When we have robots

The end of labour is to gain leisure.” Commonly attributed to Aristotle

Whilst my recent blogs have been about decision-making, politics and different ways of thinking about things, this topic has come up so often in conversations with three different sets of friends that I felt I had to explore it with you all.

Here is the scenario. It is 2030 and someone has finally produced a multi-functional household robot. It is more capable than a robotic vacuum cleaner and not restricted to performing a tightly defined task. It can move around the house, learn from you and become better at doing things over time.

Five years ago, that idea felt distant, but now feels oddly inevitable given the images from the World Humanoid Robot Games in Beijing and the advances in artificial intelligence. There will still be safety questions about having a humanoid robot at home, but I think people will get past it once autonomous cars take them home from a night out.

The four questions

After the initial excitement and a few jokes about our friends’ robotic personalities or dance moves, the conversations would centre around four recurrent questions:

1. What would you happily outsource?

2. What would you never outsource to it?

3. What would you pay for it?

But the most interesting question would come at the end:

4. What unintended consequences might there be?

This last question ended up veering towards lifestyle design and what might we miss if so much of our daily functions change. We also optimistically thought about what we might gain too.

What would you happily outsource?

The easy place to start is with activities most of us already try to avoid:

1. Make your bed

2. Hoover

3. Dust

4. Do the laundry

5. Iron clothes

6. Empty the dishwasher and put things away

7. Do all the garden chores

8. Bring in the shopping and put it away

9. Accept deliveries

To limit the scope we also assumed the robot would not do anything online. The precise list would be different for everyone, but the underlying rule is straightforward. We would outsource activities that take time, provide little enjoyment and hold little wider meaning. For example, if a robot cleaned my house perfectly while I was at work, I doubt I would return home mourning the loss of the experience.

The same principle already underpins many of the machines in our homes. Washing machines have removed much of the work involved in washing clothes. Dishwashers have done the same for dishes. Few people believe that using either has taken something important away from their lives.

A learning robot would extend the principle. Rather than automating one activity, it could learn how a particular household works and take responsibility for a wide range of repeated tasks. Imagine the robot knowing your preferences so well that it could recognise how you like your pillow and duvet set up, or put a lemon water by your bedside for wake-up.

What about food?

The conversation started to move from “what can I remove from my life?” to “what should we keep?”. Food preparation (e.g. cracking eggs and chopping vegetables) and cleaning up were easy tasks to give to the robot, but would we want to give up the act of self-indulgence or the service to a loved one that cooking brings? For example, would you want a robot to prepare Sunday lunch for friends or family? What about making a loved one’s favourite pancake recipe?

What would you never outsource to it?

So, when we have the perfect robot, what would we keep doing?

It is obvious that we cannot outsource exercise because the whole point of the activity is what it does to us. But what other positive effects do we risk losing if we outsource too much?

There are many other activities where getting it done is only part of the objective. What about reading to a child, where there is the practical act but also the opportunity to bond and talk through questions that might come up? Or entertaining a relative who is visiting town? What about filling in a form for a friend who does not understand what she is being asked?

The dreaded framework....

The discussion so far identifies at least five dimensions to an activity.

First, how much time does it take? Second, what is its cost? Third, do we enjoy doing it? Fourth, is the act of doing it good for us, whether through health benefits, relationship-building or meaning beyond the activity itself? Fifth, does doing it connect us to other people?

We can easily measure time and cost, but it is much harder to weigh up, or even identify, joy and meaning in a world where we feel obliged to optimise every activity. It is here, in what we enjoy and where we find meaning that our unique traits shine through.

The final aspect that feels pertinent right now, when loneliness is discussed more frequently, is whether outsourcing to robots makes it more or less likely that we meet people. Does it make room for more interactions with friends, family or a community built around a shared interest, or does it isolate us further?

When we make a change, we often consider the first-order effects, but rarely consider the higher-order ones. I see this lower-order thinking when people decide to outsource their thinking to AI or models, or, more practically, start swapping real food for processed supplements.

