Updated: Aug 12, 2019
We are now in the future. When you look at the past two to three decades, you should have noticed how fast we evolve in sharing and processing of information through various technologies. From that familiar sound in the telephone line when you use a dial-up Internet Service Provider (ISP), to the wireless fifth generation (5G) cellular network that may or may not harm honey bees and humans (not proven yet!); just the thought of where we have been and to where we are going in terms of technology is terrifying.
In the corporate world, these technologies can be the weapon to eliminate competition, dominate an industry or revolutionised an era. If you noticed it, the top corporations in the world are either dependent on technology or have developed technologies that we are dependent on, such as Google, Amazon, Microsoft and Apple. There are many more on the list such as Cisco and IBM dominating the network and cloud services. Wherever or whatever these corporations are, they have vast amount of patents and thousands of on-going innovations.
I might go on and have too much to babble on about different innovations. But in this article, I just want to focus on five things. You might have encountered the terms Big Data, Machine Learning, Natural Language Processing, Robotics Process Automation (RPA), and Internet of Things. If not, don’t worry. It only means these corporations are doing a great job trying to monopolise them. But, where is Artificial Intelligence (AI) in all of these? I can tell you now that there is no true AI yet. Maybe in a few months or in a few more years, we will encounter them. However, in order to build a true AI, the five elements must be incorporated into one.
Let me start off with “Big Data,” the first ingredient in building a true AI. The human brain can store millions of data that help us remember how we eat, how we understand and process speech and make decisions to what useless items we add in our Amazon carts. In the past decades, we don’t have the technology that can store big data. But now, we have those tools. By using bigger data storage (in terabytes!), faster database processing and even the use of Cloud, we can store millions of data that can help a machine to learn and decide. Still, corporations are now more than ever scrutinised on how they protect personal data they collect. Data should be secretly and enviously guarded by us and corporations that collect them as they are very potent munitions. Hopefully, you’ve heard the Facebook Cambridge Analytica scandal? But, that my friends, will have to be in another article.
Now that we have the capability of storing big data, we can move on to the next important ingredient in AI, which is Machine Learning. This technology is the scientific use of algorithms, patterns, and statistics from big data sets, to ultimately perform a task or provide a decision. With machine learning and big data, we are now close to AI than ever. Machines can now utilise historical data to come up with decisions or actions. However, there is always the risk of Machine Learning bias, whereas the gathering or usage of data leads to improper conclusions about data sets. This is often due to human intervention or lack of cognitive assessment. Types of cognitive bias that can be inadvertently applied to algorithms are stereotyping, bandwagon effects, confirmation bias, priming and selective perception. If asked who will be the next president of the United States, for example, the Machine Learning will conclude that it will be a 40 year old white male. This is due to the historical data that more than 96% of the US presidents are white and 100% male.
The third element to form an AI is the use of Natural Language. It is basically the ability for computers to interact with human’s natural language. Currently and during the ancient days of programming, computers are only able to understand bits of zero and one. We have slowly developed several programming languages to make it easier for the computer to understand the instructions we humans provide. If a natural language can be perfected, computers can now understand how to read, decipher, and make sense of the human languages in a manner that is valuable. Scary as it may seem, Sophia the robot is already using concepts of natural language. You might have encountered chatbots that can frighteningly make sense of some of the key words you’re inputting and almost accurately respond like Amazon’s Alexa.
Fourth on the list is RPA. It is essentially a software that can emulate and integrate the actions of a human interacting within the digital systems to execute a process. Imagine a software programmed to open a Microsoft Excel and start typing and doing v-lookups on a scheduled basis. If you combine Big Data, Machine Learning and Natural Language with RPA, you will be surprised if one day you can see an email as well as websites fully controlled by machines. If you’ve heard the news, in the next few years, Business Processing Outsourcing (BPO) industry will collapse due to RPA. From outsourcing repetitive processing e.g. Journal Entries and Reconciliations to BPOs, RPA can basically do it 99 times faster and more cost efficient. I don’t want to say that people will be losing jobs, but I know it will be a factor in an ever increasing unemployment ratio around the world. For now, we can be assured that RPA is only limited to processes that are repetitive in nature and with a simple logic to guide the software.
Lastly, there is the ever glorious Internet of Things. We are all connected, not close enough to be plugged into the Matrix yet, but our lives now are subservient to the Internet. Having a machine that is able to connect to the internet and can access your smart phones, smart tv, smart shower head, smart dog feeder and anything smart, can therefore have access to even more data. This is how an AI will learn, gather and be simulated more into humans. Tesla has already built a smart car, Tesla X, using most of what I have discussed above. If you say “Start,” the computer within it can recognise your speech, link up to its data through the use of the internet, learn your habits and control the other software in the car based on process automation. Imagine an AI having access to all there is that is connected to the internet, you can now feel how much closer we are to the movies.
I have mentioned only some of the elements that can ultimately build an AI. Tesla fiddled with the art of AI, but turned away, tails tucked behind their backs. What have they seen? Now that we are witnessing the rise of the intelligent machines, are we ready?