Peter Cochrane: Why we should stop worrying and start engaging with AI
We need a generalised species of AI that is all-knowing, good at everything, and tuned to our individual limitations and fallibilities
My involvement with AI started in the 1980s at an R&D lab pushing the boundaries of the technology and tech driven change. At the time, AI went hand-in-hand with artificial life, robotics, wearables and implants - but all at an early and crude state of capability. It was at this juncture that I stumbled across new truisms that have become foundational in systems thinking today. Perhaps the most important of these are:
- We are the only species that optimises anything
- Mother Nature and life seek to minimise entropy
- We design complex systems to address simple tasks, whilst Mother Nature evolves simple entities/systems to deal with the highly complex
And looking more deeply:
- Life is an emergent property of complexity
- Intelligence is an emergent property of life
- Sentience is an emergent property of intelligence
- All forms of life exhibit some level of intelligence
- Almost all lifeforms/intelligences are sentient
These all came from the study of biology, evolution, man made hardware and software facsimiles, and the unpicking of the core processes. One outcome of the 70-year progression of AI has been an increasing symbiosis that sees me wearing three different interactive AI systems that are augmented in my home, office and car, with evermore invisible variants embedded in my appliances, networks, and devices. However, I do not interface with this last category as I might with Alexa, Google, Cortana and Siri. They operate in the background, silently looking after my day-to-day operations, improving my life and efficiency, by seamlessly fulfilling my needs with near faultless accuracy.
Despite these successes, AI remains domain specific, outstanding at what it does, but limited by scope. We actually need a generalised species of AI that is all-knowing, good at everything, and tuned to our individual limitations and fallibilities. However, it is our complex nature: in the analogue way we think, act, react, and communicate, that sees this particular dream still some way off. Nonetheless, before us is an exciting and opportunistic journey!
Among my ‘favourite' AI frustrations and irritations is the multiplicity of interfaces and ‘command lines', voices, and the limited ability to master my accent. I am also awash with irrelevant search results! So, in my ‘wish list' top-ten is a search engine that exploits AI to monitor what I am working on, who I am meeting and talking with, and the content of all interactions, so that my search for ‘protein folding and cancer' zeros in on my specific agenda and not some global average derived from the entire user population. I would also value an effective ranking of academic publications steering me toward the high ‘Reward-on-Reading' content rather than the predominant dross, trivial, poor and inconsequential documents that make searching so very tiresome and irritating.
For as long as I can remember, the media have worked had to highlight the ‘potential' downside of AI. Hollywood in particular excels with such characterisations such as Terminator. In contrast, home, office and shop floor sees AI acceptance to be steady and sure-footed, as does the integration with robotics and industrial systems.
It turns out that many of the negative authorities promoting their worries have never engaged in any design-and-build or basic use of AI for real.
They also seem to omit the very significant upside. The numbers of people being saved by AI in the medical sector, our increasingly safe transportation systems, our devices and appliances of unrivalled quality and performance, along with our clothing and food standards all rely on advancing AI.
My advice is to stop worrying, and start engaging with AI, so our systems can learn faster and so can we! The alternative is to be left behind, to become increasingly disadvantaged.
Peter Cochrane OBE is professor of sentient systems at the University of Suffolk, UK