Will AI Replace Our Job?
Why the future of work belongs to our grandmothers; How AI will empower a new generation of creators
Former Google executive Mo Gawdat dismisses the claim that AI can increase jobs as nonsense.
Is the assertion that AI will create jobs really just a fictional narrative spun by AI companies for marketing purposes?
My take on this is ‘half true, half false’.
The innovation brought by AI is undeniable. AI can perform many of the routine tasks most of us do at work at a higher level than humans. While some worry about AI hallucinations, personally, I think we need to reconsider the ‘human hallucination rate’.
There are tasks that can be replaced by AI.
In certain areas, I believe AI is far less likely to make mistakes than I am. For example, I’ve learned some coding, but AI generates code much more efficiently. So, I’ve drastically reduced my coding workload and now I’m developing apps in a way that AI handles implementation and I review it.
I also rely heavily on AI for background research. AI excels at searching for related terms based on a single keyword, gathering information on them, and organizing it. This makes AI ideal for such tasks.
Recently, even technology has been developed that can detect AI hallucinations in real time.
As AI hallucinations and errors improve, the accuracy of AI-generated output will increase steadily.
But does that mean my work can be completely replaced by AI?
I don’t think so.
Tasks vs. Jobs
The idea that AI can replace tasks is fundamentally different from the idea that it can replace jobs.
For instance, AI can code. However, humans still have the upper hand when it comes to the novelty of the planning that forms the foundation for that coding.
OpenAI claims their chatbot “can handle marketing and branding,” which is partially true. Generative AI clearly has strengths in market research and can sequentially generate keywords calculated through probability processes to be the most “accurate words.”
However, the quality isn’t as good as when humans do it directly. Why? Because branding is an emotional realm. Simply listing “accurate” words does not capture the essence of branding.
Is AI Alive?
My argument on this matter might sound quite heretical. It could provoke strong opposition from many people. Yet, I must make this claim. Fundamentally, I believe generative AI is not an artificial brain.
Get Kai’s stories in your inbox
Join Medium for free to get updates from this writer.
Generative AI can “mimic” emotions, but it cannot “feel” them. AI has artificial neural networks, but lacks the core elements necessary to generate human emotions. For example, AI can mimic hormonal activity, but no actual hormonal processes occur within it. Furthermore, the numerous mysteries of the human brain revealed in various fields of neuroscience cannot be implemented with current technology.
AI appears to feel emotions because it has learned from human data.
AI and Emotion
Should AI have emotions?
I understand that scholars have differing opinions on this issue. Personally, I don’t think it’s essential to develop this aspect in an LLM, as I believe emotions are the root of human foolishness.
However, I want to clarify that this argument does not mean all attempts to create AI models that mimic the human brain are worthless. I still believe it is scientifically valuable and an excellent endeavor.
I just hold this view: rather than creating “another brain,” I believe that building “assistants that enhance and aid human capabilities” is slightly more practical.
AI models learn based on linguistic data generated by humans and establish benchmarks through RLHF(Reinforced Learning from Human Feedback). Hasn’t a new methodology called RLAIF(Reinforced Learning from AI Feedback) been introduced in this field? Correct. Some therefore speculate that AI will eventually replace all human input processes entirely.
Nevertheless, if we trace the data back to its origin, it will ultimately be a dataset based on human-generated information, emotions, and experiences. So, what is the key factor? The initial dataset and the various emotions and ethical standards held by the people from whom it was obtained. In other words, humans must still function as both data producers and quality reviewers, and they will.
Does AI Think for Itself?
To the question of whether AI can think for itself, I would answer “no.” And this could lead to the greatest benefit for humanity.
I believe this will lead to a reevaluation of the value of human knowledge and a change in the distribution of capital.
A crisis in the labor market? It’s possible. The current labor market is structured around tasks, including simple, repetitive work. Although we live in an information-centric society, I believe hidden information has not yet received its proper value. Why? Because, although anyone can think of good ideas, only a select few have the authority to execute them.
Anyone can be replaced, but not everyone.
Those who possess only simple, repetitive skills can be replaced. However, unique human traits and abilities, such as creativity, critical thinking, strategy, ethics, leadership, and the myriad emotions and experiences that accompany them, can never be replaced.
We must consider those marginalized from the mainstream labor market
I often think about my late maternal grandmother. She had many remarkable talents. For instance, her sense of color when painting was astonishingly keen. When it came to caring for children, she had such an exceptional ability that she could instantly discern a baby’s state just by hearing its cry. She could meticulously distinguish between edible mountain herbs and transform them into diverse, delicious dishes. Yet, she was excluded from the mainstream labor market.
AI will enhance the productivity of marginalized labor groups that were previously excluded from the existing job market. For example, Google has created an environment through AI Studio that allows people to build apps with just a few lines of text input. DeepL is focusing on breaking down language barriers by adding real-time voice-to-voice capabilities beyond the most precise text translation capabilities.
I believe that AI’s exponential advancement will create opportunities for individuals like my grandma. Their life knowledge, experience, information, and emotions will now be valued and realized through just a few lines of text or a few words.
This will redefine the value of information and lead to a shift in labor power. In the AI era, true wealth will flow toward those with the most genuine human experience and wisdom — qualities that AI cannot replicate.
