16 Dec 2019
SIM Chief Learning Officer Dr Kevyn Yong went on radio station MONEY FM 89.3 recently to discuss about artificial intelligence (AI) with presenter Michelle Martin in her radio show ‘Influence’. This popular programme features thought leaders on their views over a wide range of topics.
Read on (excerpts and edited) to learn how AI can help enterprises innovate and create value-added services and products, and why companies should use AI to improve consumers' lives.
Michelle: What do you think AI can do to help companies build disruptive innovation?
Dr Yong: AI is a tool for innovation, and companies can leverage this technology to make customers’ lives simpler and better.
Let’s take Spotify for example – it has gone exponential and has made people’s lives simpler. Research based on Spotify data found that people stopped discovering new music around 33 years old, whereas Spotify’s competitor Deezer also did a study and found that people stopped discovering new music around 30 years old. So Spotify looks at your playlist and matches it with somebody else’s. The idea is that if two persons’ playlists overlap on 70-80% of the songs, chances are that they have the same taste in music. Depending on how I respond to these new songs that are recommended to me, Spotify then updates the algorithms to incorporate my preferences.
Michelle: How can AI help businesses find the right problems?
Dr Yong: What we want is to look for unexpected patterns in the data generated by AI to help us find new opportunities to innovate for customers.
One simple example – Management guru Clayton Christensen developed the “Jobs to be Done” theory of innovation (affectionally known as the Milkshake Theory). He sent his research associates to go hang out at a restaurant to collect data and they found that around 40% of milkshake sales occurred before 9am. They realised the types of people who drank milkshake before 9am included those who drove to work, those who commuted to work by train, and those who did not have time for morning breaks. The milkshake thus served as their breakfast. It served a purpose and fulfilled a job. It made their lives simpler. This insight came out from data. AI can be used to collect all that data through the transactions in the restaurant, process the data, and provide the patterns.
Michelle: Can AI help us make sense of the data or it just shows up the patterns?
Dr Yong: AI shows up the patterns; it does not help us make sense of them. AI doesn’t tell us whether the data is good or not. It’s kind of objective in that sense.
Another example – James Scott, author of the book AIQ, was touring the town of Ypres in Belgium where classical music was playing through outdoor speakers, and he thought he heard Justin Timberlake’s Sexy Back played amongst the classical music. Out of curiosity, he enquired at the tourism office about the music playlist being used, and was told that the tourism office was simply using Spotify. It seems that Spotify’s AI ‘thinks’ that listeners of classical music will enjoy Justin Timberlake’s Sexy Back just as much!
Michelle: What are some common myths about AI you want to debunk?
Dr Yong: That women are terrible at AI. In fact, I want to start a new myth – the reason why there are not enough women in AI is because women didn’t grow up playing computer games. Women, like guys, can do AI. This is because we have all done mathematics. In fact, the very first person to build a working AI is a lady called Grace Hopper! In fact, she became the first human ever to “speak” to a computer by developing the very first ‘compiler’ that uses only English words. In so doing, computer programmers could now use English to programme computers.
Michelle: As AI gets even more popular, what would you say to the enterprises out there about staying relevant?
Dr Yong: Get back to the basics. Always understand your customers and your users. Don’t get caught up with the technology. Don’t let the technology drive the innovation. Let the innovation drive the technology: Think about where and how you can create value for your customers and find the right technology to help you achieve this goal.
Listen to the full interview here: Influence: Why the extent of how much AI can be leveraged depends on the user behind it