Scott Hunter, cloud specialist at Google, speaks about digital divides between countries, socio-economic risks stemming from AI use, and the big question of data privacy
In the first part of this conversation with Unravel, Scott Hunter, cloud value advisor at Google, talked about the concept of ‘AI for good’, and the growing role of AI in food security, healthcare and supply chains. We also looked at the role of AI in the development of emerging markets and in taking financial services to the unbanked.
In this second part, Mr Hunter speaks with Siddharth Poddar and Shivaji Bagchi of Unravel about how societies are coping with changes brought about by AI, the data privacy dilemma and emerging disruptive technologies in the light of COVID-19.
Unravel: Do you think AI could further the divide that already exists between technologically advanced countries and others, or narrow it?
Scott Hunter: I actually think it narrows the divide if done correctly. Countries like Singapore have done a fabulous job about thinking through this whole issue of what AI and ML can do. How do I retrain my workforce and move them into better paying jobs? How do I move them through society? It is about how technology companies apply AI and ML.
I also think it has to do with the government – about how it trains workers along with refreshing and retraining them such that they can go out into the workplace in a different capacity. How will government efforts have them go back into the workplace in a different capacity? This is especially true in manufacturing. When I go to factory floors now, I see a lot of robotics. Therefore, it is necessary to retrain and reskill workers, and direct them towards better paying jobs. It really is a balance between the commercial side, education and the government.
Like I explained in our chat earlier, if AI is applied correctly in developing economies, it can uplift farmers and fishermen and enable them to enjoy better prices at the marketplace. Likewise, it can benefit an underserved population enjoying banking or insurance capabilities. In these ways, it helps narrow the divide.
It is a thorny issue though, and it needs to be addressed by more than just technology companies.
Unravel: One of the things we often hear about is how developing economies are well-placed because they can leapfrog and move to more advanced technologies directly. Do you think that logic works within the AI and ML space as well?
Mr Hunter: Let’s take a step back and examine how we provided internet to the society. In China, it was actually developed with the military. In Japan, the government ensured there was fibre optics in every house and that is how internet was provided to families. In the US, it started off as an expensive commercial model. It then got cheaper and now it is at a price point where most can enjoy it, and there is even free internet in many cases. I think it will differ by country and society.
Tech adoption depends on where a particular country is in its technology cycle and what policy decisions the government takes that drives the next trend. Certain industries will embrace AI faster than others.
About your question about how we leapfrog? There are other things that have happened in technology. Let’s take the telephone, for example. In Europe, the Middle East or North America, for instance, there was considerable landline infrastructure. However, Asia and Africa didn’t have such infrastructure, so they went straight into mobile and leapfrogged other parts of the world in many ways, using mobile technology.
It therefore depends on where a particular country is in its technology cycle and what policy decisions the government takes that drives the next trend. Certain industries will embrace AI faster than others. Call centres, for instance, are applying AI to perform customer service roles by using chat bots to actually provide a better customer experience.
Unravel: In what ways do you think AI is challenging traditional businesses?
Mr Hunter: The change will continue, and at a very rapid pace. Let me give you an example. I got to deal with Minolta, the camera company. They realised that film cameras were starting to disappear, and they bought Konica and renamed themselves Konica Minolta. They realised this acquisition was not enough. Then they started making printers, only to discover that it was also a dying industry. They changed track to specialise in printer services and made better margins too. But they realised that other players were coming into the printer service market. It’s simple – they said they understand about giving service, so let’s figure out what we can do to change or transform ourselves so that we have a path forward.
Minolta Konica continued to change and transform itself. It went on to make over 50 acquisitions, and in a span of about six years, it went on to become the largest robotics service provider globally. From a camera company, they evolved into a robotics services company. And the thing is, if they hadn’t transformed, they would have died.
It is similar with AI — if you don’t transform, you will become obsolete. It used to take 20 years for a company to become worth a billion dollars. Now we’re seeing companies become worth a billion dollars in a matter of a year. So, you’re watching the Fortune 500 change significantly and companies dropping off. Even for a Fortune 500 company, if they do not transform, about 50% of them will not feature on that list within a decade’s time. And that space will be filled by someone else, someone who has transformed and embraced technology.
Another very exciting development is that half the new Fortune 500 companies by 2030 will be from Asia. There is this massive movement because of technology and the shift of wealth from the West to the East. So that’s another area you have to look at – about the speed or shift and then the adoption of technology and AI.
If you don’t transform, you will become obsolete. Even for a Fortune 500 company, if they do not transform, about 50% of them will not feature on that list within a decade’s time. And that space will be filled by someone else, someone who has transformed and embraced technology.
Unravel: What are some of the broader socio-economic risks relating to AI in emerging markets?
