By Muhammad Ubale Kiru
Whether Artificial Intelligence (AI) can surpass human intelligence is a complex and debated topic. Many scientists, AI users, and observers have argued whether what we see in movies regarding AI surpassing human intelligence will come true. I have asked this question several times, and colleagues at work and friends on social media have asked me whether this myth can be true. Since then, I have been gaining momentum, strength, and proof to be able to answer this question.
The above statement has directly or indirectly revealed that if you agree to use ChatGPT, you must surrender to the fact that OpenAI will collect personal user information for research and training purposes. A non-specialist will not understand the implications or consequences of that. One may think it is an ongoing activity because social media companies like Facebook, X (formerly Twitter), etc.) collect users’ personal information for business and quality assurance purposes.
So, what is the real implication here?
It is simple. AI and machine learning algorithms are like weeds on a plant. They rely heavily on data to learn. The more data they consume, the more intelligent they become. Most of us are already using AI to solve our day-to-day activities and problems. For instance, tasks that used to take me seven days to complete can now be done in 10 minutes. I am handing over my tasks to AI to handle them for me.
Each time I ask AI to handle my task, AI is learning the task more and more. Humans perfect their skills through constant and regular learning. Now, I’m handing over most of my tasks to AI; AI learns while I lose because previously, I learned from my work experiences, and now AI does the work for me. AI is becoming more intelligent and capable, while I am becoming less intelligent and less capable. By the way, I’m not the only one in this mess. Nowadays, even programmers who rely on constant practice to improve their coding skills are also using AI to generate codes or programs that used to take months to complete.
ChatGPT, for example, is used by millions of users daily. When it was first developed, they used random internet data to train its learning models. Now, they are using real-time human input (data) to train the AI. If you look at the core foundation of any AI in the world, it is designed to capitalise on learning from its environment. Our phones are AI-enabled, laptops are AI-enabled, web apps, games, calendars—everything is now AI-enabled. The more we use AI, the more AI learns about us.
Today, your phone keypad knows more about your words and thoughts than you know yourself. As you begin typing, it completes the rest for you. So, with time, your AI-enabled devices would learn more about you than you could ever learn about yourself. Thus, what is left of us if AI has learned everything about us? In Sun Tzu’s book, The Art of War, he says, “Knowing your enemy is akin to winning half the battle. Understanding their strengths and weaknesses provides a strategic advantage that can pave the way to victory.”
The question of whether AI can start a revolution or take over the world, as we have seen in movies, is another debate for another day. The Tesla CEO Elon Musk and AI guru Lex Fridman are among the few people in the world who are always concerned about the potential danger of AI and have continued to call for regulations before AI gets out of hand. The technology has very speedy and staggering growth potential. It is growing at a breakneck pace right now.
To this end, I urge policymakers and regulatory bodies to take necessary precautions before AI gets out of control. AI is undoubtedly powerful, and if unleashed without caution, it can create devastating chaos.
Let me hear your thoughts in the comment box.
(c) Muhammad Ubale Kiru