All posts made in Nov. 2015:

Filtered for air and light and war and stories


I was complaining out loud the other day about the distracting man I was sitting next to, hammering his keyboard, typing like a donkey falling downstairs. But then it occurred to me, I always blame external factors for ruining my focus when really I lack it for internal reasons. If I genuinely had focus, nothing could disturb me.

Eric piped up with this poem by Charles Bukowski, air and light and time and space.

if you're going to create ... you're going to create with part of your mind and your body blown away



Muji's mission statement:

MUJI's goal is to give customers a rational satisfaction, expressed not with, "This is what I really want" but with "This will do." "This is what I really want" expresses both faint egoism and discord, while "This will do" expresses conciliatory reasoning.


Machines generating stories about images.

He was a shirtless man in the back of his mind, and I let out a curse as he leaned over to kiss me on the shoulder. (Looking at an image of two sumo wrestlers grappling.)

Uses a technique with the astounding beautiful name of skip-thought vectors, a machine which is able to reconstruct the surrounding sentences of a passage in a book.

See also: A video of the same stories-from-images trick being performed from a live webcam feed: a man is eating a hot dog in a crowd.


From this explanation of Soviet Deep Battle theory, an insight into military science:

War is no longer a series of short and sharp engagements but rather a flowing affair, with larger, strategically oriented battles ('operations') that often encompass several smaller, shorter battles-within-battles (tactical engagements).

Which leads to approaches:

Deep Battle, or Deep Operations in particular first begins to develop as a theory in the 1920s. Like most developing theories of Mobile Operations at this time, it had one, over-arching goal: Get the battlescape moving, and keep it moving



I've been skiing like once and my main metaphorical takeaway was that it's easier to course correct when you're in motion. Try to turn when you're going forwards slowly, you'll tumble. There's a lesson there for company strategy, and I find myself reaching for this metaphor again and again. But now it turns out that military science understands movement, ability to adjust to circumstances, and flow, in a far richer way than me and my experience on the side of a mountain in Canada.

We tell stories to ourselves about what we experience, then we use those stories to approach the world. What stories we choose matters.

War has a vocabulary and a philosophy all of its own, and the fact I don't know anything about it tells me I'm missing out on something valuable -- as unpleasant as the subject matter is.

See also: Frieze magazine on the Israeli Defence Forces (from 2006) who, it turns out, are heavily influenced by contemporary philosophy:

Most important was the distinction [Deleuze and Guattari in A Thousand Plateaus] have pointed out between the concepts of "smooth" and "striated" space ... In the IDF we now often use the term "to smooth out space" when we want to refer to operation in a space as if it had no borders. ... Palestinian areas could indeed be thought of as "striated" in the sense that they are enclosed by fences, walls, ditches, roads blocks and so on. When I asked him if moving through walls was part of it, he explained that, 'In Nablus the IDF understood urban fighting as a spatial problem. ... Travelling through walls is a simple mechanical solution that connects theory and practice.'

A startling article.

Like to Continue, a fictionbot

I wrote a poem on Twitter. It's 36 tweets long, and happens entirely in your notifications panel.

Or maybe what I made is a fictionbot. You say "hi" to it and it tells you a story. You get sent each line only when you like the last. The story is about liking, and continuing.

So it's called @liketocontinue and you should just introduce yourself to start. Then watch out for what it tweets at you, and like to continue.

If it gets too much attention it'll break, that's part of the fun.

Whys and hows

You can tell I'm interested in chatbots and - with my business hat on - I'm especially excited about digital coworker bots, being pioneered by the likes of Howdy which helps you run meetings (see screenshots). All the energy is around Slack which is bot-friendly group messaging for work... a great product and a great marketing strategy: They've figured out how to make virality work in enterprise by having a frictionless on-ramp below the expense threshold and treating the team as the viral atomic unit.

And back in the day, I used to make chatbots that you used individually on AIM. For instance, googlematic let you search Google -- and that got me a bunch of nice attention, and in a bunch of trouble too.

But I'm into Twitter. Twitter is something between these and something different too. Twitter is a place where people talk to each other and groups. It's not quite personal, and it's not focused on work... it's public. I'm curious about what you can do with bots in public space. I'm in love with @mothgenerator and its gorgeous computer-generated moths. But more than that, there's something for me about interactions that happen over time, and interactions that can start with one person and widen up to more people, sometimes deliberately and sometimes accidentally because they're visible. It seems like there's a lot of creative potential there. Stories! Text adventures! Collaborative poems!

