Optometrists, Octopii, Rubber Ducks & Centaurs: my talk at Design for AI, TU Delft, October 2022

I was fortunate to be invited to the wonderful (huge) campus of TU Delft earlier this year to give a talk on “Designing for AI.”

I felt a little bit more of an imposter than usual – as I’d left my role in the field nearly a year ago – but it felt like a nice opportunity to wrap up what I thought I’d learned in the last 6 years at Google Research.

Below is the recording of the talk – and my slides with speaker notes.

I’m very grateful to Phil Van Allen and Wing Man for the invitation and support. Thank you Elisa Giaccardi, Alessandro Bozzon, Dave Murray-Rust and everyone the faculty of industrial design engineering at TU Delft for organising a wonderful event.

The excellent talks of my estimable fellow speakers – Elizabeth Churchill, Caroline Sinders and John can be found on the event site here.

Video of Matt Jones “Designing for AI” talk at TU Delft, October 2022

Slide 1


Slide 2

This talk is mainly a bunch of work from my recent past – the last 5/6 years at Google Research. There may be some themes connecting the dots I hope! I’ve tried to frame them in relation to a series of metaphors that have helped me engage with the engineering and computer science at play.

Slide 3

I won’t labour the definition of metaphor or why it’s so important in opening up the space of designing AI, especially as there is a great, whole paper about that by Dave Murray-Rust and colleagues! But I thought I would race through some of the metaphors I’ve encountered and used in my work in the past.

The term AI itself is best seen as a metaphor to be translated. John Giannandrea was my “grand boss” at Google and headed up Google Research when I joined. JG’s advice to me years ago still stands me in good stead for most projects in the space…

But the first metaphor I really want to address is that of the Optometrist.

This image of my friend Phil Gyford (thanks Phil!) shows him experiencing something many of us have done – taking an eye test in one of those wonderful steampunk contraptions where the optometrist asks you to stare through different lenses at a chart, while asking “Is it better like this? Or like this?”

This comes from the ‘optometrist’ algorithm work by colleagues in Google Research working with nuclear fusion researchers. The AI system optimising the fusion experiments presents experimental parameter options to a human scientist, in the mode of a eye testing optometrist ‘better like this, or like this?’

For me to calls to mind this famous scene of human-computer interaction: the photo enhancer in Blade Runner.

It makes the human the ineffable intuitive hero, but perhaps masking some of the uncanny superhuman properties of what the machine is doing.

The AIs are magic black boxes, but so are the humans!

Which has lead me in the past to consider such AI-systems as ‘magic boxes’ in larger service design patterns.

How does the human operator ‘call in’ or address the magic box?

How do teams agree it’s ‘magic box’ time?

I think this work is as important as de-mystifying the boxes!

Lais de Almeida – a past colleague at Google Health and before that Deepmind – has looked at just this in terms of the complex interactions in clinical healthcare settings through the lens of service design.

How does an AI system that can outperform human diagnosis (Ie the retinopathy AI from deep mind shown here) work within the expert human dynamics of the team?

My next metaphor might already be familiar to you – the centaur.

[Certainly I’ve talked about it before…!]

If you haven’t come across it:

Gary Kasparov famously took on chess-AI Deep Blue and was defeated (narrowly)

He came away from that encounter with an idea for a new form of chess where teams of humans and AIs played against other teams of humans and AIs… dubbed ‘centaur chess’ or ‘advanced chess’

I first started investigating this metaphorical interaction about 2016 – and around those times it manifested in things like Google’s autocomplete in gmail etc – but of course the LLM revolution has taken centaurs into new territory.

This very recent paper for instance looks at the use of LLMs not only in generating text but then coupling that to other models that can “operate other machines” – ie act based on what is generated in the world, and on the world (on your behalf, hopefully)

And notion of a Human/AI agent team is something I looked into with colleagues in Google Research’s AIUX team for a while – in numerous projects we did under the banner of “Project Lyra”.

Rather than AI systems that a human interacts with e.g. a cloud based assistant as a service – this would be pairing truly-personal AI agents with human owners to work in tandem with tools/surfaces that they both use/interact with.

And I think there is something here to engage with in terms of ‘designing the AI we need’ – being conscious of when we make things that feel like ‘pedal-assist’ bikes, amplifying our abilities and reach vs when we give power over to what political scientist David Runciman has described as the real worry. Rather than AI, “AA” – Artificial Agency.

[nb this is interesting on that idea, also]

We worked with london-based design studio Special Projects on how we might ‘unbox’ and train a personal AI, allowing safe, playful practice space for the human and agent where it could learn preferences and boundaries in ‘co-piloting’ experiences.

For this we looked to techniques of teaching and developing ‘mastery’ to adapt into training kits that would come with your personal AI .

On the ‘pedal-assist’ side of the metaphor, the space of ‘amplification’ I think there is also a question of embodiment in the interaction design and a tool’s “ready-to-hand”-ness. Related to ‘where the action is’ is “where the intelligence is”

In 2016 I was at Google Research, working with a group that was pioneering techniques for on-device AI.

