After hearing a lot of news and updates on Artificial intelligence, machine learning, deep learning, neuralink etc we often times wonder whether these technological concepts are different or whether they are related, well if you have been in Information Technology domain for quite some time, I am certain you are quite familiar with these terms.
As we all are towards a era of digital transformation, these technological developments will change or rather evolve our existing environment to a more dynamic state. So today i thought it is important to help everyone form a non- technology background to understand what is machine learning? why is machine learning different? what are the advantages of machine learning and moreover the limitation of machine learning?
TERMINOLOGY “machine learning”
The term “Machine Learning” was a phrase first used by Arthur Samuel in the year 1952. This was when Arthur Samuel while working for IBM had designed a computer that could play checkers and learn various winning patterns algorithms each time it is used for a play. This was back in the year 1950. The concept of machine learning in simple words is derived form a machines ability to decipher and learn a function at the same time, by getting better each time based on the data accumulated in its memory.
In the late 1700’s the use of machine became a necessity because of it input vs output ratio, quality and more importantly improved the efficiency at an extreme level, due to industrialization. When these machines performed varied tasks in the manufacturing segment the entire workflow became very robust and efficient, but this still needed human intervention as these machines can only perform a task under human supervision or help.
This still was not enough as many manufacturing units had to invest on both machines and labor relentlessly, that’s when the thought of computer arised where in machines could complete a task successfully based on the need of the task.
That’s when the thought of modern computer came to rise in the early 1870’s and this equipment was less mechanical and more technical. which was still far away from achieving a machine that works and achieves a single task with utmost efficiency. In the early 1900’s, a British Mathematician, James Hilbert went on to identify the possibilities on how a computer can solve a problem or a task by deciphering information in the form of data, there were multiple challenges he was thrown at, the most important one included whether a computer can handle multiple tasks with perfection and creativity.
That’s when Arthur Samuel and team of experts came up with the discovery of the computer that played checkers in the early 1950’s. The fundamental of machine learning further progressed to what it is today because of a major contribution by late Dr. Alan Mathias Turing, who went to explore on the concept “Can machines think like humans?” based on his further research identified that a computer accurately completes and learns a particular task based on the readily available data or the data it has acquired through experience.
This still didn’t solve the challenge in hand, which was can machines multitask and learn simultaneously, here is where Dr. Alan Turing visualized the entire process by which he created the fundamentals of computer science, i.e. In order for a machine to perform a series of task, it must have a set guidelines that it could follow in order to achieve the set goal simultaneously learn different other possibilities based on the pre-recorded data. This entire process is called computing in modern world.
ADVANTAGES OF MACHINE LEARNING
We are living in fast paced world, where information is the key. This information in the environment comes down to binaries i.e. 1’s and 0’s. Today complex problems are solved with the aid of machines and it has helped us to grow to a significant level. These are some of the advantages that machine learning has contributed to and may contribute in future.
- Machines have been able to help us automate various types operational activities across various industrial verticals through digital transformation.
- Machine learning has helped us a step further in creating computers with its own brain, which are able to multi-task with minimal errors.
- With the help of machine learning, large amounts of binary data available in for each process are easily stored, analyzed and are made available at any given point of time.
- As machine learning function improved, it created opportunities around other technological functions such as Big Data Analytics, IoT, SAAS, PAAS, IAAS, etc.
- Today, machine learning capabilities are applied in the fields of Healthcare, Archaeology, Manufacturing, Stock Market, Banking & Finance Services, Trade & Commerce, etc.
LIMITATIONS OF MACHINE LEARNING
Though machine learning has contributed effectively in advancing our technological infrastructure, there are numerous downsides of this piece of technology
- As I have explained earlier machine learning is made possible when the there is enough data for it to process a take the needful action instructed to the machine at the time of programming, but this piece of information in wrong hands can impact to technological terrorism.
- We are well aware of eCommerce companies, social media platforms use Artificial intelligence and machine learning algorithms to understand and exploit the buying behavior. Many may disagree with this fact, but think again are you really buying what you need? or are you just buying what they want you to take? Information is good but, it must not be exploited for business or entrepreneur or political gains. We have also seen multiple times the social media platforms support certain political views and misguide the users to win their agenda in the form of campaigns.
- We know that the COVID-19 situation itself caused Pay-cuts and loss of jobs even in sectors which were not impacted, it clearly shows that on entrepreneurs exploit the situation at hand, so just imagine, if the machine learning was adopted globally and narrow artificial intelligence is adopted too, do you really think you would still keep your jobs? I have heard this argument a couple of times that use of AI will create new jobs, my question to them is for how many? and what type of knowledge and credentials are required to apply for such a position?
We would love to hear from you so please feel free to leave a comment below and also please feel free to share this with your friends, colleagues and family!