Daily Archives

2 Articles

Posted by Tom Cheesewright on

Will a robot hold your hand? Technology in the future of care

Will a robot hold your hand? Technology in the future of care

This week I gave a short address at a panel debate on the future of care, hosted by frequent client, the rather forward-thinking Freeths solicitors. Here’s what I said. Or at least, what I intended to say when I wrote my script.

When you’re looking to the future, you need to understand two factors. Firstly, what are the pressure points that the sector you are examining is facing today? In my experience, these are always the points of failure or opportunity where change happens first. Secondly, you need to understand what is causing that change. What are the major trends?

For me, the best way to understand those trends is to look at technology. Technology is the means by which we enact change. Described in the broadest terms, technology the application of our understanding of the world. From the first rock a caveman or woman sharpened, through language, to the modern smartphone.

Pressure Points

You’re all familiar with the pressure points facing the care sector today. Rising demand, declining budgets, and a catastrophic lack of skilled permanent staff. The demographic changes we’re facing mean the areas with the most demand often have the least access to staff.

Throughout history, technology has been deployed to address issues like this. If you can mechanise a process, you can repeat it at lower cost and higher frequency with fewer staff. It doesn’t matter if it’s a steam powered loom or a computerised call centre. But can we really apply technology to replace skilled people in care? For me, the answer is a very strongly qualified ‘yes’.

First of all, let me tell you what we won’t see, which is some form of robot nurse, capable of all the things a person can do. Human beings are extraordinarily adaptable, both physically and mentally, and this flexibility is enormously challenging — and expensive — to try to replicate.

Rather, what we will see is a much more distributed and pervasive suite of technologies designed to help people support themselves better, for longer. To smooth their entry into more formal care settings. And to assist them in overcoming their challenges throughout.


I’ll start with the first category — what we might loosely term remote monitoring.

Have all heard the buzzphrase, the internet of things? How about Moore’s Law? What this really means is that the price of adding computing power and connectivity to just about anything has collapsed over the last fifty years — even the last decade. And at the same time, the accessibility of the devices and the knowledge to do this has dramatically increased. You can now, with really only a junior school education, programme a machine to monitor basic environmental factors such as temperature and humidity, and send that information off over the Internet. That device might cost you five pounds.

Scale this up and add some grown-up intelligence, and you can start to monitor more things: activity, energy consumption, carbon dioxide levels, doors opening and closing. You can know if someone is active and what sort of conditions they’re living in.

None of this is new, you might say. We’ve been able to get this information over a phone line for years. Sure. But two things have changed. Firstly, the cost: it now costs less than £5 a month to monitor basic environment factors and activity in someone’s home. The hardware is so cheap that there is no up-front cost. And it’s all battery powered so you don’t even need a specialist installer. It can just be stuck to a wall or ceiling, just like the fire brigade installs smoke alarms.

Secondly, intelligence. Computing power is so cheap now that we can throw enormous amounts at monitoring and interpreting this data for very little money. To find the exceptions, the behaviour changes. To identify the risk factors and intervene early — and cheaply — rather than later when the issue is acute.

Over the next few years I think we will see a massive expansion in the application of home monitoring technologies, not just by concerned children but by the state in a bid to manage the costs of care.


The second class of technology I want to talk about is robots. This is perhaps the area that has caused most consternation when its application in the care sector is discussed. People don’t like the idea of a warm nurse being replaced by a cold machine. And I understand that, but we shouldn’t leap to the conclusion that all automata in a care setting are bad.

I have a sideline reviewing gadgets for the BBC, and before Christmas I got sent a Cozmo to play with. Did anyone’s kids get one of these for Christmas? Lucky kids. This is a tiny toy that looks a little bit like a cross between Wall-E and a forklift truck. It borrows the incredible processing power of your smartphone to approximate an artificial intelligence. It can recognise your face and play a series of games with you, using some special cubes that come with it.

The most interesting thing about this toy for me was not the level of tech packed into its tiny shell, but the way that my children projected an identity onto it. This shouldn’t have surprised me. It’s a very human trait: we anthropomorphise everything. Just look how much intelligence and personality we ascribe to our pets, or kids do to totally inanimate dolls.

In a very short space of time my kids created a connection to Cozmo and clearly felt a real sense of reward from interactions with it. The same behaviour has been witnessed in adults interacting with automata in a care setting.

Machines can’t care. But they can provide mental support and stimulation. They can answer questions, guide people, control the environment and entertainment, and increasingly, chat. We can even project a level of love and companionship onto them — even when we know deep down that they are not capable of reciprocating. Because this is clearly what the human brain does. We shouldn’t reject that possibility out of hand.

Augmentation: Physical and mental

What these robots can also do is collect information, store it, and replay it. This is something that we all struggle with, particularly as we age or if our mental faculties are starting to decline. There’s a serious opportunity for us to start to augment our minds with technology. In fact I’ve been arguing for a few years that the process has already started: we are all bionic now.

How many people used a smartphone to get here, looking up the time or location, using GPS and maps? I was born with a terrible memory and basically without a sense of direction, so the advent of such technology has been an absolute boon for me.

Imagine if you could make the interface to this information even more natural. So low friction that you barely notice where you end, and the machine begins. Take these inserts for the sole of your shoe, for example, which vibrate to tell you when to turn left or right. Imagine a verbal prompt through a bone-conducting earpiece. Imagine a digital overlay on your vision.

All of these things are real today or within a few years. They are still both expensive and a little rough around the edges, but that Moore’s Law I talked about will make them widely accessible. I’m betting that this technology is what ultimately replaces the smartphone.

This doesn’t help those with physical frailty of course, but here again, Moore’s Law is our friend. Just ten years ago, strength augmenting suits were the stuff of science-fiction and military fantasies. Now they are commercially available, both in commercial contexts and to help the paralysed to walk again. In another decade or two, as battery and motor technology continues to improve, it’s easy to see articulated walking frames helping people to recover mobility. We’ve already seen such a revolution happen: how many mobility scooters do you remember seeing 20 years ago?


Technology is not the answer to our care crisis. That requires political intervention to raise funding and wages, improve conditions, overcome the looming threat that Brexit presents, and to address the threat to employment and employment quality that technology also so clearly presents. But whether or not these interventions are made, technology represents an opportunity to improve care. To give people more self-sufficient lives for longer, to ensure earlier interventions when they are needed, and even to provide a level of companionship to those who need it. I’d argue that we need to overcome our squeamishness and embrace it.

Posted by Tom Cheesewright on

Political choices in the face of rising automation

Political choices in the face of rising automation

Human beings don’t act out of naked self interest. Given the choice, we prefer to pursue a fair outcome rather than one that leaves one side with everything. This was amongst the findings of Daniel Kahneman and Amos Tversky in their work that came to define the field of behavioural economics. (The pair are the subject of a new book, The Undoing Project, by Michael Lewis, author of The Big Short — well worth listening to his interview with Adam Buxton).

Large companies don’t have this compunction. Allowed to do so, they act in the interests only of their shareholders. In fact, they are compelled to do so. The only question is whether they are looking to maximise returns over the short term or the long term. The longer term the thinking, the more likely it seems the interests of shareholders and the rest of the planet might align, though this is hardly a firm rule.

That they are focused on profit to the frequent exclusion of social good isn’t a criticism of companies or their leaders. Much as we might dislike their behaviour sometimes, they’re playing the game by the rules. Given multiple opportunities, we’ve chosen not to change those rules and perhaps for good reason. There’s at least a solid argument to be made that the game to date has been a net positive for most of us.

The question is whether it will continue to be so in the future.

