I’d like to begin with a story.
I was born in the UK — in Birmingham — although obviously I don’t have the accent! My parents came from Hong Kong, but we didn’t visit it until I was a few years old, since it’s quite the trip for any family.
The approach to the old Hong Kong airport in Kowloon Bay is hair-raising. You descend between skyscrapers, so close that you can practically see inside their windows. We were staying with relatives near the airport, which was fun, if noisy.
Me and my brother did the rounds of our aunts and uncles and grandparents, but eventually it was time for my parents to see their own friends. We were left with our cousins and the world’s greatest collection of pirated Famicom and Sega Megadrive videogames.
Now, these cousins. Their great aunt Agatha lived with them. As I was told it, she’d travelled the world, sailed the seas, fallen in love with all sorts of people, and made her fortune. Now in her eighties, she was still as sharp as a tack, with photographic memory and a wickedly funny tongue.
Agatha couldn’t easily walk any more, so more often than not, she’d sit in her armchair in the corner, situated just so she could see the whole living room and kitchen and hallway, and watch everyone coming and going. She wanted to know what was going on in the home, but more importantly, she wanted to be useful — and she was.
If you were on your way out but you’d forgotten to get pick up your keys, auntie Agatha would remind you (very loudly). If you were looking around for a letter or book you’d misplaced, she’d know precisely where you’d left it. She’d even watch you while you were doing your chores and tell you just which spots you’d forgotten to dust. Her job, as she saw it, was to help the household flourish, and keep them safe.
I’m sure some of you have figured out where I’m going with this. Almost forty years later, we all have auntie Agathas, watching over us in every room of our homes.
Today, in 2027
8 out of 10 households in the UK and US now have multiple home cameras. It’s one of the most astonishing success stories in the history of technology, with an adoption curve almost as impressive as smartphones in the previous decade. But unlike smartphones, we’ve bought many more than one per person.
What fuelled the rise of home cameras? Let’s start with the devices themselves.
Why did the home camera revolution only begin in 2018 and not earlier? Fast and cheap internet was an essential condition, allowing owners to monitor their homes on the move and abroad. Another boost came from the ‘smartphone dividend’, which reduced the price of camera components.
But beyond 2018, two technological revolutions fuelled the rise of home cameras: charging and sensors.
Nowadays, it’s hard to believe that almost all home cameras in the mid-teens were wired. These cameras had no batteries and had to be tethered to a power outlet at all times, constraining their placement within homes and generally causing an unsightly mess.
From 2018 to 2023, home cameras adopted batteries lasting one week to one month — a massive improvement over tethering, as they could be mounted anywhere, including outdoors and in bathrooms — but arguably more irritating than wires, as their “low-power” chirping became a frequent sound in many homes.
It wasn’t until the full rollout of resonance charging, or more broadly speaking, ‘charging at a distance’, that cameras truly permeated every room and corner of our homes. Freed from the need to be wired or retrieved every month, and completely weatherproofed, they were stuck in the corners of ceilings, thrown onto roofs, hung on walls, mounted on gates, and balanced precariously on shelves. Providing they remained within range of a resonance station, they could be placed and forgotten for years.
The improvement in the sensor capabilities of home cameras has been even more extraordinary. In 2018, most cameras had a laughably-named ‘high-definition’ resolution of 1920 x 1080 — barely enough to distinguish small objects across a room. Matters were soon improved with the introduction of ‘High Speed 4K’ sensors that could examine minute changes in skin bloodflow to monitor people’s heartrate and emotional state. Soon after, cameras reached beyond the visible spectrum to infrared and ultraviolet, essential for home security and health applications.
It wasn’t until the introduction of multipath LIDAR in 2024 that the supremacy of cameras in our hearts and homes was assured. Various primitive forms of LIDAR had been present in earlier cameras, as an aid to home VR and augmented reality through precision depth mapping and 3D positioning. Multipath LIDAR, however, multiplied the reach of our cameras by using reflections to see around corners into other rooms; to interpolate new camera angles; and to even see inside objects. It finally provided total awareness of all objects within a home, without the need for excessive numbers of cameras.
In fact, the most advanced multipath systems now pose a threat to the business model of the camera manufacturers who’ve emphasised quantity over quality. Now that a single camera can take the place of many, it’s likely that overall camera shipments could begin falling.