Moving the machine learning models and operations to a device gives great advantages in privacy and performance – but perhaps most notably in energy use.

If you process things ‘where the action is’ rather than firing up a radio to send information back and forth from the cloud, then you save a bunch of battery power…

Clips was a little autonomous camera that has no viewfinder but is trained out of the box to recognise what humans generally like to take pictures of so you can be in the action. The ‘shutter’ button is just that – but also a ‘voting’ button – training the device on what YOU want pictures of.

There is a neural network onboard the Clips initially trained to look for what we think of as ‘great moments’ and capture them.

It had about 3 hours battery life, 120º field of view and can be held, put down on picnic tables, clipped onto backpacks or clothing and is designed so you don’t have to decide to be in the moment or capture it. Crucially – all the photography and processing stays on the device until you decide what to do with it.

This sort of edge AI is important for performance and privacy – but also energy efficiency.

A mesh of situated “Small models loosely joined” is also a very interesting counter narrative to the current massive-model-in-the-cloud orthodoxy.

This from Pete Warden’s blog highlights the ‘difference that makes a difference’ in the physics of this approach!

And I hope you agree addressing the energy usage/GHG-production performance of our work should be part of the design approach.

Another example from around 2016-2017 – the on-device “now playing” functionality that was built into Pixel phones to quickly identify music using recognisers running purely on the phone. Subsequent pixel releases have since leaned on these approaches with dedicated TPUs for on-device AI becoming selling points (as they have for iOS devices too!)

And as we know ourselves we are not just brains – we are bodies… we have cognition all over our body.

Our first shipping AI on-device felt almost akin to these outposts of ‘thinking’ – small, simple, useful reflexes that we can distribute around our cyborg self.

And I think this approach again is a useful counter narrative that can reveal new opportunities – rather than the centralised cloud AI model, we look to intelligence distributed about ourselves and our environment.

A related technique pioneered by the group I worked in at Google is Federated Learning – allowing distributed devices to train privately to their context, but then aggregating that learning to share and improve the models for all while preserving privacy.

This once-semiheretical approach has become widespread practice in the industry since, not just at Google.

My next metaphor builds further on this thought of distributed intelligence – the wonderful octopus!

I have always found this quote from ETH’s Bertrand Meyer inspiring… what if it’s all just knees! No ‘brains’ as such!!!

In Peter Godfrey-Smith’s recent book he explores different models of cognition and consciousness through the lens of the octopus.

What I find fascinating is the distributed, embodied (rather than centralized) model of cognition they appear to have – with most of their ‘brains’ being in their tentacles…

And moving to fiction, specifically SF – this wonderful book by Adrian Tchaikovsky depicts an advanced-race of spacefaring octopi that have three minds that work in concert in each individual. “Three semi-autonomous but interdependent components, an “arm-driven undermind (their Reach, as opposed to the Crown of their central brain or the Guise of their skin)”

I want to focus on the that idea of ‘guise’ from Tchaikovsky’s book – how we might show what a learned system is ‘thinking’ on the surface of interaction.

We worked with Been Kim and Emily Reif in Google research who were investigating interpretability in modest using a technique called Tensor concept activation vectors or TCAVs – allowing subjectivities like ‘adventurousness’ to be trained into a personalised model and then drawn onto a dynamic control surface for search – a constantly reacting ‘guise’ skin that allows a kind of ‘2-player’ game between the human and their agent searching a space together.

We built this prototype in 2018 with Nord Projects.

This is CavCam and CavStudio – more work using TCAVS by Nord Projects again, with Alison Lentz, Alice Moloney and others in Google Research examining how these personalised trained models could become reactive ‘lenses’ for creative photography.

There are some lovely UI touches in this from Nord Projects also: for instance the outline of the shutter button glowing with differing intensity based on the AI confidence.

Finally – the Rubber Duck metaphor!

You may have heard the term ‘rubber duck debugging’? Whereby your solve your problems or escape creative blocks by explaining out-loud to a rubber duck – or in our case in this work from 2020 and my then team in Google Research (AIUX) an AI agent.

We did this through the early stages of covid where we felt keenly the lack of informal dialog in the studio leading to breakthroughs. Could we have LLM-powered agents on hand to help make up for that?

And I think that ‘social’ context for agents in assisting creative work is what’s being highlighted here by the founder of MidJourney, David Holz. They deliberated placed their generative system in the social context of discord to avoid the ‘blank canvas’ problem (as well as supercharge their adoption) [reads quote]

But this latest much-discussed revolution in LLMs and generative AI is still very text based.

What happens if we take the interactions from magic words to magic canvases?

Or better yet multiplayer magic canvases?

There’s lots of exciting work here – and I’d point you (with some bias) towards an old intern colleague of ours – Gerard Serra – working at a startup in Barcelona called “Fermat

So finally – as I said I don’t work at this as my day job any more!

I work for a company called Lunar Energy that has a mission of electrifying homes, and moving us from dependency on fossil fuels to renewable energy.

We make solar battery systems but also AI software that controls and connects battery systems – to optimise them based on what is happening in context.

