Unlocking Behaviour Change

In the previous episode, Robert castigated economics and economists' concept of 'utility' as fundamentally flawed, leading to models whose predictions keep being wrong. This week, Robert sketches out a way of thinking about utility that could improve economic modelling, yield more accurate predictions, and provide a more rational basis for macro-economic policy. A bold claim ... but this is a podcast, not an academic paper! Want more? Check out our latest books:   Energise: The Secrets Of Motivation: https://shorturl.at/bbA7y  React: Harness Your Animal Brain: https://shorturl.at/yiFQ9   Sign up to Robert’s Substack: https://robertwest100.substack.com/   Subscribe to the UBC YouTube channel: https://www.youtube.com/@unlockingbehaviourchange   Ask a question at: https://www.unlockingbehaviourchange.com 

Robert has a bone to pick with economic theory . . . In fact, he wants to pick its bones clean and start over with a much better theory for everyone! Join the duo as they discuss what economic theory gets wrong and how it might be put right . . .   Want more? Check out our latest books:   Energise: The Secrets Of Motivation: https://shorturl.at/bbA7y  React: Harness Your Animal Brain: https://shorturl.at/yiFQ9   Sign up to Robert’s Substack: https://robertwest100.substack.com/   Subscribe to the UBC YouTube channel: https://www.youtube.com/@unlockingbehaviourchange   Ask a question at: https://www.unlockingbehaviourchange.com  

Derisory amounts of money paid for a piece of work called an 'honorarium' seem ok when if they were a 'fee', it would be insulting! Robert and Jamie discuss how we react differently to the same things when they are classified and labelled differently. This includes clinical diagnoses, where receiving a 'diagnosis' can be a huge relief to sufferers but can also carry stigma. Want more? Check out our latest books:   Energise: The Secrets Of Motivation: https://shorturl.at/bbA7y  React: Harness Your Animal Brain: https://shorturl.at/yiFQ9   Sign up to Robert’s Substack: https://robertwest100.substack.com/   Subscribe to the UBC YouTube channel: https://www.youtube.com/@unlockingbehaviourchange   Ask a question at: https://www.unlockingbehaviourchange.com    

Robert and Jamie discuss the crucial question of how we classify and label things: one of the most basic things that we do in our mental and social world. But in an infinitely complex universe, putting things into boxes is fraught with difficulties, and classifications lie at the heart of important disputes. Want more? Check out our latest books:   Energise: The Secrets Of Motivation: https://shorturl.at/bbA7y  React: Harness Your Animal Brain: https://shorturl.at/yiFQ9   Sign up to Robert’s Substack: https://robertwest100.substack.com/   Subscribe to the UBC YouTube channel: https://www.youtube.com/@unlockingbehaviourchange   Ask a question at: https://www.unlockingbehaviourchange.com  

Robert and Jamie pay tribute to Dolly Parton - and Jamie plugs his video on her which can be found here - and then discuss how our intuitions about probability can be seriously misleading - leading to disastrous decisions. Robert explains why decision-makers must understand numerical probability. Robert claims that 'stats beat experience every time (almost)'! If we can estimate probability based on evidence, we should always do that. If we can't work out the probability confidently, then trying to do so tells us something important about our ignorance. After this episode was recorded, it was widely reported that an AI insider estimated the probability of AI killing all humans within 10 years was 10 per cent! So it is being claimed that there is a 1 in 10 chance that human life will be obliterated by AI in the next 10 years. And Geoffrey Hinton (often acclaimed as the godfather of AI) said that that was not an unreasonable figure.  But when asked about what scenarios could this involve, he struggled. He should have said something like 'We cannot estimate a numerical probability for this outcome but we should act on the possibility that it could happen.' So, while we should always try to work out numerical probabilities, if there is no sound basis for doing so we should express likelihoods in non-numeric terms.