My name’s Ben, and I’m a software engineer with a background in neuroscience and philosophy. I read On Intelligence a looooong time ago, and Jeff’s ideas always stuck with me as a logical approach to achieving human-like AI.
I’m very interested in developing a hardware platform with sufficient sensorimotor capabilities to test TBP Monty and see if we can get a robot to perform similarly to human infants on some of the earliest developmental cognitive tasks.
I’m really looking forward to learning from this awesome group of like-minded folks. Thanks for having me!
Hi Ben! It’s great to hear about your interest in our approach, and I’m hoping to hear more about your work on a sensorimotor hardware platform and how this might integrate with Monty in the future! Always happy to answer questions as they come up.
Welcome
I am at an advanced stage of developing such a platform with a similar goal in mind.
But human infants are the (current) pinnacle of evolution with sensorimotor systems far beyond anything we can emulate with today’s technology.
I propose to start with much simpler artificial creatures. There is a great deal to learn about how to encode behaviour, natural learning and goal seeking in an artificial neural network.
I was hoping to use Monty to control the sensorimotor platform I am designing but from what I have read about it I don’t think the match is ideal. Someone may correct me, but I think object recognition is quite a long way up the evolutionary tree of understanding and intelligence. There is a lot to be learned before we get that far. However, if someone wants to try !
Please contact me directly if you would like to know more about the hardware platform.
