On November 29, 2018, top management levels of convened at Siam Kempinski Hotel Bangkok at the Thailand Management Day 2018, setting the stage of discussion for the future of Smart Connectivity on upcoming opportunities and challenges. TMA Thailand Management Day 2018 began with a Welcoming Speech by Ms. Wanweera Rachdawong, Chief Executive Officer of Thailand Management Association.
Rachdawong stressed on the impact of smart devices and seamless connectivity as a game changer for future opportunities in 4 aspects as follow:
• Connecting Human with Smart Devices – the core topic of today’s discussion and how human should be able to leverage from the progression of increasingly intelligent smart devices to increase efficiency and productivity.
• Connecting Smart Devices with the Network – with increased accessibility and speed, seamless connectivity will reinsure that the users will be able to access the network without fail at all places and all time.
• Connecting Traditional Machines – where smart connectivity includes not only new devices or machines, but also includes increasing the connectivity of traditional devices and machines such as television, refrigerators, and other traditional devices by a single control point such as the Smart Application on Smartphones.
• Connecting Smart Devices with Personal Data – where Big Data is the core to the development of the intelligence of smart devices in understanding human needs, as well as increasing opportunities through leveraging data management.
And through single entry point, Smart Connectivity will open up many more doors and opportunities for improving our quality of life. However, there are many challenges that we must face when dealing with Smart Connectivity. The TMA Thailand Management Day 2018 is an open forum to discuss the opportunities and challenges of Smart Connectivity.
Rise of AI and the World of Machine Learning
Dr. Peter J Bentley, Chief Technology Officer at Braintree Ltd., and Honorary Professor, Department of Computer Science, at the University College London opened his session on the Rise of AI and Machine Learning with the question, “What is Artificial Intelligence?”
Bentley pointed out how the definition of ‘Artificial Intelligence’ or AI has changed from when he first started his studies 25 to 30 years ago – from when the understanding of artificial intelligence was very limited, where people used to think about AI as artificial examination, or some secret intelligence aimed towards taking over to world.
It was not until 1951 when Claude Shannon, the pioneer in Computer Science, started talking about a machine that could solve mazes, that could remember and forget things, and learn languages. In fact, Claude Shannon was one of the pioneers of artificial intelligence. As we can see, the concept of Artificial Intelligence backdated to even before 1951; Artificial Intelligence could even be traced back to 1945, to the time before the first computer even existed. Artificial Intelligence, as Bentley demonstrated, is hardly a new concept.
If we are to ask, “What is Artificial Intelligence?”, it is the oldest computers. AI has always been inspired by biology.
– Dr. Peter J Bentley
Bentley then gave his favorite definition of Artificial Intelligence as “Something that could make computers do something they currently cannot do”, because the thinking behind artificial intelligence is to be able to make computers do things that nobody thought they could do – composing music, reading handwritings, or painting a picture. And ironically, as soon as we could get computers to do all those things, we would soon forget that all those capabilities are Artificial Intelligence.
If we were to predict what would be the future of Artificial Intelligence, the simple answer would be – to be able to do the things we cannot achieve today – predicting complex systems. A highly complex system is basically a system that is highly connective with many different components leading to unpredictable behavior. So when we talk about Smart Connectivity, we are in fact talking about a highly complex systems. Predicting the behavior of complex systems could not be achieved with simple mathematics, and this is where AI comes into play.
In order to conceptualize how predicting the behavior of a highly complex system could be achieved with Artificial Intelligence, Bentley started off with trying to predict the behavior of Software Developers. Take the App Store, for example, is it possible to predict the behavior of millions of App Store users and developers? Is it possible to predict which application would become more successful than others? Is it possible to predict which Marketing Schemes would work better? With the computing powers available today, what we can do is to focus on the individual elements in a system, in this case, the element of concerned is People – those who buy applications, those who develop applications. We could in fact create virtual users, virtual developers – and with enough computer power, we could create millions of virtual developers, each one developing different Software; at the same time, we could create millions of virtual users with million different personal preferences. From this virtually created environment, we could attempt to predict which applications App Store users are likely to download by creating different models. We could even model who these virtual users might talk to each other via Social Network, we could even model the effects of publicity. Bentley further explained that in order to be able to create these models, we would need to be gathering an immense amount of data, where these data will be used to calibrate the models so that the virtual environment would match the actual environment. With these models, we can conclude that Mass Media, for example, is a very contagious mean of advertising and application.
This concept could be applied to any complex system. By leveraging the use of data, you could even predict the likely success or failure of different strategies of your business. Bentley gave another example of how AI can be used to deal with other complex systems – the progression of a cancer tumor. The development of a cancer tumor of one person is different from another. Being able to predict the mutation of cancer cells in each person is very vital for the development of cancer treatment. Applying Artificial Intelligence to this case, we could model the various types of tumors, creating a virtual environment to model the progression of each virtual tumors and how they develop in different environment in a cellular level. Running computer simulation of how each cell mutates, we would be able to understand how different environments could affect each type of mutations, and thus, come closer to understand the causes of certain tumors. This is one other example of using Artificial Intelligence to understand a highly complex, highly connected system.
In another example of predicting people’s behaviors in a workplace in the US, Bentley demonstrated how Artificial Intelligence could be use to understand the collaborative environment of a workplace by leveraging data collection to create a virtual computer model in order to understand how different solutions could affect people’s interactions for each person with different personalities, and how different solutions affect the team’s collaboration, productivity and efficiency. In this case, with Artificial Intelligence, we are able to predict which solution would allow the team to work for effectively, predict which team are most likely to fail, or to succeed. By creating different models, we are also able to evaluate which factor has more effect than others, for example, you could isolate different elements to analyze, such as the characteristics of the team members, or the nature of the tasks at hands.
Bentley pointed out how today, with Machine Learning, we are starting to see a more structured data, and how we are able to make use of these structured data. With structured data, we are able to extract useful information, analyze and learn from those data to achieve more useful features. This is particularly applicable in the case of Social Network, where information is interconnected in a complex network of data. And if we are to make use of these interconnected data, we will be able to predict and extrapolate new solutions from recognizing interesting patterns of behaviors. We could also enhance the structure of these data as we learn more from them, and with this, we can create new machine learning algorithms that can repeatedly learn from these data more quickly with more computing power.
Smart connectivity is another impressive machine learning like brain. It takes structure data and made changes the link between the data as they learn. Your brain can figure itself, change itself, and it change connections. This is inside-your-head smart connectivity.
Ending his session, Bentley emphasized on the fact that AI has always been inspired by natural systems like human biology, so in the end, artificial intelligence should somehow be superhumans, they should always be inspired by how our brains work. And our brains are indeed designed to allow us to survive in a highly complex environment, endlessly predicting the unpredictable. As simple as it sounds, this is the future of Artificial Intelligence.
Source: Thailand Management Day 2018