So much potential.

Which is why I'm taking my own advice and exploring the potential with art. Well I say art. Amateur poetry really.

I wanted to explore the feeling of a like and in particular waiting for a response, especially because Twitter just shifted from faves to likes. So that's what I wrote. Made. Wrote.

Technically, I have a basic Python 3 app that I use to get started on any new project. It has everything I want already set up... sign in via Twitter, a database capable of storing emoji, nice web templates, email error logging, solid deployment to my webserver, and an asynchronous loop to run background tasks like listening for tweet activity. Custom for how I tend to work. It's taken me a while to get happy with this (my coding is rusty) but it's neat that I can get something written and live in an hour instead of a week.

And I've learnt a ton about the tech things like Twitter limits and what you can and cannot see via the API (such as: you can see @-mentions from users you don't follow, but you won't get notified of their likes on your tweets). And lots of details about how to make a system where it won't break in-progress stories when I edit the words.

But mainly I've been seeing how reading (and having to like!) tweets feels, versus lines on paper, and how that changes what I write. So I've spent most of my time on the words not the code, which is just as it should be.

I want to keep digging with fictionbots. Like I said above, there's so much potential. If you'd like to collaborate, I'd be up for chatting... it would be great to work on a little project with someone who can actually write!

Anyway, nice to have shipped something, no matter how simple, or rather, snuck it out the door. Or rather rather - because it's a poem - published. I hope you like it.

Filtered for space and smell


2.8 million years ago, a nearby supernova caused apes to come down from the trees.

Based on the concentration of Fe-60 in the crust, Knie estimated that the supernova exploded at least 100 light-years from Earth - three times the distance at which it could’ve obliterated the ozone layer - but close enough to potentially alter cloud formation, and thus, climate.

(You know when Carl Sagan says we are made of star stuff. Fe-60 is a type of radioactive iron that is created only in stars, and gets out only when they explode. There is a very thin layer of Fe-60 dusting our planet, 2.8 million years down in the deposited mud at the bottom of the oceans.)


Around that time, the African climate dried up, causing the forests to shrink and give way to grassy savanna. Scientists think this change may have encouraged our hominid ancestors as they descended from trees and eventually began walking on two legs.

What an alarm clock for humanity! Where's the snooze button, let me stay in my tree.


Video of the new sport of drone racing.

First-person drone racing through the trees: You wear a virtual reality helmet, and race the drone along the track through the trees, seeing through a camera mounted on the front.

The future!


Future Forms is a collection of space-age electronics, primarily dating from the 1960s to the 1980s.


See also this gallery of Soviet PCs. Look at how much bright orange there is! Yum.

I wonder. I wonder. Was it because these PCs were made in the heady days of the Atomic Age?

A uranium glaze on ceramics has a wonderful glossy, bright red-orange finish. Was the orange uranium aesthetic carried over to plastic PCs?

It reminds me of the days in the early 2000s when all electronics had to use blue LEDs to look futuristic -- the technology of blue LEDs having been commercialised just a few years before. I saw my first blue LED - hot out of the labs in Japan and brought back to Oxford - in 1998. There were probably a couple hundred of us in the room when the LED was connected to a battery and lit up, and all of us saw the purest blue we had ever seen, all for the first time, the colour of the future. Physics. There was a collective sigh of awe.


Natural gas is used domestically for cooking and heating. A leak can explode or suffocate. So an odour of rotten eggs is added, in order that we can tell it's there when it's there. For safety.

So you wake up one morning after drunkenly playing some VR game and passing out, you forget you're still wearing your retinal projection contact lenses and your ear plugs are still connected to the virtual world, not amplifying the real. What's the odour we add to VR so you know you're still in it? What should be the smell of the virtual?

See also, this slide and the next: Blue is the colour of hyperlinks.

Filtered for reading


Rod McLaren's beautiful meditation on spreadsheets, and his spreadsheet art: Sandcastles and Spreadsheets.

The spreadsheet's unreality is dangerously doubled because, while their ordered data and formulae always comfort you that you have authored a controllable certainty, most spreadsheets are mere conjectures, provisional plans, ideas or hopes.

Spreadsheets are dreams.

Read that, next time you have a rainy day with your head buried in Excel.

Rod's essay is part of Beeker Northam's larger project, Hand & Brain with William Gibson and several others contributing. Something to get lost in.