For example this recent (September 2022) typhoon warning in Japan where we have a large fleet of batteries controlled by our Gridshare platform.

You can perhaps see in the time-series plot the battery sites ‘anticipating’ the approach of the typhoon and making sure they are charged to provide effective backup to the grid.

And I’m biased of course – but think most of all this is the AI we need to be designing, that helps us at planetary scale – which is why I’m very interested by the recent announcement of the https://antikythera.xyz/ program and where that might also lead institutions like TU Delft for this next crucial decade toward the goals of 2030.

Pharaoh Lovelock

Kamasi Washington’s obit of Pharoah Sanders

“Here was a man who played spiritual, cosmic music, from whom I wanted to know the secrets to the universe. But he was more interested in being in the moment and recognising the power of being in the moment. He showed me that connecting with the great beyond is sometimes about the simplest things.”

John Gray remembering James Lovelock

“Jim attributed his great old age to long daily walks – he lived to 103 and right up to the end his mind was very vivid. I joined him sometimes wandering through his grounds, where he’d let Gaia have its will. He had a cat and once the cat sat on my shoulder through the entire walk.”

Driving an EV from London to Porto


It’s fine. You could probably do it in three days, we took it easy with four. Stopping along the way to charge and sleep is wonderful. Driving is insanely boring, no matter what the advertising industry tries to tell you, even in a shiny new EV. The eurotunnel is an engineering miracle. France’s charging infrastructure is great, Spain a little less so, and Portugal less so again – but improving. Charge point mapping/route apps are your friend.

I’ve been meaning to write this up for a while – well, since this summer of 2022.

Told the tale to a number of folks who are thinking of moving from away from cars that use fossils fuels to make tiny explosions inside them, but are feeling uneasy about long journeys (that they probably undertake once a year if that).

I was lucky to get our family EV (an ID4) late last year before supplies got very constrained – and even then it was a a six month wait from having made the down payment in the summer of 2021.

This summer we decided to drive instead of fly to Arouca, Portugal – a lovely small town in the mountains / river valleys about an hour to the east of Porto.

It’s a beautiful place – with loads of great nature / hikes to be had including the epic Arouca suspension bridge and ‘Passadiços do Paiva’ riverside walk along wooden platforms. If you’re near Porto, go visit!

Our Route

We started out from London early for the Eurotunnel.

It was my first time so I was a little trepidatious.

I needn’t of worried – it was all very smooth and simple. 30 or so minutes after driving our car onto the train we were rolling off in France.

From there to our first charge, which was one of the Ionity network.

Quick and painless for us – but the person in the next bay (with a very shiny Porsche Taycan EV) was struggling.

The flummoxed face of a stranger looking at a cellphone or frantically trying to tap a multitude of RFID cards against a charging station would be a familiar sight this trip.

We gave them our VW WeCharge card which allowed us to use Ionity, and they were able to make their return ride on the eurotunnel back to the UK.

A brief diversion stating the obvious on RFID cards, charging networks and apps etc.

There are too many.

If I have a combustion engine, I can roll up anywhere and pay cash or use a debit/credit card to get refuelled. Not so with electrons. You have to give companies data, and membership as well as money. A few EV charging networks now will accept money alone, but that’s the minority it seems.

A good alternative is to join something that allows you to tap into different networks. It seemed though that not one of these would work across everything on our route. ChargeMap is a great app – and certainly was our mainstay – as well as the “ChargeMap Pass” RFID card you can get from them that allowed us to use a number of different networks on the way. Our VW “WeCharge” card also helped – but in Portugal – Miio came into its own.

Mostly I found navigating between Ionity stations. Ionity is expensive for sure – but fast – and fairly ubiquitous. Also – there tend to be 6-8 of them at a location instead of 1 or 2. Driving to a way point only to find that the one working EV charger is occupied is very dispiriting.

Night One: Saumur

After traveling through Normandy and North-east of Paris we got to our first stop Saumur in late-afternoon. Settled into the cheapish-but-nice hotel we’d booked (which had an EV charger… most of the big hotel booking sites now let you filter for locations that have EV charge points) and walked the short distance into the town centre (not a hardship after driving all day).

Quickly the pattern of the journey was established – this was all about driving through the morning and early afternoon, arriving somewhere to eat, sleep and recharge ourselves as well as the car. It also reinforced that the journey was going to be part of the experience rather than the thing merely facilitating it. I’d had that mindset going into this, and planning it – but it had been theoretical until the cold beer and charcuterie arrived…

We were also travelling through a Europe that was experiencing a heatwave – and tangible reminders of the climate emergency were everywhere. It was sobering to see the levels of the Loire for instance

Night Two: San Sebastián

The next morning we departed pretty early as we had a long leg ahead of us – through south western France, past Bordeaux – ending up in San Sebastián in the Basque region.

Despite traffic jams all around Bordeaux scuppering our plans to have lunch there we made ok time and did two full charges on the way. French service stations continued to impress – with one offering a model for the future: more or less an outdoor festival site with beanbags and hammocks in the shade while you let your EV recharge!

Again – thinking about charging as being something you enjoy rather than an annoying hindrance is made a lot easier to embrace when you have a motorway stop like this – reading your kindle for an hour in a beanbag with a cold diet coke, while the electrons flow…

We passed over the border around 5pm and got to San Sebastian not much after that.

We had a pretty posh hotel booked in the centre of the city which had been listed as having an EV charge point, but the logistics of plugging in our car defeated the staff on duty… So, it was into town for a sunset pintxos crawl…

Night Three: Salamanca

From the Basque region through Spain, the charging situation started to get a little more vexing. Chargers were on many different networks, some were faster than others – many only being 50kW, and maintenance became a little bit of an issue. Again Ionity was the saviour – but we managed to scrape through to our next stop – a vineyard/hotel outside Salamanca.

This, being more of a ‘destination’ had a Tesla ‘destination charger’ – which in Europe are all CCS standard rather than the proprietary Tesla connector. They’re low powered – 7kW – but again leaving it overnight to charge while we slept was no hassle.

In the morning we drove the remainder of the way to Arouca in Portugal.

Arouca is a pretty small town but has a 50kw charger in it’s centre which we used every couple of days we were staying there. We found the Miio app invaluable for using the chargers in Portugal – which was a bit sparser in its coverage again from Spain. We’d also bought a EU-outlet compatible trickle charger with us to plug into socket where we were staying in the countryside outside the town, and allowed us to leave a few weeks later on 100% for the route back… which was tackled with considerably less trepidation after the success of our journey there.


Long road trips in EVs are fine, entirely doable with a bit of casual planning – and maybe more pleasant than the equivalent in a fossil-fuel car, once we adjust your mindset a little… Stopping and recharging your car and yourself for a while on the journey can be a very positive thing – especially in France where they already seem to have it down.

There’s an interesting new typology of rest stop yet to emerge around EV charging and longer road trips, but you can kind of see the beginnings of it – if you squint.

EV charging networks though really need to improve. Ionity is the best in class for sure – but hopefully others will follow their lead. But – spare us the memberships, apps and terms of service… just let us pay like the people buying the dinosaur juice!!!

I hope – if you have an EV or have been considering changing over to one – this post is useful in someway.

Until the next one…

The right to longevity

After the EU “right to repair” should there perhaps be a “right to longevity” for connected objects, to enable them to be reanimated by open-source code and platforms once they have had their motivating spirit of software removed by the shuttering of whatever service they were originally the avatar of?

I reanimated my Little Printer last year with the help of the good folks at Nord, who in turn were able to do that because Matt Webb had open-sourced the code when BERG’s Little Printer / Bergcloud shut down.

Google just announced that they are to shutter Stadia. I was an early adopter – and have one of the quite handsome Stadia controllers. It’s a really nice games controller! Look – it even got a mini documentary about it’s design engineering presented by Baratunde Thurston!

So – as Stuart Horton says:


And let’s not stop there – let’s mandate that anything that is manufactured with atoms, but animated by software and services *must* have it’s firmware at minimum open-sourced if the service is shutting down.

Feels like a good bit of ‘endineering

Allow old phones, aibos, connected juicers, jibos, Nabaztags to rise like polycarbonate phoenixes from their slumbers.

An EU right-to-longevity might keep more atoms out of landfill by letting the bits be free.

Update 2023-01-17: Stadia team are working on making the controller work with other devices / Bluetooth https://9to5google.com/2023/01/15/stadia-controller-bluetooth-certification/

Partner / Tool / Canvas: UI for AI Image Generators

“Howl’s Moving Castle, with Solar Panels” – using Stable Diffusion / DreamStudio LIte

Like a lot of folks, I’ve been messing about with the various AI image generators as they open up.

While at Google I got to play with language model work quite a bit, and we worked on a series of projects looking at AI tools as ‘thought partners’ – but mainly in the space of language with some multimodal components.

As a result perhaps – the things I find myself curious about are not so much the models or the outputs – but the interfaces to these generator systems and the way they might inspire different creative processes.

For instance – Midjourney operates through a discord chat interface – reinforcing perhaps the notion that there is a personage at the other end crafting these things and sending them back to you in a chat. I found a turn-taking dynamic underlines play and iteration – creating an initially addictive experience despite the clunkyness of the UI. It feels like an infinite game. You’re also exposed (whether you like it or not…) to what others are producing – and the prompts they are using to do so.

Dall-e and Stable Diffusion via Dreamstudio have more of a ‘traditional’ tool UI, with a canvas where the prompt is rendered, that the user can tweak with various settings and sliders. It feels (to me) less open-ended – but more tunable, more open to ‘mastery’ as a useful tool.

All three to varying extents resurface prompts and output from fellow users – creating a ‘view-source’ loop for newbies and dilettantes like me.

Gerard Serra – who we were lucky to host as an intern while I was at Google AIUX – has been working on perhaps another possibility for ‘co-working with AI’.