Here are some activities and the dimensions we discussed above:

Rank	Adult responsibility	Time	Cost / effort	Enjoyment	Good for us / meaning	Bonding	Verdict
1	Dusting	3	2	1	1	1	Outsource completely
2	Ironing	4	2	2	1	1	Outsource completely
3	Emptying dishwasher / putting things away	3	1	1	1	1	Outsource completely
4	Laundry	4	2	1	1	1	Mostly outsource
5	Routine garden maintenance	3	3	2	3	1	Outsource selectively
6	Food preparation and clearing up	4	2	2	2	2	Outsource the boring parts
7	Cooking a family meal	4	3	4	4	4	Keep, but use the robot as sous-chef
8	Helping a child with homework	3	2	3	5	5	Mostly keep human
9	Helping/caring for an elderly relative	4	4	3	5	5	Outsource tasks, never the relationship
10	Reading your child a book	2	1	5	5	5	Do not outsource

The AI question

The household robot may still be a few years away, but it is a thought experiment for the same decision we already face with AI.

I am happy for AI to search through information, summarise a long document, suggest a structure or check whether I have missed something. These activities can save time and make the final product better. The boundary becomes less obvious when AI moves from organising our thinking to doing the thinking for us.

There is an important difference between using AI to improve the expression of an opinion and asking it to supply the opinion. There is also a difference between asking it to identify weaknesses in an argument and accepting its conclusions without working through the issue ourselves.

Writing is a good example. The output is merely a jumble of words, but the activity is also a way of clarifying what we think and forcing us to explain our ideas using arguments. In my opinion, much communication, especially on social media platforms is to gain favour with the algorithm to find clients/followers. This has had the impact of more content and much less thinking. If we outsource writing, we may produce content but miss the more important part that develops us.

The same applies to many professional activities. Preparing for a meeting, analysing an investment or working through a difficult decision can feel inefficient. Yet that effort is often how we build knowledge and judgement. If we only consume the finished answer, we may save time while gradually weakening our ability to challenge it. Capabilities we stop exercising will eventually weaken.

The point is that technology should be a tool to improve our lives. The question is not only what can we outsource, but what should we outsource.

So how much would people pay for a robot?

Assume a family of four and a robot capable of making beds, cleaning, doing laundry, ironing, emptying the dishwasher, maintaining the garden, putting away shopping and accepting deliveries. For the moment, let us assume that, for safety reasons, it would not cook, look after children or complete online tasks, but that it could remove around fifteen hours of household work each week.

There are two ways it might be sold, direct purchase or the sharing economy.

Model 1 - buy or lease the robot and pay an annual maintenance charge.

A robot is more useful than a washing machine, should last longer than a mobile phone and, unlike a cleaner paid by the hour, could theoretically work every day.

Assume

• a freelance domestic helper charging £20 per hour.

• the robot cannot do a better job than a human

• you place the same value on the work whether it is done by a robot or a person.

£20 × 15 hours per week × 50 weeks gives us a gross annual value of £15,000.

This is clearly a big number and would immediately price most people out. However, it feels to me as though a robot could soon become available for less than this, making the decision real within a relatively short timeframe. This pricing also assumes that one household has exclusive use.

Model 2: Fractional ownership/rental

Given that this is a serious piece of kit and would sit idle for much of the time, I would imagine a fractional ownership or rental model would emerge. A household could use it on demand, bringing the individual cost down to an all-in price that suits different budgets and household arrangements.

I would imagine there is significant unmet demand for domestic help if the price is right.

What unintended consequences might there be?

Whilst having a domestic robot is an exciting proposition for me, it does raise some concerning questions:

• Is this creating unreplicable comfort? If I do not need to lift a finger at home, would this suddenly make the allure of a holiday or children going to university a much less enjoyable experience?

• How would I fill my new free time? Would it be an excuse to wade into more "productive" high intensity activity, or be a passive watcher on TV?

• Would I feel the same connection with my home? Would I miss that positive sense of satisfaction when I re-order my wardrobes for summer and winter?

Those questions seem to hang, until we thought more about the positives that would emerge:

• The marginal effort of hosting would decrease significantly. This extends to having people stay at your house, fostering an old-school set of experiences.

• By reducing the prep and clean up time of cooking, we might end up cooking more at home, and cook more ambitious meals.

• It would reduce arguments over household chores (we of course never argue over household chores)

• By having the domestic front taken care of, it could also save a lot of mental load.

• The benefits to the elderly of either remaining in their home or being looked after in yours have to be the biggest.

So what?

Exciting technologies are being developed, but it is worth really thinking through what are valuable skills you might gain or lose, and second order impacts around enjoyment and meaning.

It is an amazingly exciting time to be alive.

Next week's blog will be "Citizen's Assemblies". As always get the blog delivered directly to your inbox on Home | Deciders | for mental fitness | change your mind.