Artificial Intelligence (AI) V Physical Intelligence (PI): Collaborative Intelligence (CI)?

There are many aspects to how the human species is going to interface itself with computers, its software, and physical technologies and resources around the planet. What I see is a convergence between artificial intelligence and physical intelligence.

What do I mean by AI? There are a range of interpretations of what this means in 'reality'. However, the 'reality' is that we are using more and more mathematical algorithms to solve 'real world (physical)' problems. These algorithms are enshrined in software code based on a physical system such as smartphones, laptops, desktops and cloud-based server farms. I was at a Summer solstice party a couple of years ago and I spoke to a bearded computer science student who described AI as basically using software, based on computers, to solve problems. That could be as simple as using a sensor such as a moisture meter to tell an Arduino microcontroller to turn on the water in a wicking bed to water the plants. It could be as complex as modelling the climate for change, mining a cryptocurrency and searching out new exoplanets in the universe for human habitation.

What do I mean by PI? An old film director, known as 'Smokin' Joe' who directed the movie 'Mystery Train' lived in a guesthouse at Nakano in Tokyo. I was staying there in 2006 while I was teaching English as a Foreign Language. We had a number of conversation about physical intelligence. What Joe was trying to get to was the interface/relationship between people and planet. He described how people were losing their ability to think about how they interacted with the physical environment losing their connections to physical resources in the face of globalisation. How is it that we could be so connected to everyone else yet many so dimly aware of our own impacts on the physical resources of the planet? At that time, Japan was importing 90% of its soya beans and starting the smartphone revolution. I was teaching many students at that time who were coding apps for smartphones. A few years later I went and studied a Diploma of Permaculture and realised then that scarcity of resources was manufactured and that design science embedded in human/physical systems was the main driver of future sustainable societies.

So, the question on my mind is interface. I've been playing with the hardware and software interface in my music studio for 20 years. I have seen a steady trend in integration between hardware and software over that time. It's at a point now where a company called 'Native Instruments' have achieved such tight software-hardware integration that one doesn't even need to look at a computer screen while producing electronic music. Access to all the sounds is at your fingertips. Smartphones are no different - apps are getting smarter than we are and information, or intelligence, is available in an instant. Big data is collected by most software companies and more and more sophisticated algorithms are being coded to create increasingly efficient ways to mine data and get the results the user might be looking for.

But, is efficiency a substitute for sustainable technology? I would argue that efficiency can be more sustainable in certain applications that might save energy usage, such as smart grids and variable controlled greenhouses/warehouses eg. 'Plenty' and shipping containers eg. 'Terrafarms'. However, if the designs of those systems are creating an increased demand in the physical resources based in those systems, are we creating an overall larger ecological (physical) footprint? As we see a growing middle class around the world with more sophisticated consumer demands, AI applications may float more boats and make life more convenient with less physical resources due to efficiency increases. More importantly resources need to be used more than once and have multiple functions. I have found that most aspects to biochar technology cannot be substituted with software algorithms telling me what to do or even doing it for me. A software program can't harvest the biomass, the raw product for biochar, could integrate with a laser cutting machine with additional hardware, can automate expensive and complicated kilns, can't operate simple kilns such as the 'Carbon' to produce the biochar and can't physically distribute the biochar to its various cascades such as animal feed and bedding, construction (eg.hempcharcrete), water filtration and soil fertility. But I can write a physical algorithm aka operating/usage instructions to use the resource within a designed ecosystem. 

I imagine that there will be greater convergence with AI and biochar in the future but ultimately automation and efficiency algorithms can't solve many of the physical problems posed by managing physical resources such as biochar. Arguably, it is only a matter of time for AI and big data to catch up with PI, but will probably never completely take it over - just move closer to a converged state of mind eg. Mapping. There's plenty of ways people can collaborate with mapping on the internet and it seems to be the ultimate way to communicate between the mind(s) and machine(s). Mapstory.org is one example though it's still in beta and contains many incomplete datasets but houses some pretty stunning maps. Google 'MyMaps' (also shareable) seems to be an easy way to collaborate in mapping something basic like georeferenced points eg. Biochar kilns, as long as you've got a Google account. These can probably be embedded in a website too. Google Maps 'Mashups' can also be easily created with many tools freely available on the internet. Call me biased, but I love a good map though as a wise man once told me 'The map is not the territory'. Enough said.