While this is back in the realm of LLMs and language rather than image generation, I am a fan of the approach: creating a shared canvas that humans and AI co-work on. How might this extend to image generator UI?

Open Lecture at CIID: “Keeping up with the Kardashevians”

Back in April 2022, I was invited to speak as part of CIID’s Open Lecture series on my career so far (!) and what I’m working on now at Moixa.com.

Naturally, It ends up talking about trying to reframe the energy transition / climate emergency from a discourse of ‘sustainability’ to one of ‘abundance’ – referencing Russian physicists and Chobani yoghurt.

Thank you so much to Simona, Alie and the rest of the crew for hosting – was a great audience with a lot of old friends showing up which was lovely (not that they spared the hard questions…).

I was on vacation at the time with minimal internet, so I ended up pre-recording the talk – allowing me the novelty of being able to heckle myself in the zoom chat…

CIID Open Lecture, Matt Jones, Apr 5 2022

CIID Open Lecture, April 2022

Hello, it’s very nice to be “here”!

Slide 1

Thank you CIID for inviting me. 

I’ll explain this silly title later, but for now let me introduce myself…

Slide 2

Simona and Alie asked me to give a little talk about my career, which sent me into a spiral of mortality and desperation of course. I’ve been doing whatever I do for a long time now. And the thing is I’m not at all sure that matters much.

Slide 3: My life vs Moore’s Law

I’m 50 this year, you see. I’m guessing everyone has some understanding of Moore’s Law by now – things get more powerful, cheaper and smaller every 18mo-2yrs or so. I thought about what that means for what I’ve done for the last 50 years. Basically everything I have worked on has changed a million-fold since I started working on it (not getting paid for it!)

Slide 4: Design vs Moore’s Law

If you’re a designer that works in, say, furniture or fashion – these effects are felt peripherally: maybe in terms of tools or techniques but not at the core of what you do. I don’t mean to pick on Barber Ogersby here, but they kind of started the same time as me. You get to do deep, good work in a framework of appreciation that doesn’t change that much. I get the sense that even this has changed radically in the last few years as well – for many good reasons.

Slide 5: A book about BERG would make no sense.

Anyway – when I was asked to look back on work from BERG days over a decade ago, it’s hard to pretend it matters in the same way as when you did it. But perhaps it matters in different ways? I’m trying here to look for those threads and ideas which might still be useful.

Slide 6: Pace Layers & Short-Circuits

In design for tech we are building on shifting (and short-circuiting) pace layers. I’ve always found it most useful to think them as connected, rather than separate. Slowly permeable cell membranes/semiconducting layers. New technology is often the wormhole or short-circuit across them.

Slide 7: BERG

So with that said, to BERG.

Slide 8: About BERG

BERG was a studio formed out of a partnership between Matt Webb and Jack Schulze. Tom Armitage joined – and myself shortly after. From there we grew to a small product invention and design consultancy of about 15 folks at our largest, but always around 8-9. It was a great size – I’m still proud of the stuff we took on, and the way we did it.

Slide 9: All you can see of systems are surfaces.

One of the central tenets of BERG: All you can see of systems are surfaces. The complexity and interdependency of the modern world is not evident. And you can make choices as designers about how to handle that. Most design orthodoxy (at the time) and now is to drive towards ‘seamlessness’ – but we preferred Mark Weizer’s exhortation for ‘beautiful seams’ that would increase the legibility of systems while still making attractive and easy to engage surfaces…

Slide 10: Moore’s Law meets Main St.

Another central tenet of the studio: “What got just got cheap and boring?” We looked to mass produced toys and electronics, rather than solely to the cutting edge for inspiration. Understanding what had passed through the ‘hype curve’ of the tech scene into what Gartner called ‘the trough of disillusionment’. This felt like the primordial soup of invention.

Slide 11: A tale of two Arthurs

We got called sci-fi. Design fiction. I don’t think we were Sci-fi. We were more like 18thC naturalists, trying to explain something we were in, to ourselves. think we were more Brian Arthur than Arthur C. Clarke. We didn’t want to see tech as magic.

Slide 12: The nature of technology

Brian Arthur’s book “The Nature of Technology” was a huge influence on me at the time (and continues to be.) An economist and scholar of network effects, he tries to establish how and why technology evolves and builds in value. In the book he explains how diverse ‘assemblages’ of scientific and engineering phenomenon combine into new inventions. The give and take between human/cultural needs and emergent technical phenomenon felt far more compelling and inspiring than the human centred design orthodoxy at the time.

Slide 13: Thinking through making / chatting is cheating

This emphasis on exploring phenomena and tech as a path to invention we referred to as ‘material exploration’ as a phase of every studio project. We led with it, privileged it in the way contemporaries emphasised user research – sometimes to our detriment! But the studio was a vehicle for this kind of curiosity – and it’s what powered it.

Sllide 14: Lamps for Google

“Lamps” was a very material-exploration heavy project for Google Creative Lab in 2010. It was early days for the commoditisation of computer vision, and also about the time that Google Glass was announced. We pushed this work to see what it would be like to have computer vision in the world *with you* as an actor rather than privatised behind glasses.