Enough about technology — why did people invite cameras into their homes, and what did they use them for? I’ve identified five broad applications, in rough chronological order:
1. Baseline Sensor Platform
By the late teens and early 20s, desktop and laptop computers were gradually departing our homes, and with them, all the keyboards, mice, and trackpads used as input devices. Their replacements? Touchscreens, wearable tech, and heads-up displays; devices that were useful and valuable precisely because they were so personal. But not everyone had those devices, and those who didn’t — infants, children, and the elderly — still needed to interact with home technology.
Enter home cameras.
Cameras also acted as a baseline platform for other rising technologies such as VR and AR. Yes, our headsets and glasses have their own positioning systems, but they work far better when coupled with high-power in-home LIDAR. It’s a venerable use of cameras, dating all the way back to 2010 with the Microsoft Kinect, and later, the HTC Vive.
The ancestors of home cameras are, of course, CCTV and security cameras. Likewise, our own cameras give us valuable peace of mind when we’re at home and away, by monitoring our kids, pets, valuables, and keeping an eye out for accidents.
And as we added more and more ‘smart’ (sic) locks and doors to our homes, cameras became essential lubricants in the so-called ‘Socialisation of Everything’ by making deliveries and sharing and cleaning safer and more reliable. It took a while to reach a détente, but most hosts and guests settled on ‘angel snapshots’ as an acceptable mode of surveillance (i.e. only recording rooms while guests were not present in them — enough to spare guests’ blushes while still sufficient evidence to prove whether they damaged anything).
Like millions of people, my last home workout was guided by cameras that tracked my body with millimetre-accuracy while also quantifying my muscle exertion and oxygen consumption. To be honest, I’m not entirely clear how all of that data is used, but providing that I follow the real-time workout instructions, I recover faster and I don’t get cramp any more.
Home cameras have been monitoring our health since the mid-twenties — not just during workouts, but all the time. Obviously there’s only so much they can infer from visuals and LIDAR, but they can still detect a surprising range of symptoms long before humans can: everything from an incipient cold to the tell-tale tremors that precede Parkinson’s, and plenty more.
Our cameras can tell whether you’ve left out food for too long; nag you about your poor posture; note the signs of depression; and even remind you to clean up that festering mug — particularly useful for student houses.
As such, it’s hardly surprising that parents quickly fell in love with home cameras, even before they introduced dedicated child-monitoring features. They began with safety features — whether a child was venturing too close to something dangerous — but when they began to monitor and improve children’s learning and development… well, it was hard to keep the cameras in stock.
Cameras now count the number of words spoken by children every day; they scrutinise their walking patterns and physical dexterity; and they perform gaze-tracking to assess social development. It only took a few well-publicised success stories about detecting language problems and developmental issues for home cameras to become an essential purchase for all new parents.
4. Environment & Economy
An unexpected second-order effect of home cameras has been on our environment and economy. It’s hard to believe today, but in the mid-to-late-teens, every home robot and drone contained its own sensor array, which might include cameras, ultrasound, or even LIDAR. This was an astonishingly wasteful practice which only ended when robot manufacturers decided they could save money by relying on the more accurate and ever-present cameras in every home.
In a similar vein, cameras reversed the absurd trend for ‘smart toothbrushes’ and ‘smart water bottles’ that would record the most basic of human activities. Why use a fancy expensive toothbrush when cameras can check whether you (or your kids) have brushed properly tonight, at no extra cost?
The computer vision that powered home cameras also negated the need for cumbersome forms of product tracking such as RFID tags, QR codes, and barcodes. Cameras could identify objects from sight, and even distinguish two apparently identical objects; when deployed across an entire distribution chain from factory to warehouse to home (physical retailers having steadily receded over the past decade), a product could be digitally tagged and tracked without any physical markings whatsoever.
Back when home cameras were new and viewed with unalloyed suspicion, I’d always sell them to my friends and family by saying how I never lost track of anything any more. I’m not a forgetful person, but even I would occasionally forget my keys or my wallet when leaving the house. Not any more! And no more misplacing your nail clippers or your passport or your glasses when every single object at home is tracked in real time.
These days, cameras are also capable of tracking object usage. This apparently minor feature has spawned entire new industries, since we can now predict when any given product might need to be repaired or replaced. The same data can be used to calculate an object’s real-time monetary value — and consequently the value of the entire contents of your home. Quite helpful for insurers (or thieves, as demonstrated by the events last December).
