Artificial intelligence is incredibly trendy at the moment, with barely a day going by without a story of some kind about the increasing ability of things to act with a degree of intelligence and automation.
Whether it’s driverless cars, algorithmic trading, automated drones or the Watson super computer, there is no shortage of interesting examples of the improvement in AI in recent years.
Most of these attempts at improving AI have rested on the desire to enhance the capabilities of individual machines however. An alternative approach, as advocated by AI doyen Nick Bostrom, is to ensure greater connectivity, both between humans and other humans, but also between humans and machines.
Why connectivity matters
Whilst the web has done much to improve this connectivity, it still often fails at the task of directing us to the best content as it struggles to find relevance in the huge amount of information shared online each day.
The potential for this kind of collective intelligence is evident through the many crowdsourcing projects that have flourished around the world. It’s evident in the way everything from markets down to teams support a collective gathering of ideas towards a common goal.
This kind of intelligence is exceptionally good at solving challenges that can easily be broken down into smaller chunks that can be tackled both independently and in parallel. The kind of work we’ve seen so well in open source communities or in various citizen science projects.
Enhancing our collective intelligence
To improve our collective intelligence, we can focus on enhancing the intelligence of individual components, but we can also enhance how well they function together, or their organization if you will.
Here, Bostrom advocates that efforts should focus on enhancing our norms and conventions. We should look to improve the epistemic institutions that allow for knowledge to be shared, whether that’s the scientific journal, the patent system, our way of peer reviewing papers or our ways of disseminating the latest thinking more quickly.
The last few years have seen tremendous gains in securing a much greater proportion of the worlds intellectual capacity towards productive problem solving.
This kind of intelligence enhancement doesn’t get quite the same kudos as the cool work being done with AI, but it is nonetheless just as important for our organizations, and indeed our species.
AI has failed to even understand the nature of the field in which it operates. A silly field of study, for silly people. That's it, really.
You could argue that man is just a stepping stone on the way to the machine future. By starting the process and inventing and designing computers we are in effect initiating our own demise. In the future machines will be far more sophisticated than anything yet devised and become self replicating. Eventually they will probably take over. We may survive – but only as curiosities or pets – a bit like the family dog.