This sounds like the periodic resurgance of interest in neural nets that has been going on for decades. While our brains do well with neural nets, they use a lot of neurons and a lot of synapses to figure out patterns. Once the pattern is understood: "IF the temperature rises quickly THEN ring the fire alarm", it can be programmed on a trivially simple logic gate. I think that neural nets often are used to address problems which are not well understood and the programming of the neural nets is not well understood either. Knowing what we want and finding an efficient way to get to the intended destination is more likely to lead to programming success.
It's great to see experimentaiton in architectures. However these usually come at the cost of needing a whole new programming approach which means outside of the chip devekopers, no one knows how to write software for this beast yet.
What are the engineering and design challenges in creating successful IoT devices? These devices are usually small, resource-constrained electronics designed to sense, collect, send, and/or interpret data. Some of the devices need to be smart enough to act upon data in real time, 24/7. Specifically the guests will discuss sensors, security, and lessons from IoT deployments.