Last week, my camera pointed out that I’ve only used my slow cooker twice in two years, and that I could sell it for £50. I gave it the nod, and the next day it let someone in to collect it while I was out at work. And just today, it encouraged me to trade in my well-used exercise bike to a newer model, chosen exactly for my preferences.
Believe it or not, one very early application of home cameras was for fashion advice. As far as I can tell, it still doesn’t well all that well even today, but cameras can take very accurate body measurements and identify which clothes would fit you best, or order personalised clothes and accessories. This reduced a truly massive amount of waste caused by millions of people overordering and returning clothes that didn’t fit.
5. Behaviour Modification
Undoubtedly the most advanced and most controversial application of home cameras. To continue with clothes as an example, not only do cameras measure your body and see how well clothes might fit, but they can remember often you wear your clothes and how they affect your mood. It was just a hop, skip, and a jump for cameras to recommend the clothes you should wear to improve your mood, based on stress level predictions derived from your schedule and personal communications. You don’t need to listen to those recommendations, of course — but most people do.
Clothes may seem like a frivolous form of behaviour modification to you (although not to me!) — but what about the conversation tracking, gaze tracking, and emotion tracking that many households now employ in the service of their whole families? These days, cameras recommend all sorts of things to improve your home’s overall mood and happiness, from suggesting the perfect playlist, to reminding you of that book you meant to read, to turning off a distracting TV show. In some households, they’ve skipped over recommending straight into doing. It’s just faster that way.
All these applications work better the more data our cameras have and the more permissions you enable. It’s no good if you turn your cameras off half the time or you keep them out of your bathroom and bedroom.
And then you also have the discovery problem — how are people supposed to know that all these applications even exist? No-one’s going to watch a thirty minute video or read a manual before starting. So all camera features and applications are turned on by default, because in surveys, most people seem to want them.
We’ve known for a long time that we act differently when we are being watched. It’s true. These days, we don’t lose things any more. We waste less, buy less, lend more and exercise more. We don’t buy things twice. We don’t buy things we won’t need. Our clothes fit better, our food is fresher, our homes are cleaner, and our bodies and minds are healthier. And we are happier — we think.
Everything in our homes is watched, including us. Everything in our homes is known, including us. And everything in our homes is optimised.
Back to 2017
Now, I know what you’re thinking. No way will people sign up to this! But this is not hypothetical. People have shown time and time again that they value money and convenience more than anything else; we already have millions of Amazon Echos in the wild, and all the Echo spinoffs announced in mid-2017. Here in the UK, we have something like sixmillion CCTV cameras watching us out in public. And of course, worldwide we have billions of smartphones that track us all the time.
That is not to say people are stupid for wanting home cameras. They will simply be too useful and too cheap to avoid, just as search engines and smartphones have become too useful and too cheap to avoid, despite their drawbacks.
In any case, believe it or not, I think it’s possible to get too hung up on the current arguments surrounding privacy and advertising. To be very clear — I think that personal data privacy is incredibly important, and I think advertising has had a negative effect on our personal desires. But we shouldn’t assume that present-day practices for privacy and advertising will hold true in the future.
Right now, we know that aren’t prepared to pay for true cost of device and operating costs, so we defray them with advertising, which obviously needs to be targeted — not just to make money, but to make better and relevant. But if you show the cost of removing ads (e.g. Amazon fire tablets) people may choose differently. This should be mandated.
Likewise, right now our personal devices need to be cloud-enabled so we can crunch the numbers better, so the AIs can learn faster; so we’re all giving up our data freely, if often unknowingly. But there are glimmers of a better way. Apple announced differential privacy; they can learn from user data without compromising individual user identity. Google’s distributed machine learning, maybe shows that it doesn’t need to be this way.
I’m not a pollyanna, I don’t think things will just work out fine — present events have shown that quite clearly. My point is that things can change, even ‘cloud enabled devices’ and pervasive targeted advertising. In fact I think it’s possible that the presence of cameras in people’s homes will really help people understand the details of data ownership and retention policies.
We need to argue for these controls, and we need to demonstrate why they’re important, by building superior products and services that have privacy and data ownership controls built-in. We have to do this because this future is coming, and it’s too useful and too cheap for us to avoid.
This talk was first given at the Mozilla Speaker Series on 10th May 2017: