Daniel Susskind – @danielsusskind – co-Author of ‘The future of the professions’
Daniel Susskind co-authored ’The future of the professions’ which interviewed 100 people across 8 professions to try to make sense of the future of work. They identified two possible futures:
- a more productive version of what we have today. For example, physicians using new technology to communicate with patients in different ways
- a radically different future where technology doesn’t just streamline and optimise, rather professional roles are replaced by radically new and different uses of technology
Daniel argued that scenario (2) is the more likely outcome.
In a print-based industrial society, we have a need for professionals to solve specific problems. This has resulted in the development of professions very focused on a specific niche (e.g. medical, legal, accounting etc.). Given that we no longer live in a print-based industrial society, this is ripe for change – especially given that many professions are underperforming.
There are many great examples of people who are doing things very differently. For example Harvard online has attracted more students in one year than have attended Harvard in the whole of its history. Khan Academy has rapidly developed into a great source of learning.
Google DeepMind is advancing AI 10 years more quickly that people feel possible. Medical apps are becoming more and more popular, and WebMD is becoming a source of trusted medical advice for many people. Technology is even impacting the word of divinity, with the Roman Catholic church endorsing an app for confession.
Amongst the trends that their research identified are…
- The more for less challenge; professionals are being asked to do more and more with less resources at their disposal
- New competition; the competitors that will kill you are the ones who don’t look like you
- There is a growing trend to move away from bespoke service and offer something off the shelf
- Decomposition – professional work being broken down into its component parts and than can be done by different people (or tech)
Professional work started as a form of craft, then starts to get standardised, then systematised. We have several options for what we do next:
- Make it available and charge for it
- Make it available online and don’t charge for it (e.g. governments and charities)
- Make it available and – in the spirit of wikipedia – hold as a commons resource
1 – exponential growth
Moore’s Law predicted that processing power and data storage capability would double every two years (as a consequence of the underlying technology). There is predicted to be a 10-billion times reduction in the cost of processing power in the first 50 years of technology.
2 – increasingly capable
we can use this processing power well…
big data – most things we do now leave a trail of data, and big data can use that data to produce insights. E.g. Lex Machina is now believed to be as reliable as a patent attorney at predicting case outcomes, based on analysis of millions of data points.
Robotics – e.g. all of the work going in to creating driverless cars, despite the fact that driving cars was always predicted to be something that could not be automated.
3 – increasingly pervasive
4 – increasingly connected
There is no finishing point… the technology that we have today is the worst we’re ever going to see…. it will get better and more powerful
AI went through a first wave of interviewing people with deep expertise and then codifying and implementing the decision-making process, often based on mapping decision trees. These systems were developed for applications in law and taxation etc,, but didn’t take off massively as they were costly to develop, was in competition to what the profession actually wanted, but then was superseded by that arrival of the web.
Garry Kasparov’s defeat by Deep Blue was pivotal in that Deep Blue wasn’t ‘better’ at chess than Kasparov, but it defeated him by brute force of being able to process and remember more alternative scenarios. In a similar way, professionals are at threat from AI not because computers will be ‘better’ than them, but because they will compete differently. For example, professionals are often called upon and respected for their judgement whereas the real problem is uncertainty, and AI can come up with alternative (and more effective) ways to deal with uncertainty. They won’t provide better judgement, but they will provide better outcomes.
what does it mean for us? At the moment, we tend to think about jobs but we probably need to start thinking about tasks as there will be a decomposition of our roles. It is unlikely that out jobs ‘will be replaced by robots’ but many of the tasks we do might well be automated or done very differently. By 2020 we need, as professionals, to either compete with AI or we can get ahead of the curve and build the machines! This second strategy of being the person who understands and designs AI solutions seems a more attractive option.
What lies beyond the professions? How do we produce and distribute practical expertise to where it is needed? The answer used to be through the professions. The Internet gives us many more options including…
- Networked experts – workers on tap
- knowledge engineering
- communities of experience – recipients of professional work (e.g. PatientsLikeMe is a forum which has lots of discussions about medicine without any clinicians involved).
It isn’t helpful to pretend that this isn’t happening to us, or to try to protect ‘our patch’. Things we can do involve:
- explore new roles
- explore new models
- start with a blank sheet of paper
- change your mindset
There are, however, loads of opportunities. Don’t go into professions to do what your parents or grandparents did. Rather, go into them because you really want to make a difference. For example, go into law because you care about justice, go into medicine because you are interested in overcoming disease. We now have amazing new opportunities to do more if we are agnostic about the solutions and not wedded to how we did things in the past.
This was live-blogged during a session at the CIPD Learning & Development Show 2017 – I’ve tried to capture a faithful summary of the highlights for me but my own bias, views – and the odd typo – might well creep in. I’ve also curated the story of the session as told through the tweets of the attendees (you might need to tap ‘load more tweets’ to see the full story):