Discussion Forums » Topic of the Month
Topic for April/May 2023 - AI and re-shaping our society
-
The topic for April/May is AI and Re-shaping our Society.
How can we both exploit and control AI and other innovation to improve society for the ‘good of all’?
Introduction
There is no doubt that AI is by far the most significant development humanity has made to date, with immense implications, both good and bad. It will become totally pervasive in every aspect of life. On the one hand the disruptive consequences and opportunities are so great, that we have the opportunity (and need) to reshape society and our economy to what we want it to be. On the other hand, there are also immense possible harms and some see an existential risk to the human race if we don’t control it.
It is important to realise that we are still in the foothills of AI. It is also important to realise that AI isn’t just one thing, there are a multiplicity of approaches and techniques. At a top level there is also an important distinction between ‘narrow AI’ which is what we are seeing deployed at the moment and confined to a particular domain and Artificial General Intelligence (AGI) which is illusively just round the corner. AGI will be able to deploy its learning and concepts developed in one environment in a previously unseen environment just as we humans do.
All the focus of AI by Government to date is around organisation benefits (Public and Private Sector). Probably42 has been inputting to Government and other influencers for the last few years that they appear to be failing to look at:
- Using AI to improve society for the good of all – taking a societal perspective, not just a business-led perspective
- Managing and overcoming the downsides and unintended consequences of AI deployment, especially in relation to labour market disruption
The alarmist media reporting of the open letter from Elon Musk and others in the AI community about pausing the most advanced deployment may serve to move Government out of its complacency but it may also drown out a more thoughtful debate. This agenda is an opportunity to do some ‘blue sky thinking’ ourselves about our future society and to develop our existing thinking and ideas in both the above areas.
Agenda
What are our top-of-the-mind feelings about AI? What are the main issues and opportunities that we see?
Do we have a feel for what AI is likely to be able to achieve? Do we have any specific predictions for major developments? How important is it to explain AI to the public more fully?
How far can AI go? Will it overtake human ability in every sphere of life? Do we have to get used to AI being more intelligent than us? Are we just sophisticated pattern-matching machines that will be surpassed.
‘Blue sky thinking’. What sort of society would we like in 20-30 years’ time, what is our vision for the future? What phases might AI go through and what game-changers might there be?
Assume that AI can: improve the quality of everything we do; can bring order of magnitude improvements in productivity; can dramatically increase the speed of research, innovation and deployment of that innovation; can manage the production and delivery of our basic needs; can solve problems that have eluded us; can understand and exhibit empathy.
Consider improvements to:
Our individual and national wellbeing and that of future generations
Our lifestyles – enjoyment, fulfilment, contribution, freedom from fears
Our health and fitness
Our communities and social environment
Our work and our economy
Our democracy
Our political system
Our international relations and the international order
What can get in the way of the society we want? What potential downsides do we foresee? How can we tackle or mitigate these downsides?
Responsibility of supplying and deploying organisations
Regulation and ethical guidelines
Societal acceptance
Ownership of AI and the benefits of AI
Depletion of natural resources
Existing economic model unable to cope
What is work?
Digital dependence, external threats and malicious intent
Criminal use of AI proceeds faster than benevolent use
Need for transparency and accountability
Managing development and implementation of solutions
Disparity and Division
Autonomous AI weapons
Existential threats from scientific advances
-
The output of our discussions is now available. Unusually we have divided this into two separate documents as each deserves their own focus: