Sci-Fi flicks will forever have a place in my heart. From A.I. Artificial Intelligence to I, Robot, these flicks hauntingly illustrated futuristic worlds shaped by technology that gave rise to advanced A.I. and demonstrated astonishing examples of A.I. in action. But instead of Doomsday scenarios with humanity submitting at our metallic rulers, AI is slowly becoming one of the most significant forces that shape today’s digital age. It has the potential to not only impact the corporate world but also influence our day-to-day lives.
From Zoology to astronomy, A.I. has redefined how the info was being collected, integrated, and analyzed, leading to more insights and better outcomes. These technological breakthroughs are made possible by Deep Learning.
Deep Learning, which is a part of machine learning, is concerned with giving computers the ability to learn by themselves, and it’s led to many beneficial developments in language understanding, video/image understanding, and audio understanding.
Despite the huge potential of A.I., genius minds like Sam Harris, Elon Musk, and Stephen hawking still warn of the impending A.I. apocalypse. In fact, Elon’s Neuralink aims to stop a Terminator-style apocalypse by integrating AI in human brains.
Should We Start Worrying About Artificial Intelligence Now?
After all, A.I. is, in essence, artificial. Whether accurate or not, it will still follow its programming. And therein lies the biggest challenge: Startups pursuing A.I. tech lack a strong foundation of structured, ethical, and accurate data – the type of ground truth a machine needs to learn over time.
According to the MIT Technology Review, lack of quality data was the second biggest challenge for employing AI, just behind the shortage of internal talent. What’s more, more than half of all AI projects will not pass trials. The data challenge is difficult to overcome. That’s why you can’t just feed algorithms without having solid, clean data.
The Terminator Analysis
This reminds me of The Terminator. It’s famous for portraying a dystopian world with A.I. hellbent on destroying all of humanity. In the movie, The Terminator is on a mission to eliminate Sarah Connor. But the machine doesn’t know which Sarah Connor to eliminate. The robot only knows her name and city. Unsure of who the main target is, The Terminator dispatches all Sarah Connors on the phone book. The Terminator then intercepts the intended target rather quickly and spends the remaining time in utter destruction.
So, as intelligent and advance as the antagonistic Terminator is, it essentially must guest which Sarah Connor to kill because it lacks foundational, basic data – info needed for training a Machine Learning Model. It’s actually not until many sequels that The Terminator can identify the intended target because it has already been fed with enough data and learned from previous, failed attempts. (And there’s that time travel part that just gives me a headache.)
So, Artificial Intelligence, whether it’s a killing machine or just a chatbot, needs accurate, unique, timely, and complete data to make important decisions. If Skynet had given the Terminator the ground truth on its intended target, its mission would’ve been a walk in the park. Indeed, that would’ve meant a shorter movie and the terrifying scenario of Judgement Day would’ve been the conclusion.
There’s no question that A.I. will change the world’s future. But, don’t worry about the Apocalypse…yet. Truth be told, humans are still vital in the R&D of Machine Learning models. And, no matter how sophisticated robots may become in the near future, they will never be at their full potential until data is reliable enough for us to register.