The Future Starts Now
Once the staple of sci-fi movies, artificial intelligence is shaping our present
By Tellisa Ramadhani
Tuesday, August 15, 2017
Among the existing AI applications out there are intelligent personal assistants such as Amazon’s Alexa, Apple’s Siri, Samsung’s Bixby, and Microsoft’s Cortana. AI is also being applied in the automobile industry, an example being what Tesla is doing with its self-driving car. AI is definitely gaining popularity and is ever more influencing the way we live and interact.
AmCham Indonesia spoke to the CEO of Dattabot, Regi Wahyu, on the development of AI in Indonesia. Dattabot has been in the AI industry since early 2015 and is a leading player and pioneer.
AmCham Indonesia: How do you define artificial intelligence (AI)?
Regi Wahyu: From what I have been observing, whenever I go into business meetings there are three pieces of jargon that often come up: artificial intelligence, big data, and the Internet of things [IoT]. So I think it is necessary for the players, along with business associations like AmCham and media outlets, to educate the market and manage expectations. Because big data or artificial intelligence is not a magic wand, like it will suddenly make everything so technologically advanced.
For us, AI is to make something as practical as possible; how to create a machine with intelligence so it can enable people to be more productive in their jobs. We input the machine with a training-set that will allow it to learn to process the data by itself to suit the objectives.
In Indonesia, the most applicable AI is the chatbot. Chatbots work as an assistant to cater to people’s questions, complaints, or requests. A chatbot is made by inputting the data of previous records of emails and phone calls, which is pretty much based on millions of FAQs. Indonesia has a sizable amount of data to be developed; it is not easy, but it is doable.
Aside from chatbots, is there another AI application happening in Indonesia?
The application of AI in Indonesia is not limited to chatbots. We have published this project in the legal sector this year, together with a law firm as our partner. There are thousands of regulations and laws here in Indonesia, so we created legal taxonomy of those regulations and within that taxonomy we can see the correlation from one regulation with another. This will cut working time for lawyers by hours, who in the past have done their research manually going through one regulation after another. With this project, we can also see which regulations have been annulled or amended. What is great about this is it learns and understands the vocabularies and abbreviations we often use. Sometimes we look up on a search engine an abbreviation but what comes out is not what we want, but with this machine it learns to adapt to our vocabulary. The data service team behind it is teaching the machine that these terms are actually the same and connected.
Another application that we are doing is in sales distribution. We have a map of Jakarta with the complete demography of the people and density. Each salesperson has a phone that we can use to track their route and gives them an automated route each day for where they should go based on the data assessed of sales effectiveness. So AI is now able to give advice to get a maximum result for sales. As a result of more efficient distribution, we can increase sales by 40 percent - in the past five years it had always been only 8 percent - and are able to cut costs through slashing the number of trucks needed.
We are also partnering with GE in agriculture, using their software platform Predix as a base. We are building an application on top of it to create connectivity or machine-to-machine. The objective is to get real-time information of what is happening in the field, which is automated and can be accessed on a mobile application with our satellite prototype connecting all the data to the GE platform. We are optimistic about this application but at the moment we are still in the data mining stage. This is a challenge for us because in Indonesia we mostly have offline data, so we have a big job to convert this data to online records.
AI needs to be built step-by-step. We first need to make sure that the data is out there and there is enough of it, and after that we can create a technical model and recommendation engine – in the end the machine can keep learning from it.
The business sectors that have already applied AI are e-commerce and finance/banking. These are the kinds of industries that gather a lot of data in their business. Supposedly, the telecommunications industry should be able to adopt AI already.
From what I have seen, consumer goods and retail have not yet incorporated AI.
We see that there will be an industrial revolution, specifically the industrial Internet. Development of this is happening to allow the machine to “speak” to us with its predictions on which spare part must be replaced before it breaks down. This could reduce the cost of maintenance.
What do you think of people’s awareness of AI in Indonesia?
People do not often realize that we are actually using AI. We have plenty of Facebook users in Indonesia and Facebook is incorporating AI into its platform, for example, with automatic tagging and face recognition.
We are now in the third wave of innovation, the incorporation of the two technologies in the first and second wave, which are the foundation and application to go to the real sector. What is happening is we now don’t only need people, technology, and platforms, but also we require policy. We need to ensure that innovation and advocacy for policy goes hand-in-hand and cooperate with the government.
We really appreciate what the government has been doing, because there are already so many initiatives out there. We are part of the Indonesia Fintech Association and the Indonesia IoT Forum, and through them we regularly update and discuss with the government. We are the only company in the Fintech Association who is actually not working in fintech, because we support the growth of fintech.
The way I see it, with the current government, it foresees the future and supports initiative, which is complemented by the regulations it has or is drafting.
How do you see AI being applied in the financial and fintech sectors?
We can look up to the countries whose fintech industries are already very mature, such as China and the US. China is so advanced in its fintech because of the population, creativity and freedom to develop the business. The market in the US has a high financial literacy and is aware of its rights and obligations. With the fintech trend, it will make each individual more responsible and have a higher awareness of their own financial situation.
That kind of impact is what we need for our people. We need to build a system for the people to have better financial literacy and, in this digital era, it will help Indonesia’s economy as a whole. And there needs to be collaboration between players. We have seen that big banks collaborate with start-ups, which is good.
I think Indonesia needs to support the players on data, just like what they do in China to create a better ecosystem.
What do you think of the future of AI in Indonesia?
I think Indonesia is ready for the future. We are living in a parallel world, so what is happening out there will be happening here as well, it is one call away.
I had a discussion on how Indonesia can speed up its innovation culture. There are two answers: one is if we can grow the entrepreneurial spirit, that the number of entrepreneurs within one mile is a lot so that innovation happens many times due to the competition for survival, that is what is happening in the [San Francisco] Bay Area. The second answer is the failure culture, so when we have a failure we have got to embrace it and move forward. When there is a failure, people should not be ashamed so that they end up not sharing their failure and this disables other people from learning from that failure.
In Dattabot, I encourage the people here to celebrate our failures and I never call them employees, they are my future partners. The way we develop our human resources is based on what they dream of so they can grow as much as they can. I believe innovation is unlimited.
There is still not much talent for AI in Indonesia because the opportunities to experiment and to learn and fail are not that many. Most companies will play it safe when it comes to testing something new. Creating AI is not going to be successful from the start, to build the learning measurement it needs lots of trial and error.
But I am confident with the raw material that we have, because without that Dattabot would not have been able to survive to this moment. Our office offers some lessons and sharing moments on AI and we welcome anyone who would like that because we are both learning from one another.
There is a big opportunity for AI to grow faster in Indonesia. I’d say that within three to five years more people will be adopting AI, and I’m optimistic about this because I work in the business. The challenge will be how the big companies, such as most of the members of AmCham, will collaborate with small start-ups and create like an innovation garage. There is a lot of potential, not just to work as a vendor but to collaborate to open a new market and business model.