Misconceptions Prevalent In ML & AI

GUPTA, Gagan       Posted by GUPTA, Gagan
      Published: June 21, 2021
        |  

Enjoy listening to this Blog while you are working with something else !

   

Blame unrealistic technology hype and media misinterpretations for all this misunderstandings. AI and ML are science's that's often misunderstood by those who are outside the field - and sometimes even within the realm. AI and ML related misconceptions have arisen primarily due to the following:

- Lack of authoritative facts.
- Evolving nature of the industry.
- General flaws in human logic.

Myth: AI and ML are the same and interchangeable

Both are closely associated; Machine Learning is a subset of Artificial Intelligence. AI is an umbrella term that includes ML, robotic process automation(RPA), deep learning(DL) and natural language processing(NLP) and much more. Machine learning is the statistical-based methodology used for training AI systems.

Myth: Machines learn from experiences

In fact, the opposite is true. Machines do not know what to learn. Machines do not understand how to learn. Human programmers design the learning architecture for an AI. The learning can be supervised, unsupervised or reinforced learning. Machine learning isn't dependent on experiences, but rather on data. You can't just turn a computer loose to attempt to solve a problem-machines need data to learn from and create algorithms to apply to future situations.

Myth: AI systems work exactly like the human mind

Any 3 year old can tell you the difference between a cat and a dog in less than a second with 100% accuracy. So far, no AI algorithm can match this. When we describe AL and ML, often we use some terms such as 'neurons,' 'think,' 'learn,' and understand,'; loosely. Although, this can simplify the meaning of the software function, this can lead people to mistaken conclusion that AI or ML works in the same way as human minds. Computers are good at doing certain tasks; fast and repetitive; provided they are programmed that way. And that's it.

Myth: AI will replace human workers

Arguably the most widespread and potentially dangerous misconception about AI is that it will take away jobs from humans and will make them useless. Yes, automation through computers is leading to the increased redundancy of a number of certain low-skilled jobs, but this trend has been significantly overblown in recent years. In addition, most scientific estimates demonstrate that computers create jobs than it displaces. According to my estimate, roughly 20% of the global workforce is directly working with in the IT industry. They are the one making the computers to work; and rest 80% or so are using it as a tool. Did car industry put bicycle companies out of work ? And then again, how many people are employed in car industry ? Just think about it.

Our On-Premise Corporate Classroom Training is designed for your immediate training needs
Misconceptions Prevalent In ML & AI
Misconceptions Prevalent In ML & AI

Myth: ML will make AI conscious in the future

Yes, they can definitely make a good movie on this idea. You can sure watch I,Robot or Edge of Tomorrow or Terminator. The creation of a self-conscious and fully autonomous system is hardly possible at all, at least at the current level of technological development. Remember Dolly; the year was 1996. How many of your clones do you see moving around in your vicinity? Like explained in the previous myth, to explain AI/ML, people resort to use loose term, which gives a false perception that it is actually mind we are talking about. A machine can be made to "learn" and take "action"; with given limitations. All depends on the data it has been trained with. But we still can never map why machine gave a particular 'reasoning' based on that data. Humans can give reasons for their actions because they have 'conscious'. How many times has Tesla crashed so far ?

Myth: AI is too difficult for laypeople to understand

At the end of day, AI and ML are just computer codes (0's and 1's). If you can learn to code, you can learn to code AI or ML. Very simple. With so many libraries around us, it is possible to create a fully working AI code with just 5 lines of code. I do it all the time. AI and ML is like a working black box, you feed the data in, you get the results out. Pretty straightforward. My 2 year old niece, knows how to use youtube. Behind the curtains, youtube executes one of the most complex AI algorithm ever written. AI and ML are here to augment the human capacity in performing certain tasks. 40% of entire global mobile internet traffic is youtube. Everybody understands youtube, without any user manual or a system guide.

Myth: AI yields immediate results

This one makes me laugh. I hear it the most often, and specially from the people in my own industry. Too many organizations implement AI-based resources with the expectation that they can 'fill it, shut it and forget it.' This approach is sure to lead to disappointment and, potentially, a significant loss in ROI. AI is complex and requires a significant time investment in order to properly work and to produce any sort of meaningful result. Problem is, even money and talent will not give you the results in AI. Every AI technology is based on data, rules and other kinds of input from human experts. Your AI and ML algorithm is as good as your Data. End of story !

Myth: AI algorithms are neutral and objective processes

AI is simply neither neutral nor objective. The code is as good as the people behind it or the data behind it. Quality and trustworthiness are not the attributes of a computer code. Because all humans are intrinsically biased in one way or another, so is the AI. They are ethical topics, and governed by the very people who create the algorithm. If companies want to steal your data, companies will steal your data. There are laws to monitor and stop these practices. Often, AI is easily mis-sold to people who have too much faith in the technology itself.

Our On-Premise Corporate Classroom Training is designed for your immediate training needs

Myth: My business does not need an AI strategy

Business can survive without it, but to grow, every organization should consider the potential impact of AI. Organizations should investigate how this new technology could be used in solving their business problems or planning the next phase of automation else enterprises will be at a competitive disadvantage. Find use cases that leverage AI's power to augment human work, decisions and interactions.

Myth: AI is expensive

This one is True ! Implementing a fully automated system at the enterprise level isn't cheap. Neither is building an analytics team to make sense of terabytes of data. Costs stack up very quickly when you think of the data and the training. The key is for each business to figure out what they want and apply AI as needed, with respect to their unique goals and company scale. Remember, AI is not a one time cost. The more the AI system is used, the more the AI system keeps changing, thus triggering cost for updates. Once implemented correctly, AI can do wonders, but it is best to start on things with eyes open and after considering every possible pitfall.

Myth: AI can undertake any task

Maybe one day in future, far away from now. Not in the near future, not at least with the current pace of development. With our current level of technology, an AI can identify a human figure with a given accuracy. But, it cannot be made to recognize its gender, emotions, mood, intentions, etc. all at once. Another problem is large, diverse and significant datasets.

For example, to design an AI solution for predicting possibilities of extraterrestrial life, you would need relevant and extensive data set about the universe itself, all life forms, evolution, planetary systems, and numerous conditions for sustaining life, etc. Creating and cleaning such a massive data set is inherently impossible.

AI solutions cannot replace complex tasks as of now. Take performing surgery as examples. These jobs require complex information processing and taking the best possible decision at all moments. Current ML architectures limit the machines to do so. Right now, the AI which are widely used serve a single purpose and are called Narrow AI.

Conclusion

Some misconceptions were born out of unrealistic expectations, while some resulted from the misinterpretation of information.

A search on Google defines common sense as 'sound, practical judgment concerning everyday matters, or a basic ability to perceive, understand, and judge in a manner that is shared by nearly all people.'

No one yet knows how to capture such knowledge or abilities in machines. However, due to the advancement of technology, this may become possible in the future like in some AI movies or comic books, may be !

Over to you. What do you think about AI and ML and their future? Let us know.

Support our effort by subscribing to our youtube channel. Update yourself with our latest video`s on Data Science.

Looking forward to see you soon, till then Keep Learning !

Our On-Premise Corporate Classroom Training is designed for your immediate training needs
Misconceptions Prevalent In ML & AI
                         



Corporate Scholarship Career Courses