Slide 15: Smart Light

The premise was instead of computer vision reporting back to a personal UI, it would act in the world as a kind of projected ‘smart light’, that would have agency based on your commands.

Slide 16: Build your own ARKit, too early

To make these experiments we had to build a rig. An example here of how the pace-layers of past design work get short-circuited. Here’s our painstaking 2010, 10x the size, cost, pain DIY version of ARKit… which would come along only a few years later.

Slide 17: watch the “Dumb things, smart light” video here

This final piece takes the smart light idea to a ‘speculative design’ conclusion. What if we made very dumb interactive blocks that were animated with ‘smart light’ computation from the cloud… There’s a little bit of a critical thought in here, but mainly we loved how strange and beautiful this could look!

Slide 18: Schooloscope

We also treated data as a material for exploration. One of the projects I’m always proudest of was Schooloscope in 2009 (one of the first BERG studio projects for Channel4 in the UK) – led by Matt Webb, Matt Brown and Tom Armitage which did a fantastic job of reframing contentious school performance data from a cold emphasis on academic performance to a much more accessible and holistic presentation of a school for parents (and kids) to access. Each school’s performance data creates an avatar based on our predisposition to interpret facial expression (after Chernoff)

Slide 19: Suwappu

Another example of play – Suwappu was an exploration for a toy/media franchise for Dentsu. Each toy has an AR environment coded to it, and weekly updates to the the storylines and interactions between them was envisaged. Again a metaverse in… reverse?

Slide 20: Cars for Intel

A lot of the studio work we couldn’t talk about publicly or release. I don’t think I’ve ever shown this before for instance, which was work we did for Intel looking at the ‘connected car’ and how it might relate to digital technology in cities and people’s pockets.

Slide 21: Car as playhead in the city – video proto

A lot of it was video prototyping work – provocations and anticipated directions that Intel’s advance design group could show to device and car manufacturers alike – to sell chips!

Slide 22: Light-painting interaction surfaces in cars

We deployed our usual bag of tools here – and came up with some interesting speculations – for instance thinking about the whole car as interface in response to the emerging trend of the time of larger and larger touchscreen UI in the car (which I still think is dangerous/wrong!!!)

Intel cars: video proto of smart light car diagnostics – watch here

Here’s another example of smart light – computer vision and projection in one product: an inspection lamp that makes the inscrutable workings of the modern car legible to the owner.

Slide 24: manifesto

Something we wrote as part of a talk back then.

I guess we were metaverse-averse before there was a metaverse (there still isn’t a metaverse)

Slide 25: William Gibson’s obituary for BERG

I left BERG in 2013 – it stopped doing consulting and for a year it continued more focussed on it’s IoT product platform ‘bergcloud’ and Little Printer product.

In 2014 it shut up shop, which was marked by some nice things like this from William Gibson. Everyone went on to great things! Apple, Microsoft – and starting innovative games studio Playdeo for instance. In my case, I went to work for Google…

Slide 26: Google career vs Moore’s Law

So in 2013 I moved to NYC and started to work for Google Creative Lab – whom I’d first met working on the Lamps project. There I did a ton of concept and product design work which will never see the light of day (unfortunately) but also worked on a couple of things that made it out into the world.

Slide 27: Google Creative Lab: A Spacecraft for all

Creative Lab was part of the marketing wing of Google rather than the engineering group – so we worked often on pieces that showed off new products or tech – and also connected to (hopefully) worthy projects out in the world.

This piece called Spacecraft For All was a kind of interactive documentary of a group looking to salvage and repurpose a late 1970s NASA probe into an open-access platform for citizen science.

Slide 28: Google Creative Lab: A Spacecraft for all

It also got to show off how Chrome could combine video and webGL in what we thought was a really exciting way to explain stuff.

Slide 29: Project Sunroof

Another project I’m still pretty proud of from this period is Project Sunroof – a tool conceived of by google engineers working on search and maps to calculate the solar potential of a roof, and then connect people to solar installers.

Slide 30: Project Sunroof

We worked on the branding, interface and marketing of the service, which still exists in the USA. There were a number of other energy initiatives I worked on inside Google at the time – which was a much more curious (and hubristic!) entity back in the Larry/Sergey days – for good and for ill. 

Slide 31: Google Clips – On-device AI

One last project from Google – by this time (2016) I’d moved from Creative Lab to Google Research, working with a group that was pioneering techniques for on-device AI. Moving the machine learning models and operations to a device gives great advantages in privacy and performance – but perhaps most notably in energy use. If you process things ‘where the action is’ rather than firing up a radio to send information back and forth from the cloud, then you save a bunch of battery power… 

Clips is a little autonomous camera that has no viewfinder but is trained out of the box to recognise what humans generally like to take pictures of so you can be in the action. The ‘shutter’ button is just that – but also a ‘voting’ button – training the device on what YOU want pictures of.

Along with Clips, the ‘now playing’ audio recognition, many camera features in pixel phones and local voice recognition all came from this group. I thought of these ML-powered features not as the ‘brain-centered’ AI we think of from popular culture but more like the distributed, embodied neurons we have in our knees, stomach etc.

