Like any other revolutionary technology that has been—and will still be—produced, it will receive a fair share of doubts and fears. While it is only natural for any individual to feel this way, us as responsible members of society ought to take extra care in educating each other on topics such as these.
Think of it like this: If an inaccurate/incorrect detail around AI is shared it could spread a dark cloud of misunderstanding and consequently hinder/ decelerate the evolution of a valuable technology. Hence, it is vital for everyone to do their part in debunking any misconceptions of AI; as it not only beneficial to a single individual but the progression of society as a whole.
We will be diving into common misconceptions surrounding the development of this technology as well as revealing the truth to these concepts.
We’ll start with the most common misunderstanding pointed out in an article by the CIOL Magazine (India): Machine learning—a category within the broad field of artificial intelligence—is merely inspired by the processes of the human brain; however, it is differently not one in the same.
The magazine elaborates on this further as says that AI is much more efficient and accurate in recognizing faces than humans are; although this same algorithm will be unserviceable when faced with a mathematic problem. In a nutshell, artificial intelligence is simply not as adaptable as humans.
A fear that plagues society today is the fear of employment due to artificial intelligence. It is a joint understanding that this technology will rid the world of human-based labour. This is entirely false and if anything, artificial intelligence will provide something even better… human-centred labour. This means that human evolvement will not cease to exist within the workplace, but it will be supported instead—similar to how CAD (Computer Aided Design) software like Maya Autodesk aid designers in the planning of buildings or cars (CIOL; Bengaluru, 2019).
Another misconception is that AI learns on its own and that it is free of any forms of biasness. This (arguably) unfortunate for many of us to hear. While AI can learn by machine learning algorithms, it simply cannot do this on its own and has to pumped with data to study and analyse. This means data scientists and others working on its development must constantly provide new pieces of data if they wish for the AI to become “smarter”.
Which brings us to the next point. AI is biased. This is not its own fault, but the fault of those who train it. Keeping data unbiased is an extremely difficult task and comes to no surprise why AI tends to have these “opinions” (Chowdhary, 2019).
So, what can we take away from this post today? Here is a list of common misconceptions surrounding machine learning and the workings of artificial intelligence, some of which will be explained in more depth in other posts:
- AI is not the human brain, and the human brain is not AI. Each entity has their own flaws and perfections—where AI is designed for accuracy and repetition; the brain is designed for adaptability and creativity.
- AI is not meant/nor has the capability to completely replace human labour. The main intention behind it, is to simply aid humans in unusual cases.
- AI cannot learn on its own and depends solely on the data fed too it. This means it follows the basic principles of GIGI (Garbage-in, Garbage-out ).
- There is a high change of AI software being biased based on the data fed to it. While this is a major problem that is trying to be corrected, scientist can only ensure that the river flowing into the dam…is clean.
References
Chowdhary, S. (2019, March 21). Debunking myths and misconceptions about artificial intelligence. Financial Express; New Delhi. Retrieved November 4, 2021, from https://www.proquest.com/docview/2194405362/fulltext/998EC6864CF64A18PQ/5?accountid=188233
CIOL; Bengaluru. (2019, February 14). Gartner Debunks Five Artificial Intelligence Misconceptions. CIOL; Bengaluru. Retrieved November 3, 2021, from https://www.proquest.com/docview/2179758501/fulltext/998EC6864CF64A18PQ/2?accountid=188233
The AI Guy 