Sufficient data and open technological development are required to grow Artificial Intelligence and its branches in Indonesia
Artificial Intelligence (AI) has become a hot topic in the recent years, with Indonesia being one of the countries that has started to take a look at it. I got the opportunity to sit down with Adrian Li (Convergence Ventures), Irzan Raditya (YesBoss), and Jim Geovedi (computer scientist, machine learning practitioner) to discuss the future of AI in the country, especially for bot technology which has been gaining traction through chatbot services.
Artificial intelligence, bot, and the things they are capable of
Referring to the Oxford Dictionary, AI is being defined as the “theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.”
Geovedi explains that AI itself is synonymous with the word ‘robot’ (often abbreviated into ‘bot’) which is commonly used nowadays to refer to mechanical beings or virtual computer programmes (software agent). Furthermore, the term ‘bot’ is often combined with other terms to create a new terminology (portmanteau), such as chatbot or botnet.
“Basically, AI and bot use advanced computation to take over human responses and simple tasks. For example, [to play] games that involve the use of strategy such as chess, go, and interaction simulation. The real AI is capable of learning through interaction and develop a sense of self-awareness. However, on a practical sense, today AI is being used to automate basic tasks such as scheduling meetings or processing Natural Language Processing (NLP) through chatting bot,” Li says.
“The AI that many are currently developing is more towards the direction of NLP, though [the technology] itself has a wide scope … While chatbot is the product here, it does not have to take a [form of] AI … AI is present as the brain that enables conversation, as similar as possible to a conversation with human,” Raditya adds.
Though in the past 50 years AI had experienced rapid progress, it had also experienced a downturn during its development in the 1970s. Geovedi reveals that this condition is nicknamed the “AI Winter” among developer communities. It took a whole decade for the technology to be able to rise again after several commercial initiative with realistic, less ambitious targets compared to the previous attempts.
“Up until now, the real AI is not yet available. However, the answer to the big question of ‘What can they contribute to our life?’ can be seen through the implementation of AI [programmes] … such as the SIRI programme as developed by Apple, […] friend or post recommendation on Facebook, the ‘OK, Google’ service […] or autonomous cars as developed by Tesla, Google, or Baidu,” Geovedi explains.
Also Read: This artificial intelligence funding program wants to put Asia’s AI startups on the map
According to Geovedi, several fellow scientists and AI developers stated that the momentum for the return of research in this field began in a workshop on “Deep Learning for Speech Recognition and Related Applications,” held in conjunction with NIPS annual conference in 2009. Apart from that, progresses in software research, hardware development, and the increasingly affordable cost of big data processing also contributed to the rise of AI.
In Indonesia itself, chatbots (which are identical to AI) are experiencing a rise of popularity, and is considered to have a strong business potential. Li explains that one of the reasons why chatbots are starting to turn heads is its simple and easy-to-use user interface (UI), which can be used by Internet or e-commerce newbies to interact. Companies such as YesBoss, that has developed NLP and enabled people to book all sort of services through chat features, is one popular example.
Furthermore, Li stresses that “Though chats have implemented a strong UI that is able to traditionally increase conversion and interaction, it remains unscalable. You still need human operator on the other side. However, progress in NLP and basic AI enable us to automate chat interaction, eventually making chat UI more measurable and able to be commercialised.”
Geovedi reveals the answer that he received from Yohanes Nugroho related to the rising popularity of bots.
“Users are feeling saturated towards what most apps have been offering. Apart from that, the UI/UX of these apps is often considered lacking, making it hard for users to find what they are looking for,” he says.
Another factor that supports the rise of chatbot technology, as presented by Raditya, is the widespread use of social media and the nature of Indonesian users themselves, who enjoy chatting via apps.
Also Read: Artificial Intelligence 101: A map to understand where we are now
Early stage business development approach
As one of the industry players, Raditya is optimistic with the direction taken by AI development in Indonesia. There are many sectors that can be reached by AI, starting from e-commerce, financial instutions, to government. Even Singapore is currently undrgoing talks to develop chatbots; Raditya believes that Indonesia can also do the same thing in the future, despite needing a longer time considering the country’s lingual diversity.
“In the context of chatbots, I see a potential in the implementation related to customer services, as well digital assistant services that function as product recommendation feature. My opinion is based on my personal observation on interaction pattern between users and service or product provider. For example […] there are many official accounts on Twitter or Facebook that have started to respond to customers’ complaint,” says Geovedi.
As expected, there are several challenges faced by AI developers in Indonesia. Raditya and Li agreed that fundraising is one of the greatest challenges faced in developing AI. Users’ consistency in using the Indonesian language, be it in conversation or written form, is also one of it.
Li says, “One of the main challenges in developing reliable NLP for AI chats is sufficient data input. Not only that it involves developing codes that are able to understand human’s request, but it also requires greater data inputs in building pattern recognition and understanding all variaty of requests. So, in order to develop a reliable NLP machine, companies need a massive amount of requests to study.”
Li and Raditya also stress that at the early stage, companies need to give away free services in order to develop the bot. But in the future, AI-based on-demand services will be a prosperous business, that can be offered to SMEs to improve product marketing.
“It’s all about educating the market in order to win the market share,” Raditya adds.
Also Read: Appier aims to widen the scope of marketing through Artificial Intelligence
Hope for the future of AI and bots in Indonesia
Launching an AI service is not as easy as launching an e-commerce service. Apart from having to input sufficient data to create an engine that is able to process the Indonesian language with all its existing acronyms, punctuality is also needed in the process.
“We need more collaboration between service providers, API providers, and we also need more local developers studying AI. Because without the human resource itself, it will be impossible to make this technology a reality,” says Raditya.
Meanwhile, Geovedi expresses his hope for an open development of such technology, with active contribution from the community as individuals or small-scaled companies will not be able to make progress by themselves. For the long run, he also expects stronger participation from the academics. In the case where technological contribution or investment is too hard to do, an entity may release its data sets for training purposes.
The article Jalan Panjang “Bot” dan Kecerdasan Buatan di Indonesia was first written by Adjie Priambada and published on DailySocial. English translation by e27.
The views expressed here are of the author’s, and e27 may not necessarily subscribe to them. e27 invites members from Asia’s tech industry and startup community to share their honest opinions and expert knowledge with our readers. If you are interested in sharing your point of view, submit your article here.
Image Credit: Justin Luebke on Unsplash
The post The long, winding road for bot and Artificial Intelligence in Indonesia appeared first on e27.
from e27 http://ift.tt/29IQVcJ