Slide 32: joining Moixa

At the beginning of this year I left Google and went to work for Moixa. Moixa is a energy tech company that builds HW and SW to help move humanity to 100% clean energy. More about them later!

Slide 33: Career vs Global Heating

Instead of overlaying Moore’s Law on this step of my career, instead another graph of an all-together less welcome progression. This is Professor Ed Hawkin’s visualisation of how the world has warmed from 1850 to 2018.

Slide 34: Fear, greed, despair and hope in climate tech

I’ve been thinking a lot – both prior to and since joining Moixa about design’s role in helping with the transition to clean energy. And I think something that Matt Webb often talked about back in BERG days about product invention has some relevance.

And we all love a 2×2, right? 

He related this story that he in turn had heard  (sorry I don’t have a great scholarly citation here) about there being 4 types of product: Fear, Despair, Hope and Greed products.

Fear products are things you buy to stop terrible things happening, Despair product you have to buy – energy, toilet paper… Greed products might get you and advantage in life, or make you richer somehow (investments, MBAs…) but Hope products speak to something aspirational in us.

What might this be for energy?

Slide 35: Ministry for the future & Family Guy

You probably thought I was going to reference The ministry for the Future by KSR, but hopefully I surprised you with Family Guy! It’s creator, Seth Macfarlane went on to create the optimistic, love-letter to Star Trek called “The Orville” and has this to say: 

“Dystopia is good for drama because you’re starting with a conflict: your villain is the world. Writers on “Star Trek: The Next Generation” found it very difficult to work within the confines of a world where everything was going right. They objected to it. But I think that audiences loved it. They liked to see people who got along, and who lived in a world that was a blueprint for what we might achieve, rather than a warning of what might happen to us.” – I think we can say the same for the work of design.

Slide 36: KSR – Anti-Anti-Utopia

I’m going to read this quote from Kim Stanley Robinson. It’s long but hopefully worth it. 

“It’s important to remember that utopia and dystopia aren’t the only terms here. You need to use the Greimas rectangle and see that utopia has an opposite, dystopia, and also a contrary, the anti-utopia. For every concept there is both a not-concept and an anti-concept. So utopia is the idea that the political order could be run better. Dystopia is the not, being the idea that the political order could get worse. Anti-utopias are the anti, saying that the idea of utopia itself is wrong and bad, and that any attempt to try to make things better is sure to wind up making things worse, creating an intended or unintended totalitarian state, or some other such political disaster. 1984 and Brave New World are frequently cited examples of these positions. In 1984 the government is actively trying to make citizens miserable; in Brave New World, the government was first trying to make its citizens happy, but this backfired. … it is important to oppose political attacks on the idea of utopia, as these are usually reactionary statements on the behalf of the currently powerful, those who enjoy a poorly-hidden utopia-for-the-few alongside a dystopia-for-the-many. This observation provides the fourth term of the Greimas rectangle, often mysterious, but in this case perfectly clear: one must be anti-anti-utopian.

Slide 37: Dear Alice for Chobani by The Line

I’ve been reading a lot of solar punk lately in search of such utopias. But – The absolute best vision of a desirable future I have seen in recent years has not come form a tech company, or a government – but a Yoghurt maker. This is the design of the future as a hope product.

“We worked closely with Chobani to realise their vision of a world worth fighting for. It’s not a perfect utopia, but a version of a future we can all reach if we just decide to put in the work. We love the aspiration in Chobani’s vision of the future and hope it will sow the seeds of optimism and feed our imagination for what the future could be. It’s a vision we can totally get behind. We couldn’t be more happy to be part of this campaign.” – The Line

Slide 38: Goal is not ‘sustainabilty’. Goal is to get to Type 1 Kardashev

In 1964 a physicist named Nikolai Kardashev proposed a speculative scale or typology of civilisations, based on their ability to harness energy.

Humans are currently at around .7 on the scale.

A Type I civilization is usually defined as one that can harness all the energy that reaches its home planet from its parent star (for Earth, this value is around 2×10^17 watts), which is about four orders of magnitude higher than the amount presently attained on Earth, with energy consumption at ≈2×10^13 watts as of 2020.

So, four orders of magnitude more energy is possible just from the solar potential of Earth.

A Type 1 future could be glorious. A protopia.

Slide 39: Moixa

At Moixa we make something that we hope is a building block of something like this – solar energy storage batteries that can be networked together with software to create virtual power plants, that can replace fossil fuels. It’s one part of our mission to create 100% electric homes this decade.

The home is a place where design and desire become important for change. I hope we can make energy transition in the home something that is aspirational and accessible with good design.

Slide 40: Saul Griffith’s Electrify

I’ve also been very inspired by Saul Griffith’s book “Electrify” – please go read it at once! It points out a ton of design and product opportunity over the coming decade to move to clean, electric-powered lives.

As he says: 

“I think our failure on fixing climate change is just a rhetorical failure of imagination. We haven’t been able to convince ourselves that it’s going to be great. It’s going to be great.”

