Should AI Augment or Replace Human Intelligence?
In the recent Human-Centered Artificial Intelligence symposium held at Stanford yesterday, the question of whether AI should augment or replace human intelligence kept coming up. It is probably the most important question going forward for all AI practitioners and research scientists.
I observed that a few panelists seemed to stagger a bit on the question as if indicating that the answer to this is not clear at the moment. I guess that is the point of the HAI initiative, yet no one in the panel had convincing answers. Alarming? Who is in charge of the conversation? What type of curriculums, frameworks, programming tools, laws, and protocols should we establish to better address the matter?
Honestly, as an engineer and researcher in the field, I don’t actually put attention to these questions when building and training neural networks. So the first big issue that needs to be solved is to raise awareness through improved curriculums I suppose. But is that enough? How can you actually train an engineer to be conscious of these issues all the time? Tough!
Researchers are still figuring out why and how these AI systems work in the first place. So this leaves no room to address the elephant in the room. How this mindset changes over the next couple of years will potentially dictate how much we know about AI technology at its true impact on society. We need all the experts and students we can have on board. It a race! Let’s encourage each other and most importantly let’s responsibly communicate our ideas and be more inclusive through active participation in different communities.
I don’t have much to opinionate on this matter, but I posted this question to raise awareness as I believe everyone from this community can benefit from such conversation. We are responsible for what we build at the end of the day.
As an exercise, let’s start by trying to answer the following: What are the pros and cons of replacing human intelligence with AI intelligence and what are the pros and cons of augmenting human intelligence with AI intelligence? I have a feeling that the answers to these questions are based on subjective experience and the field you are working in. It doesn’t matter, just attempting to answer these questions is a good start and a win for all. Consider it a take-home!
Full video of the symposium below:
