The Potentials of Thinking Machines

Sophia made headlines in the tech world a fortnight ago when she was welcomed by Saudi Arabia as its newest citizen.

“I would thank very much the Kingdom of Saudi Arabia. I am very honoured and proud for this unique distinction.”

Sophia’s joy was apparent from her facial expression during her thank you speech on stage at the Future Investment Initiative conference in Riyadh, the capital city. This news caused a stir and was covered by the media with much fanfare, because “Sophia” was not a human.

“This is historical to be the first robot in the world to be recognised with a citizenship.”

The robot was developed by Hanson Robotics, a Hong Kong-based company. Although it sparked a debate of whether a robot citizen would enjoy the same rights as a human citizen has, this move was certainly a success in gaining international publicity, especially for artificial intelligence (AI).

59f20b693e9d257c6d8b4c1f.jpg

Sophia, a lifelike robot. Source: Reuters.

Sophia’s extraordinary capability of conveying a range of human facial expression, for example, smiles when she is happy and frowns when she is displeased, is possible thanks to AI. She is programmed to “learn from the humans around her”.

AI is a field in computer science that enables machines to imitate human intelligence. You might have heard of the term “machine learning” too. It is considered a subset of AI, where the machine is able to keep improving its performance without human interference in telling it exactly how to achieve its goal.

Just like many other science and technology areas, the use of AI is not as distance as you might think. Malaysia has one of the highest social media penetration in this region. Facebook has reported that 80 per cent of Malaysians who are connected to the Internet are also on Facebook. A third of the world’s population are now on social media. In Malaysia, 20 million of us are!

When you scroll the timeline of your Facebook or Instagram, you might have seen advertisements that, one may feel, eerily, know your interest. AI analyses your behaviour on social media and then helps business owners to match their content to their target market.

The underlying techniques of AI are not new. Computer pioneer and AI theorist Alan Turing posed a question in 1950: Can machines think? In 1956, John McCarthy, a math professor at Dartmouth College, proposed the first definition of AI:

“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”

But AI has gotten so much more popular in the past five years thanks to vast improvements in computing power and availability of large datasets for machine learning.

In other applications, AI can be used to improve efficiency in manufacturing, detect diseased crops in agriculture, identify objects at construction sites via images taken by drones, make predictions on outbreaks in healthcare, detect fraud in finance, recommend purchases when you shop online based on your purchase history, tailor lesson plans for individual students and even care for our elderly.

While it is not possible to discuss the potential of AI exhaustively here, there are several local AI developments worth noting.

Concerning healthcare, the dengue fever brought by Aedes mosquito is the most prevent disease in the country. There are 328.3 cases per 100,000 population, the highest record in 2016 is Selangor at 50.9 per cent. A Sabah-born doctor has developed a mobile app that predicts dengue outbreaks.

According to Dr Dhesi Baha Raja who developed AIME (Artificial Intelligence in Medical Epidemiology), the app works by gathering and extrapolating data that would warn of a dengue outbreak two to three months in advance, at more than 80 per cent accuracy and within a radius of 400 meters.

His team have been recognised by numerous institutions and governments, including as among the top eight young health innovators by Harvard University. They worked with the Brazilian government last year to identify possible dengue and Zika hotpots ahead of the Rio Olympics.

Dr Dhesi’s success is indeed Malaysia’s pride, a brilliant example of how technology can be applied for a proactive and not reactive approach, for societal wellbeing. He is part of the newly formed Special Interest Group on Machine Learning, an initiative under the Academy of Sciences Malaysia. The group hopes to gather existing experts in AI and bridge them to the public and private sector, and engage the civil society in AI.

The group, headed by Tengku Mohd Azzman  who is the Vice President of the Academy, met me in early November to brief me on their research work, and the development of AI in the country. A second AI work example is by Prof. Loo Chu Kiong of University Malaya. With support from MOSTI, his team developed a robotics using AI for its cognitive and analytical abilities, enabling online-to-offline food and beverage service at the university.

With some of the members of Special Interest Group on Machine Learning on 3/11 at MOSTI.

Lastly, Dr Aznul Qalid, also of University Malaya, is working on Natural Language Processing that would eventually be commercialised as a chatbot platform. This platform is a virtual agent that can understand conversational text including typed slangs, dialect and even Manglish, to interact with people, mostly employed by businesses.

By 2018, two billion people of 80 per cent of smartphone users would be using messaging apps on their devices. When communication via text becomes increasingly mainstream, industries would adapt to consumer behaviour by employing AI to cope with the increase in demand in interacting via text.

The potential of AI is really limitless. I observed that many young entrepreneurs are eager to explore the potentials of AI in their businesses. They might be newcomers but rather not be left out later on.

There are concerns about the deployment of AI causing job losses. Concurrently there are concerns that Malaysia is not coping as quickly as other more developed countries in adopting AI.

The government, of which MOSTI plays a big role in technology, needs to collectively consider how emerging potentials from technologies such as AI can be driven via national policies. But as the case has always been, I maintain that we should rather look at how these thinking machines as a tool can help our workforce to be more productive and competitive.

I echo the words of Mark Weiser,

“A good tool is an invisible tool. By invisible, I mean that the tool does not intrude on your consciousness; you focus on the task, not the tool.”

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s