1. I hear you. I’m probably not the best person to comment on the connection of career and life purpose in the modern age. I was intrigued at an early age by the How and Why Wonder Book of Atomic Energy and became a physics professor. On the other hand, I have a lot of experience advising young people and I have seen how careers change. In my own professional career, my conception of my work has evolved, as has the toolset required. But purpose is not really the final word on our identity. Fundamentally, it is our values and the principle we adhere to which define us. I am a scientist and a scholar. From that, a certain part of my life’s purpose emerges. However, I am a scientist and a scholar because I embrace the values of those communities. From a GTD perspective, my areas of focus (roles, if you prefer), short-term and long-term goals change just like projects and next actions do, but usually more slowly. My values evolve too, but usually even more slowly.
2. I think the jury is still out on AI. I have some experience now with AI in physics research, and have seen how my wife uses it in neuroscience. I tend to agree with a very bright younger colleague who said it was “like having a not very bright graduate student.” We’re doing the experiment and we’ll see. It may turn out that we can’t afford widespread usage, or that the value/cost ratio is too low. AI is undoubtedly disruptive, but it comes at a time when higher education and STEM are both sorely stressed, at least in the US. I do not envy young college graduates today, but I am quite sure AI is being used to help justify actions in both government and industry that temporarily bolster apparent economic health.
Thank you, mcoglivie ,for such a thoughtful and wise response! I deeply respect your journey and the scientist's values you uphold.
1. On Values and Dreams.I actually really envy that you were able to realize your childhood dream. When I was young, I also dreamed of becoming a scientist, specifically a computer scientist. However, as I grew up and observed how society truly operates, I realized there is a stark contrast between the "idealized image" of a scientist and the reality of the profession. Sometimes, the realities of that path—particularly the environments I experienced and the culture of the people I encountered—actually contradicted the core values I wanted to hold on to.So as I went through my education and work, I found myself gradually losing interest in the very subjects and pursuits I once felt drawn to. Of course, I have to admit that I probably wasn't naturally gifted in this field either.Because of these observations, my values have been continuously shifting, which occasionally leaves me feeling a bit uncertain about what I ultimately want. But as you perfectly put it, our values evolve, and I am still in the process of discovering mine.
2. On AI and its Bottlenecks.I completely agree with your young colleague’s analogy of AI as a "not very bright graduate student." High-level researchers rightfully see its current flaws. However, I believe these gaps will gradually be filled.
From my perspective,
the core bottleneck for AI right now isn't just computing power, but the lack of objective evaluation standards in the real world. If we look at
AlphaGo Zero, it could continuously upgrade itself through self-play because the game of Go has absolute, unambiguous rules for success and failure.
Because it knew exactly what "right" was, it was able to evolve to a super-intelligent level far beyond the highest limits humans could achieve or even imagine. Today's top Go masters—who have already mastered every human-discovered strategy—often cannot comprehend how AlphaGo makes its decisions, and are now learning entirely new paradigms from it.
But the real world lacks a unified, ground-truth evaluation system. Because AI doesn't have an absolute metric to judge its own actions in open-ended real-world scenarios, it struggles to self-evolve efficiently. That's why we see endless new benchmarks emerging—they push AI's capability boundaries forward, but they are not the ultimate answer.
3. On the Plight of Today's College Students
You might have gotten the impression that I think today’s college students are lucky, but actually, my perspective is very similar to yours—I don't think they are lucky at all.
Earlier, I mentioned that people today have many choices, but having an abundance of choices isn't necessarily a blessing. When everyone has endless options, it doesn't make life easier—it actually makes it far more exhausting, leaving people increasingly restless and superficial.
More importantly, the very existence of AI is fundamentally devaluing academic degrees. When AI can easily handle tasks like writing papers, traditional academic credentials lose much of their weight. In traditional education, students are simply handed pre-defined questions to answer. But in the real world, problems must first be defined. In the AI era especially, the ability to figure out and define
what the problem actually is has become far more critical than just finding answers for pre-defined questions. But ironically, this practical ability has very little to do with the academic degrees these students are spending years trying to obtain. Therefore, they are essentially trapped in an unprecedented crisis of meaning.
Lastly, your point about AI being used as a tool to temporarily bolster apparent economic health by government and industry is incredibly sharp. I completely agree, though it's a heavy and complex topic I won't expand on for now.
Thank you again for the profound insights. This conversation has been incredibly rewarding for me.