– Saul Griffith

Slide 41: New manifesto!

I’ll finish with a couple more quotes:

“If we can make it through the second half of this century, there’s a very good chance that what we’ll end up with is a really wonderful world”

Jamais Cascio

“An adequate life provided for all living beings is something the planet can still do; it has sufficient resources, and the sun provides enough energy. There is a sufficiency, in other words; adequacy for all is not physically impossible. It won’t be easy to arrange, obviously, because it would be a total civilizational project, involving technologies, systems, and power dynamics; but it is possible. This description of the situation may not remain true for too many more years, but while it does, since we can create a sustainable civilization, we should. If dystopia helps to scare us into working harder on that project, which maybe it does, then fine: dystopia. But always in service to the main project, which is utopia.”

Kim Stanley Robinson

Slide 42 (of course!)

Conversation with Google ATAP / Shaping Things 2.0 vs Web3.0

Was interviewed recently by Anastasiia Mozghova for Google ATAP’s twitter feed, where it’s being published as a thread in two parts (part one here).

It centred around ambient computing and work I’ve been involved in around that subject in the past – particularly looking at notions of ‘social grace’ in computing.

This was a concept we discussed a lot in relation to Soli, the radar sensor that Google ATAP invented. I was involved in that project a little at Creative Lab, then more peripherally still (as kind of a cheerleader if anything) at Google Research.

That concept of social grace, for me, begins in the writing of Mark Weiser and continues amplified in Adam Greenfield‘s “Everywhere” – which still has some very relevant nuggets in it, IMHO, for a book on technology written in 2006 (16 years ago at time of writing!)

I also mention another classic – which is fast coming up on it’s 20th anniversary (next year?) – Bruce Sterling’s Shaping Things.

Its eco-centric vision of ‘spimes‘ rhymes (sorry) with the current hype around web3, tokenisation, etc etc – but while that is obsessed with financialisation, Shaping Things instead hints at using data sousveillance means toward resource responsibility, circularity and accountability ends in design and manufacture.

Time for a second edition?

Station Identification

Some of us know what we want: private sufficiency, public luxurydoughnut economicsparticipatory democracy and an ecological civilisation. None of these are bigger asks than those the billionaire press has made and largely achieved: the neoliberal revolution that has swept away effective governance, effective taxation of the rich, effective restraints on the power of business and oligarchs and, increasingly, effective democracy.

So let’s break our own silence. Let’s stop lying to ourselves and others by pretending that small measures deliver major change. Let’s abandon the timidity and tokenism. Let’s stop bringing buckets of water when only fire engines will do.

George Monbiot, The Guardian, 18th July 2022

“Private sufficiency, public luxury” – I can get behind that. A dream of equitable abundance. Wales recorded it’s hottest ever temperature yesterday. 37.1ºC. Fire engines are needed, indeed.

New gig, 6mo in…


Well, I had planned to post something about this sooner, but have been… busy.

At the beginning of 2022, after almost 9 years at Google I left and took on a new role as Principal Designer at Moixa…

Back to April 2013 – I move to NYC from London and join Creative Lab… One of the first things I work on is an amazingly ambitious project for Larry Page (remember him?) in what was then Google.org – to radically remake the world’s electricity grid for a zero-carbon future.

It gets a little mention here in a long article about Google’s newly opened Bayview building.

Called ‘Bottom-Up Grid’ or ‘BUG’ it aimed to make electricity infrastructure more like the internet – able to flexibly store, route and reroute power, making it far easier for renewables to scale. 

I’m making it up as I go along (which is what design is, remember?), with an awesome gang of talented people: creating speculative designs, dynamic simulations and infoviz, physical models, science-fair-like exhibitions to convince important people in nondescript buildings in mountain view.



He asks the very, very smart team we’re working with lots of questions including… how you might be able to cook pasta more quickly with high voltage.


But he gives it the nod.

The project continues for a while… until it gets a smaller focus, becomes a pilot internally, quietly sunsets. I went on with the wonderful Creative Lab teams to do many similar fantastic projects with Google Access & Energy which – sadly no longer exists now that we really need it…

Late 2021: I’ve been back in London for about 5 years, working in Google Research and having a lot of fun with a bunch of studios – imagining what an alternate decentralised Google might be in future with private, personal AI for everyone – but BUG still… bugs me…

I’d installed a solar & battery system to my home in 2018 from a UK startup called Moixa

It was the nearest thing to BUG I could find on the market: harvesting what energy you can from the UK sunshine (!), then storing it in a battery and sharing using innovative networking software…

And then… I saw they were hiring for a Principal Designer.

So… I figured I needed to go for it.

At Moixa we make something that we hope is a building block of something like this – solar energy storage batteries that can be networked together with software to create virtual power plants, that can replace fossil fuels.

It’s one part of our mission to create 100% electric homes this decade.

The home is a place where design and desire become important for creating change and impact in the climate fight.

I hope we can make energy transition in the home something that is aspirational and accessible with good design.