Bottleneck

Our human interface with reality

Category Archives: Learning

Elon Musk thinks that accessing the brain directly will enable us to interact faster with AI. He is making a fundamental error.

Speaking about AI at the Code Conference 2016, he said: “Constrained by input output”, “If we can create a high-bandwidth neural interface with your digital self”, “Access directly to cortex”,  “How to establish a high bandwidth neural interface”

Unfortunately Elon Musk does not realise that the bottleneck is the brain’s ability to integrate new information. Our sensors already have greater input capacity than our biology based brains can process, so unless we replace the CPU (the brain), we will never be superhumans.

What do we really know about our human interface – Article in Medium

Elon Musk speaks:

 

What do we really know about our human interface?

The truth about our learning bottleneck

Posted an article on Medium

 

We often hear arguments about the percentage that we are defined by the nature we are born with, and the percentage that we are defined by our experiences from our birth onward. The two sides sometimes agree a truce saying its 50% of each, which I think is a way of avoiding examining what is meant and in what context. So here are a few thoughts and observations:

Nature: If we believe that evolution has largely directed our genetic make up, then we must expect any general trends to be primarily optimised for early man, rather than for civilised 21st century man, indeed the majority of our DNA is identical to that of our earliest Homo sapiens ancestors, and is remarkably similar to that of our ape relatives. The kind of pre-programmed behaviours that we are born with are those which maximise the survival of an ape in its first few years at most. Evolutionary forces have had little exposure to the more recent activities of our species, so we cannot expect evolution to have optimised specific activities such as flaking a flint tool or playing tennis, but could reasonable expect an optimisation of physical skill in general. What about differences or defects in our DNA? These certainly can effect our physical characteristics, and often do so conspicuously through genetic diseases. However there seems to be little evidence that aspects of psychological personality are encoded in our genes and inherited.

Nurture: A newborn baby knows almost nothing of the world it finds itself in, and must build an internal idea of what is out there through observation and experience. Being incapable of survival alone, it builds its idea of the world through interaction with parents and other individuals. It is increasingly acknowledged that the entire world that we experience, is an internal construct, our best guess to date of what is out there. If we consider just how rich that experience is for adults, then we can see that almost all our daily experience is derived from what we have learned in the intervening years. The quantity of information that we subsequently learn through our lifetime completely dwarfs that which we are born with. Our behaviours are therefore likely to be dominated by what we have learned since our birth.

It is interesting to compare the inherited information with which we are born, with the very basic operating system (or BIOS) installed within a PC during its manufacture. When a PC is first switched on, the BIOS provides it with sufficient intelligence to be able to subsequently load and run a complex operating system such as Windows, and various additional programs for dealing with email or word processing. Like a new-born baby, basic PCs are fairly dumb but have huge potential. While the physical architecture (and technology) of most PCs is almost identical (as is also the case with humans), a PC can only go on to perform highly complex tasks when it has acquired such programs and additional data. The difference is that until now PCs are almost entirely force-fed with software and information, while exploration and discovery places a huge role in human learning.

So my take on this is that our specific behaviours are almost completely dominated by nurture, while our nature may influence our tendencies. Of course my deductive software may be wrong.

Letter to New Scientist in response to an article  Titled: “Into the minds of babes” by Anil Ananthaswamy. Issue: 2983, Pages: 40-43

Letter to New Scientist in response to an article
Titled: “Into the minds of babes” by Anil Ananthaswamy. Issue: 2983, Pages: 40-43