The great divide

While many might — legitimately — question whether the current technological revolution will be different to the past, in destroying more jobs than it creates, there are some worrying signs. Large companies now operate with many fewer employees than they used to, with Apple, Alphabet (Google) and Facebook all making north of $1m per employee. Though employment is at historically high levels in many places, the quality of the work being created is often low and highly precarious. Levels of self-employment have risen dramatically.

The risk is a growing division between rich and poor, with a shrinking number having access to the security of employment, or a freelance skillset that provides a reasonable income, and a growing number surviving on a mixture of benefits and gig-economy work.

This scenario is at least plausible based on the current track. What might policy responses be to such a reality?

For me there are two clear routes of intervention. We can regulate employers to force them to better look after the interests of their workers. Or we can let take a laissez-faire approach to regulation, but raise taxes to build a state system that supports people without or between employment.

Regulating employment

As the ongoing battle between Uber and its drivers shows, action is already being taken to reign in the more extreme employment practices. In fact, this is just one of many cases challenging the ‘contractor’ model common in the gig economy. Should Uber lose, it will be forced to ensure drivers receive minimum wage and other entitlements, and be subject to a much larger tax bill. But its drivers will also be subject to much tighter controls: no more logging on, as and when you want to.

Ensuring minimum standards for workers in the gig economy doesn’t come without costs to the flexibility of the model, either for workers or for customers. But it’s also not without limitations as a move: should it lose, Uber is likely to accelerate its already-rapid drive to bring in self-driving vehicles with no drivers at all. All other gig-economy driving jobs will likely follow.

You can argue that much of the work in the gig economy only exists because people are cheaper than machines right now. But we know the cost of technology only goes in one direction, and ultimately people can’t compete on price.

What about a more extreme intervention, enforcing employee representation at board level (common in places like Germany) and even perhaps some form of employee ownership? This may well improve practices over all and is likely to raise living standards for those in employment. Though it’s no guarantee of employment levels remaining high: mutually-owned John Lewis was forced to lay off many staff last year to focus on “better jobs, for better performing partners, on better pay.”

The result may be better lives for those with jobs, but it’s unlikely to do much for those without. If automation is rising, the ratio between those groups will continue to worsen.

Rebuilding the state

Another option is to accept that market-generated employment is unlikely to provide support for large portions of the population, at least at some point in their lives. If corporate profits are rising, and a smaller proportion of people are holding more wealth, then there’s an argument for increasing taxes and distributing the proceeds in a variety of forms.

Part of this might mean creating work by investing in infrastructure. Crumbling roads, a desperate shortage of affordable housing, an ageing power network and limited generating capacity: there are many places the UK could use a large state investment in improvements that would benefit citizens and businesses alike.

Part of it might mean tackling the holes in our health and social care systems: massively overworked doctors, underpaid nurses, and a critical shortage in trained care staff — only likely to worsen post-Brexit.

We might look at pushing the price of education back into general taxation and raising standards across the board, with increased investment.

And yes, we may ultimately have to look at some form of universal income, albeit one that accounts for complex differences in need. The savings from eliminating the administration of the current benefits system represent a fraction of the hundreds of billions that a universal income at any sensible level might cost.

Third ways

Of course, these routes are not mutually exclusive. We’re likely to see more of the former option in the short term, as governments stick with what they know and try to maintain the status quo. But personally, I’d like to see us put more time into considering the latter.

Taxing corporations is complex, particularly in our globalised economy. High corporation tax rates may look scary to businesses. But states generally deliver services like healthcare (and travel) much more cost-effectively than smaller-scale private options (just compare UK and US costs). Surely companies would want to invest in countries where they know their obligations are met with a single bill, rather than a complex legal framework of employee support?

Thought, if nothing else

I’m not a political scientist, or an economist. Posts like these are about exploring ideas rather than advocating them. But what I will advocate is consideration. I believe that automation threatens employment as the basis for an economy. I’m willing (and hopeful) to be proven wrong, but right now, the signs don’t look good. And beyond the surface conversation about universal basic income, I see very little consideration about what this means for the future.



Tom Cheesewright