London is full of pleasant surprises.
Yesterday evening, I travelled to The Book Club in Shoreditch, EC2A, and made my way to the social area downstairs. What’s your name? asked the person at the door. I gave my name, and in return received a stick-on badge saying
Hi, I’m David.
Talk to me about the future of humanity!
I was impressed. How do they know I like to talk to people about the future of humanity?
Then I remembered that the whole event I was attending was under the aegis of a newly formed group calling itself “Future Human“. It was their third meeting, over the course of just a few weeks – but the first I had heard about (and decided to attend). Everyone’s badge had the same message. About 120 people crammed into the downstairs room – making it standing room only (since there were only around 60 seats). Apart from the shortage of seats, the event was well run, with good use of roaming mikes from the floor.
The event started with a quick-fire entertaining presentation by author and sci-fi expert Sam Jordison. His opening question was blunt:
What can you do that a computer can’t do?
He then listed lots of occupations from the past which technology had rendered obsolete. Since one of my grandfathers was the village blacksmith, I found a personal resonance with this point. It will soon be the same for many existing professions, Sam said: computers are becoming better and better at all sorts of tasks which previously would have required creative human input. Journalism is particularly under threat. Likewise accountancy. And so on, and so on.
In general terms, that’s a thesis I agree with. For example, I anticipate a time before long when human drivers will be replaced by safer robot alternatives.
I quibble with the implication that, as existing jobs are automated, there will be no jobs left for humans to do. Instead, I see that lots of new occupations will become important. “Shape of Jobs to Come”, a report (PDF) by Fast Future Research, describes 20 jobs that people could be doing in the next 20 years:
- Body part maker
- Nano-medic
- Pharmer of genetically engineered crops and livestock
- Old age wellness manager/consultant
- Memory augmentation surgeon
- ‘New science’ ethicist
- Space pilots, tour guides and architects
- Vertical farmers
- Climate change reversal specialist
- Quarantine enforcer
- Weather modification police
- Virtual lawyer
- Avatar manager / devotees / virtual teachers
- Alternative vehicle developers
- Narrowcasters
- Waste data handler
- Virtual clutter organiser
- Time broker / Time bank trader
- Social ‘networking’ worker
- Personal branders
(See the original report for explanations of some of these unusual occupation names!)
In other words, as technology improves to remove existing occupations, new occupations will become significant – occupations that build in unpredictable ways on top of new technology.
But only up to a point. In the larger picture, I agree with Sam’s point that even these new jobs will quickly come under the scope of rapidly improving automation. The lifetime of occupations will shorten and shorten. And people will typically spend fewer hours working each week (on paid tasks).
Is this a worry? Yes, if we assume that we need to work long hours, to justify our existence, or to earn sufficient income to look after our families. But I disagree with these assumptions. Improved technology, wisely managed, should be able to result, not just in less labour left over for humans to do, but also in great material abundance – plenty of energy, food, and other resources for everyone. We’ll become able – at last – to spend more of our time on activities that we deeply enjoy.
The panel discussion that followed touched on many of these points. The panellists – Peter Kirwan from Wired, Victor Henning from Mendeley, and Carsten Sorensen and Jannis Kallinikos from the London School of Economics – sounded lots of notes of optimism:
- We shouldn’t create unnecessary distinctions between “human” and “machine”. After all, humans are kinds of machines too (“meat machines“);
- The best kind of intelligence combines human elements and machine elements – in what Google have called “hybrid intelligence“;
- Rather than worrying about computers displacing humans, we can envisage computers augmenting humans;
- In case computers become troublesome, we should be able to regulate them, or even to switch them off.
Again, in general terms, these are points I agree with. However, I believe these tasks will be much harder to accomplish than the panel implied. To that extent, I believe that the panel were too optimistic.
After all, if we can barely regulate rapidly changing financial systems, we’ll surely find it even harder to regulate rapidly changing AI systems. Before we’ve been able to work out if such-and-such an automated system is an improvement on its predecessors, that system may have caused too many rapid irreversible changes.
Worse, there could be a hard-to-estimate “critical mass” effect. Rapidly accumulating intelligent automation is potentially akin to accumulating nuclear material until it unexpectedly reaches an irreversible critical mass. The resulting “super cloud” system will presumably state very convincing arguments to us, for why such and such changes in regulations make great sense. The result could be outstandingly good – but equally, it could be outstandingly bad.
Moreover, it’s likely to prove very hard to “switch off the Internet” (or “switch off Google”). We’ll be so dependent on the Internet that we’ll be unable to disconnect it, even though we recognise there are bad consequences,
If all of this happens in slow motion, we would be OK. We’d be able to review it and debug it in real time. However, the lessons from the recent economic crisis is that these changes can take place almost too quickly for human governments to intervene. That’s why we need to ensure, ahead of time, that we have a good understanding of what’s happening. And that’s why there should be lots more discussions of the sort that took place at Future Human last night.
The final question from the floor raised a great point: why isn’t this whole subject receiving prominence in the current UK general election debates? My answer: It’s down to those of us who do see the coming problems to ensure that the issues get escalated appropriately.
Footnote: Regular readers will not be surprised if I point out, at this stage, that many of these same topics will be covered in the Humanity+ UK2010 event happening in Conway Hall, Holborn, London, on Saturday 24 April. The panellists at the Future Human event were good, but I believe that the H+UK speakers will be even better!