Mr Hunter: AI always brings risks with respect to automation. If governments or companies aren’t ready to retrain workers, then workers have an issue finding jobs at equal or better pay. Five years ago, I was speaking at a China Entrepreneur Society event in China. We were shown a short film on a fish processing plant. It showed a conveyor belt with people wearing sanitisation gear on either side with fillet knives in hand. You’d say, “Ok, that seems pretty interesting!” Then it showed what their new factory looked like. First, they had to go turn the lights on. The first thing I noticed is the conveyor belts were moving about three times faster than it was before. I then realised there were AI and camera driven water knives that were enacting the same role as physical workers. There was not one human being standing on that production line. So I realised that a massive change in technology can displace the jobs of many people, if there’s not a great culture to reskill them. We have to have empathy and compassion with employees – they must be trained and elderly employees will need more training. We must think this through more closely.
Unravel: There’s so much data that is being collected by companies and governments. Do you consider this collection of data a big risk?
Mr Hunter: You have to divide data into separate sections – personal information, and data from business. In the case of business data, you typically have an IT side such as ERP systems and then you have an operational technology (OT) side such as factory floor and operations. Over the years, these two worlds never aligned. On the OT side, a lot of the data was called dark data as it was never used. Now we are starting to see this data being harvested; it is being brought up into a data lake to run analytics or AI across the top of it. There are tremendous efficiencies yielded across the supply chain or an operations side as a result of this. So that data, I think, is good to collect to actually use. It really drives efficiency.
Then there is data that is easily collected through smartphones. These could be in the form of navigational aids such as maps or there are certain applications that people derive great joy in using, and this usage data helps in providing a better experience to the consumer.
However, data may be collected without people being aware of it, or maybe used for certain government processes. In this regard, the rules and regulations have to be brought up to speed to meet these requirements. There are now strong rules in Europe and the US with respect to data privacy. They have set an example for the rest of the world to follow on how we want to treat data. When it comes to personal data, we have to be super cautious and careful about how it is used and shared. So the data question, I think, has to be broken down into different categories in order to answer it completely.
On the operational technology side, a lot of the data was called dark data as it was never used. Now we are starting to see this data being harvested; it is being brought up into a data lake to run analytics or AI across the top of it. So that data, I think, is good to collect to actually use. It really drives efficiency.
The other kind of data can be very useful. Some of these manufacturing facilities have been collecting data off the manufacturing machines for 30 years and they haven’t done anything about that. So you start to wonder what’s there. People don’t understand how much dark data there is and this can be very useful. Separately, certain governments collect unbelievable amounts of data about temperatures or air quality, for instance, and these datasets can be very useful.
Unravel: What technologies do you think are going to be particularly disruptive over the next decade or so from an economic development perspective?
Mr Hunter: I do think the cloud will continue to be disruptive as companies will keep moving to the cloud. A lot of AI and ML apps will run on the cloud. In the year 2000 or thereabouts, a lot of what was talked about Industry 4.0 was in theory. But over the past 15-20 years, the costs of computers, IoT devices and storage have all declined – basically, the cost of computing has dropped significantly. And in this while, the speed of computing has increased astronomically. We are starting to see step changes about what is happening in Industry 4.0 and how they are moving from “on-premise” to the cloud.
We’re seeing this seismic shift in Industry 4.0 and the new capabilities and technology have really made a huge difference. I think that will continue to accelerate. We’ll begin to see things with smart cities, smart cars, self-driving vehicles and such innovations.
I am also thinking about quantum computing. Google’s made a breakthrough on that. I think you’ll see more and more of that type of research continue.
Unravel: What are the shifts we are seeing due to COVID-19?
Mr Hunter: Since COVID-19, I have been closely observing this whole thing about e-commerce and fulfilment. We’re seeing more and more customers buy their groceries online, have them delivered and it’s actually started a completely new sub-vertical about fulfilment. A lot of people talk about the shift in consumer behaviour and willingness to buy online, but not about the fulfilment element of e-commerce. This is a trend that has taken off because of COVID-19.
And then I guess the final thing that I will leave you with is that I have never seen as much activity with respect to companies and executives talking about digital transformation as I have seen since the onset of COVID-19. It’s going to become the hottest topic for business transformation and continuity.
The first part of the conversation can be read here.
Scott and his team brings a disciplined approach to support the larger value goals of a business. Over the past 30 years, Scott has developed and delivered best practices in domains such as big data, data analytics, SaaS, enterprise architecture, value demonstration, and security policies and standards to name a few. Scott has deep domain expertise in discrete and process manufacturing, oil & gas, utilities, mining and precision agriculture. He has in-depth solution knowledge of supply chain management, industry 4.0, IoT, AI, machine learning, predictive analytics and digital transformation. Prior to his tenure at Google, Scott worked for Microsoft, SAP, the Oracle Corporation, i2 Technologies, Booz Allen Hamilton, Cygnus Global and the US Army.