Donald Knuth helped define computer science. Since 1990, he no longer has an email address: I'd used email since about 1975, and it seems to me that 15 years of email is plenty for one lifetime.

Email is a wonderful thing for people whose role in life is to be on top of things. But not for me; my role is to be on the bottom of things.

He replies to his correspondence for 1 day every 3 months.

Joe Nelson ran with this... Going Off-Line. I will check email once per week, on Monday.

On Twitter and Github I'll be entirely write-only. I’ll check replies/messages/issues on Mondays along with my email. ...

I will eliminate all use of the computer that is not directly related to creating things. If I’m not coding, writing, or editing videos then there will be literally nothing to do. I am going to dissociate the computer from mindless fun, from the capacity to kill time online.

I like the idea of taking a write-only sabbatical.

(Actually, thinking about it, I reckon that's why I'm enjoying writing on my blog so much. In the old days of blogging, a blog post was part of a conversation. Not today. Barely anybody replies, and when I do get a response it's lovely or thoughtful and often both. But I don't know if anyone reads these words; the links don't get shared out so much. It's very freeing, like having a notebook with a very slight incentive to better organise my thoughts. And my thoughts improve in that process. Write-only blogging.)


Will Self on Stonehenge:

"Don't you worry about the monument ceasing to be real in an important sense," I asked. "I mean, with all this messing about isn't Stonehenge in danger of museumification?"

The balance between putting the past on a pedestal, and living in it and carrying it forwards.


A few years ago, I transcribed and put online one of my all-time favourite short stories, The Author of the Acacia Seeds by Ursula K. Le Guin. My favourite story not just for the ideas, or for the turns of phrase and the humour. But for the gentle, determined build into a wholesale decentering of what it means to be human, and a huge widening of togetherness and empathy to the entire cosmos. Which, it turns out, is what I love science fiction for.

But don't read that.

Because here is a video of Le Guin reading from the Acacia Seeds at a conference and oh my goodness it makes me tingle.

Minutes of hardware-ish coffee morning, edition 12

Hardware-ish coffee morning last week was AWESOME. Thank you for coming - in no particular order - Nathan, Nat who has just launched a new invention studio called Buckley Williams, Maximilian and Heinrich from Kazendi, Rob who is behind the Ockham Razor, creative hardware engineer Saar, Grace from manufacturer PCH International, the Tingbot massive - which is launching on Kickstarter in a matter of hours - Ben and Joe and Ken from Nord, Naomi and Nick who make biofeedback games to regulate breathing at Shift, Josh (hardware accelerator Hardware Pro), Ines (hardware investor C4V), Tony and Christiaan of the accelerating meteor Pact Coffee which recently successfully launched Nespresso-compatible speciality coffee on Kickstarter, David of Pixie Labs, No Mayo Digital's Izzy and Clare, shipped-hardware-product BleepBleeps founder Tom, hardware-to-software maker another Tom, Chelsea who is behind the Olly table-top robot currently in development, and OpenSensors Internet of Things platform founder Yodit. And breathe out.

What an amazing group for chitter-chatter and caffeine!

Including me that's 24 people, and a quick stat -- we're down to two thirds (16) who would probably identify as dudes. Which is not perfect but better than it has been, and on the right trend. Going by my gut, it seems that female founders are a better part of the mix in hardware startups than tech at large, and I hope that goes for inclusivity of all kinds. Whether or not my gut feeling is correct, I certainly want the London hardware community to be a leading edge of London's inclusivity, and that's why I track the NADQ at this coffee mornings.

NADQ = Not A Dude Quotient.

So if you're a woman or not a down-the-line dude, thanks for coming! If you invited someone, thanks! I know it's a bit weird to keep calling out my coffee morning NADQ like this, but trust me it's weirder when I sit round a table with six other men on a Thursday morning.

There were a ton of new people today. That was lovely to see.

What was discussed?

Honestly who knows what was discussed.

Hardware-ish coffee morning sprawled over about five tables, with double that many conversations and people continually moving round. I have no idea what people talked about. But I had to leave just after 11 and people were still going!

I was darting about like a headless chicken so I didn't get to say hi to anyone and my attention was always elsewhere. BOOOOO. And sorry to everyone I know that I didn't get to chat with!

BUT a couple of things I did notice.

