<aside> 💡 TL;DR: The author highlights the potential of AI language models like ChatGPT in various industries and suggests a three-layer involvement hierarchy. They believe there's no need to rush, as AI is still in early stages. Short-term job replacement is unlikely, and personal health remains a priority despite technological advancements.
</aside>
Since I began using Chat-GPT nearly four months ago, both my work and personal life have become inseparable from this handy assistant. It is the first overt AI tool I actively use daily. For instance, the English version of this article was crafted with its assistance. Compared to several past trends, this tool genuinely offers direct help to me.
Over the past few months, I have pondered how to seize this opportunity. Coincidentally, the other night, as I lay in bed, I considered this topic once more. I decided to jot down my thoughts and questions on my phone.
First and foremost, there is no doubt that applying large AI models to various industries represents a massive opportunity, as significant as the internet in 1993 or the iPhone in 2007. The connection between these innovations is that they all drive industrial development. Why is this a great opportunity? I believe the primary reason is its potential to relate to everyone and every industry that utilizes digital products.
The next question is what can be done amid this significant trend, or how should one prepare?
To answer this, I devised a hierarchical structure:
The first layer consists of the organizations that create the models, such as OpenAI and Google's related departments. In the future, this will resemble AWS services. Given my limited technical background, my chances of direct involvement are slim.
The second layer includes organizations that use the tools created from these models, encompassing all existing companies and organizations. They leverage large model "wheels" to optimize existing processes or create new ones, like Microsoft's recently released Copilot. This is where I can actively participate and where the most excitement lies.
The third layer is the end-users of these tools. In fact, we are already here. Employing these tools to enhance work, life, and even learning is already immensely beneficial.
For me, participating in the first layer is not necessary. Even as a third-layer participant, I can significantly boost efficiency and seize this wave of opportunity. Put another way, with a willingness to learn, one can already leverage ChatGPT for more efficient learning across various subjects (imagine having a knowledgeable teacher by your side). Of course, it would be incredibly fortunate to partake in the second layer.
What truly matters are the applications, not the technology itself. Now that we possess new capabilities, we can add them to our toolbox and utilize them when needed, instead of seeking nails for our hammers. From another perspective, this development benefits product-focused positions. At least for now, clearly identifying and describing needs is more crucial than implementation. By optimizing "spells," outputs can be continually strengthened. Coincidentally, one of the core aspects of product work is recognizing and describing needs.
Regarding what specifically I should do, Chat-GPT provided practical answers to my doubts and questions.
Appendix: Career advice provided by ChatGPT
Would you miss the wave if you don't join now? If you don't consider yourself divinely chosen, there's no need to rush. You don't need to win everything; focusing on specific areas is sufficient.
Based on my past experiences, nothing is truly unmissable. Take smartphones as an example: the iPhone was released in 2007, but I only started using a smartphone in 2012 after my college entrance exam, a gap of five years; considering my career, it took eight years. So there's no need to hurry. In fact, haste is of little use. Even without direct involvement in manufacturing, utilizing these tools effectively within existing processes is another form of creation.
We are still in the early stages, and a wealth of potential awaits discovery. While the current GPT-4 model's textual performance is astonishing, I believe it is only the tip of the iceberg. What truly excites me is the integration of this capability with various industries' proprietary databases, leading to numerous commercial applications. This is why I am thrilled about Microsoft Copilot's release and GPT-4's collaborations with different platforms. It is merely the beginning. Of course, there will be challenges, and progress may not be swift, but this wave illuminates the path for at least the next decade.