When Will True Human-Like AI Arrive?

I’ve done a lot of research on this and talked to numerous AI experts. There is no consensus on when Artificial General Intelligence (AGI) will arrive or how fast it will evolve. I can tell you that we’re still a ways off from developing true human-like AI. It’s not coming tomorrow.

What everyone can agree upon is that we keep raising the bar for AGI. It’s no longer a simple Turning Test using a chatbot. That goal for AGI is here already–or nearly so.

It really depends on how you define AGI. Human beings are quite complex and we have many facets to our intelligence. We’re still figuring out how the human brain actually works. It’s hard to capture everything when we don’t even understand our own intelligence.

That said, I think we’ll get true AGI that can replicate human thought in around 20 years. Between now and then, we’ll see a steady advance, along with many dead ends.

What we may need is an entirely new model that goes beyond neural networks. That’s what some computer scientists believe is necessary for true AGI.

That said, regardless of what happens, AI is going to have a huge impact in the coming decades. Lifelong ML is one of many developments taking place right now. I’ve heard of others. For example MIT is working on making AI curious. They want AI to learn to learn. And Google’s Deep Mind is experimenting with GANs and other approaches.

Whatever happens, it’s going to be an exciting ride. So buckle your seat belt.