- Published on
The new world after ChatGPT
- Authors
- Name
- Rand Xie
- @Randxie29
Beginning
It's my first blog post to talk about ChatGPT and written with the help of ChatGPT. In this blog post, I would like to share my perspective on ChatGPT based on my previous experience.
The New World
As someone who has been working in the field of artificial intelligence for years, I've seen firsthand how quickly things can change. Back in 2018, I was heavily involved in the field of feature store, and while I was successful in pushing it forward, I couldn't help feeling that it lacked the imagination and potential of more cutting-edge technologies like GPT4.
When GPT4 was released, I was in awe of its capabilities, but I also realized that it's not a panacea. There are still many areas where it falls short, such as 3D vision, manipulating the atom world, and developing new computing paradigms. But rather than feeling discouraged by these limitations, I see them as an opportunity to think bigger and pursue multi-10-year visions.
To truly unlock the potential of AI and shape the future, we need to embrace a more visionary approach. This means thinking beyond our current limitations and being open to new ideas and possibilities. We need to pursue multi-10-year visions that will allow us to push the boundaries of what's possible and create a better world for all.
One area where I believe AI can have a significant impact is in education. By leveraging AI to help mature breakthroughs, we can create a future where students spend more time creating new knowledge and pushing the boundaries of human understanding. This will allow us to unlock the full potential of these technologies and create a world that is both exciting and full of promise.
But to make this future a reality, we must be willing to embrace a multidisciplinary approach that combines expertise from all fields. Science, engineering, and product development are becoming increasingly intertwined, and we must recognize that for AI technologies like GPT4, they are almost the same thing.
For me, this means pursuing a multi-10-year vision that combines AI and robotics, allowing us to connect GPT4 with the atom world. By doing so, AI will be able to sense and engineer the real world. A small target can be embedding Large Language Model (LLM) to Boston Dynamics' Spot, with inputs more than text and image. It's interesting to see AI develops skills in a totally opposite order compared to human-beings, where AI goes from Text -> Audio -> Vision -> Action. And human-beings are born with action capabilities, then recognize things, and learn to speak and write.
Final thoughts
Writing can be a painful process, and I am glad that ChatGPT can greatly simplify our job. A lot of times, we only get a few random paragraphs from deep thinking sessions. It takes a few days' work to turn that few paragraphs into a solid article.