as part of a larger solution,” he said. “All the pieces of data that are created allow you to work with a partner and to provide them with access and value. They don’t need to own this asset to manage it,” he said.
Greeven explained this was what he meant by the ecosystem, calling it “a value proposition that provides a service that jointly offers a better service”.
“A lousy product compensated by a good ecosystem would be impossible, but a good product with a lousy partner is possible,” he said, because “you can kick them out without hurting the sustainability of your system”.
The circular economy
Julia Binder, Professor of Sustainable Innovation and Business Transformation, in her session “Driving Change: How to Champion Sustainable Business Transformations”, talked about the challenges of the shift towards a circular economy.
“We produce it, somebody consumes it, and then we just put it in landfill,” she said. “We see piles of waste no matter whether this is fashion waste, food waste or electronic waste. This is a lot of wasted resources,” she said.
It isn’t just a case of encouraging people to use products longer or to help them repair them more easily, it is a case of “looking at waste not just as a waste product, but also as valuable input product”.
At every stage of the manufacturing process, there is “unexploited customer engagement”.
She ticked off unsustainable materials, underutilized capacity, premature product lives and wasted end-of-life products.
There is value at every stage of the process, although there are challenges to closing the loop.
Binder cites opaque value chains (“Oftentimes, we just don’t have enough insight into our supply chains,” she says), as well as problems created by linear product design that isn’t designed for disassembly, linear lock in an economy not designed for circularity, as well as inefficient collection and reverse logistics.
Then at the end, there are customers.
“Customers will always want to see added value,” she pointed out. “They want to have a product at a price point they can afford and they’re willing to pay a little bit of a price premium for something sustainable, but they’re not willing to forego the benefits that it offers.”