Human vs AI. Natural vs artificial. These themes often dominate conversations about AI among the general public. Thus, the premise of how AI can help us become better humans might seem ironic to some. More often than not, AI is seen as the doombringer of humanity. Whether it’s being accused of taking our jobs or, even worse, our world being ruined by an AI overlord that can control every aspect of our lives.
However, I might be weird… No, I know I’m weird, because I am a mismatch of things usually seen as opposites.
After generating thousands of images using ComfyUI and SDXL model variants, creating Cyberpunk 2077 mods with AI-generated voices, and utilizing local and online LLMs daily for my professional work and passion projects, I have learned something that I feel is worth sharing.
How can learning Generative AI help us become better human beings?
Is there such a thing as objective reality?
In case you didn’t know, generative AI needs models to run. These models consist of billions of training data. For example, text-based models like GPT are trained on diverse text corpora, while image generation models like SDXL are trained on extensive image datasets.
This training data is their reality, which is obviously not objective. It entirely depends on the engineers in choosing the data that will become their reality.
Most people may believe in objective reality, a reality that exists independently of our subjective experiences and perceptions. For example, basic math or basic laws of physics.
However, this understanding is challenged by scientific theories like quantum mechanics and relativity, which suggest that reality is observer-dependent and that the act of observation can influence the behavior and properties of particles. There is an ancient famous question, often attributed to George Berkeley, an Anglican Bishop and philosopher in the 1600s: “If a tree falls in the jungle but there is no human around, does it make a sound?”
I do think that both types of reality (subjective and objective) exist and are not mutually exclusive. However, for me, the most important lesson from this is that learning and understanding a new perspective or new reality is crucial for us to become better human beings, so we can reduce our bias and have stronger empathy towards others.
Just like Alfred N. Whitehead said, “There are no whole truths; all truths are half-truths. It is trying to treat them as whole truths that plays the devil.”
Human greatest achievements are collective efforts
Like I previously mentioned, generative AI can be smart because it is trained with vast amounts of data. This data is undoubtedly produced by humans collectively. Billions of people throughout different ages collectively communicate and compile all of the knowledge so that the next generation could use it for the betterment of humankind, including as AI training data.
If I’m not mistaken, Yuval Noah Harari said something like, humans rule the world because we can work well together on a massive scale because we believe in the same stories that bind us together. While monkeys or other animals can’t do that.
A while ago, I was asked to give a workshop at a university to discuss the issue of AI and copyright. In one of the chapters, I talked about how originality is overrated. No, I’m not saying that we don’t need intellectual property law, nor do I support plagiarized content. However, I want to emphasize that we can’t create something out of nothing.
Even the first author, the first painter, and the first sculptor needed reality for their inspiration.
In this current time, with so much divisiveness and hatred all over the world for whatever reasons, I believe that we need to glance back a little and realize that our greatest achievements are all collective efforts.
Newton couldn’t create calculus if there were no Archimedes, Descartes, Fermat, Barrow, Cavalieri who laid the groundwork and foundation for calculus. It’s important to remember that scientific discoveries often build on the work of others.
The realization of human stupidity
One day I found this meme saying, “before we work on artificial intelligence, why don’t we do something about natural stupidity?”
I believe this is also a very important lesson that we can learn as humans so that we can be better.
There are many people who say that generative AI can do anything. Some others say that generative AI will replace human professionals. However, the truth is: No, it doesn’t. Generative AI can’t do everything and can’t replace human professionals—at least not at this current stage.
If people say that AI can do anything or can replace their job, it only means two things. First, they know nothing about generative AI. Second, they are not that good at what they do.
Let me explain.
ChatGPT is often said to be able to write an essay from scratch for us. However, this is not entirely true, at least not according to the standards of good journalistic work or academic papers.
While ChatGPT can help a lot with grammar and structure, its research capabilities can be misleading. This is tested on both free and paid models. Although, the free model is much worse. Even if the premium model can search the internet, I don’t think it could differentiate which ones are reliable sources and which ones are not.
READ ALSO: Microsoft, Indonesia, and AI: The $1.7 Billion Pledge
If you have reached this part of my writing, I can safely assume that you know that not everything you read on the internet is true, reliable, or empirical.
I don’t think ChatGPT models are already trained to differentiate those sources or put priority on some sources rather than others.
Another thing that I have repeatedly said, whether it’s on my YouTube videos, my writing, or even directly face-to-face, “Generative AI is not a shortcut. It is a tool. A very smart, advanced, and sophisticated tool, but it’s still a tool nonetheless.”
Just like any other tool before generative AI, like a DSLR camera, the internet, or even a knife, can’t make any of us suddenly become an expert.
You can’t be a great chef just because you own the most expensive knife. You can’t also become a professional photographer even if you have an overpriced DSLR camera. Just like my former boss said, “it’s about the man behind the gun.”
One more thing that I have said before, generative AI capabilities can only surpass the amateurs. It can’t give better results compared to someone who has truly studied their craft, a real professional. However, a decent professional who can utilize generative AI could do the work better and faster than the other professionals who don’t want to learn about generative AI at all.
So, if you think you’re a professional and think generative AI can do a better job, then I have bad news for you, buddy. You might not be as good as you think. It’s the Dunning-Kruger effect at its best.
Human vs AI: A Closure
So, those three points are the crucial lessons that I learned during my time dabbling with generative AI.
Of course, don’t believe anything I said blindly. Learn deeper about generative AI and hone your craft sharper. Do let me know if you found a different takeaway. But more importantly, no matter what you found, I strongly believe that every learning moment won’t do us wrong.
