I don't think so. What we are doing is creating incrementally more sophisticated algorithms that can operate faster than before on larger data sets than before. It is what we have had for quite some time just now on a bigger scale. The self-improvement is interesting but it's still just a dumb machine doing exactly what it is programmed to do with no capability to expand to something greater.

We currently do not understand how the incredibly complex human consciousness works so we cannot begin to imagine what it will take to someday replicate it. It's a hard problem that is stalled in many ways as evidenced by many competing theories that lack evidence to know which one is correct.

Can a chatbot feel like a human to the average person? Sometimes. Is it the next step to creating a real intelligence that can exceed its programming? Maybe in the way that transistors led to personal computers -- a small step on a very very long journey.