  • Tingbot is a lovely add-on for the Raspberry Pi that turns it into a hackable home gadget, a bit like an alarm clock with an app platform. Here's a pic. What's special is that it has a dead-simple programming environment, a bit like Arduino, that makes writing and running graphical, internet-connected apps very simple. Tingbot goes on Kickstarter this week -- follow @thetingbot on Twitter to get the latest.
  • Olly is a robot currently in development (in London) with all kinds of voice features. Chelsea talked about it as being table-top - which is an intriguing context because it feels like that's where the old school kitchen radio sits - and the character-driven, emotionally-responsive robot she talked about sounds fascinating. One to watch.

What's also interesting to me is that we had pre, post, and intermezzo Kickstarter projects present, from both small and pretty seriously established companies, plus investors and manufacturing. A maturing scene. There's a lot of knowledge in the room. the point that, discussing one project that is imminently adding hardware to their existing software-only offer, I can assemble the roadmap with them, 50% from my own knowledge (what the critical proof-points and bottlenecks are, what needs to happen hand-in-hand with what), and 50% simply from looking around and going: Well, first do it like company X, and then do product development like company Y, and finally you will end up speaking with someone like company Z and here are the questions to ask.

Next time

Let's have one more hardware-ish coffee morning this side of the new year. I'll take a look at my calendar and see if I can pick a Thursday which won't be too conflicted with the holidays. Join the email announce list, I'll send a note there to arrange it.

UPDATE: Tingbot is now on Kickstarter! Go read more about it there. And please back it, I want a Tingbot of my own.

Some ideas for banking apps

Something interesting is happening in UK retail banking... a transformation -- as far as I can tell it's triggered by the impending (and catchily named) European PSD2 which includes new rules designed to open up access to payment account information to third parties and the UK preparation for this, commissioned by the Treasury and led by the Open Banking Working Group, defining exactly how we'll be able to, for example, plug apps into our current accounts.

Q. What?

A. This:

If you've ever complained about your sucky bank mobile app, this will let third parties replace it. If you've ever used Transferwise because it's a cheaper and easier way to pay for a holiday hotel, this will open up competition and make the banks get cheaper and easier too. If you've ever wondered why, in the UK, we don't have apps like Acorns which automatically rounds up all your transactions to the nearest dollar, and sweeps the round-ups into your investment portfolio, this will fix that.

Naturally the UK banks are scared they're about to be commoditised AND responding with vast and impressive innovation efforts.

This is anecdotal. I have a friend who works in retail banking, I've had a meeting with a couple of other banks, and I'm hearing faint noises on the grapevine.

So I was talking to my friend, and wondering - this opening up of UK banking - which will be a cross between being able to move my mobile phone number between operators (which wasn't originally possible) and Youtube and iTunes which democratised music production... what will happen?

It might be like newspapers. It turns out newspapers were an accident of distribution. They were really good at printing for cheap and getting bundles of papers into every pair of hands in the country. But then the internet emerged as a rival form of distribution, and the newspapers were unbundled -- classified ads to Craigslist and then Facebook, ads to Google, breaking news to social media, expert comment to blogs. And as the readers go, so does ad value. It's a death of a thousand cuts, and although journalism is still useful, it no longer sits catching cash at that valuable mountain pass. We'll have to find a new way to fund it.

You know, no huge loss. It's important we find a way to properly fund investigative journalism, but it doesn't need to be these particular journalists or those particular publications.

Or it might be like Uber.

What happens when driving a car from point A to point B is no longer a specialised profession, when Google Maps can tell you what to do?

What happens to retail banking when...

  • overdrafts come from a pay-day loan company
  • the mortgage market is frictionless
  • investments are made with an app that seamlessly integrates with your current account, gamifying savings and the stockmarket both
  • getting a new third party credit card is as easy as downloading an app, thanks to Apple Pay... this could be the golden age of American Express
  • switching your standing orders and direct debits is either automated -- or unnecessary, because you take your current account number with you between banks
  • and now you've almost forgotten who provides your underlying personal bank account, and you haven't visited a bricks and mortar branch for years, you switch to another current account provide with a cuddly logo, some other bugger who can spend more on marketing because their operating costs are radically lower because they run all their banking with modern web tech instead of ancient and expensive mainframes

Goodbye existing UK high street banks.

All of which means the question becomes:

As a retail bank, what do you do to ensure you're not just the plumbing, that you provide enough value that (a) customers come to you and stay with you; and, (b) customers use you enough that there multiple low-friction upsell opportunities to those services that actually make a profit?

In short, how can my bank be a platform more like iOS (as gorgeous as Android is, I'm never going to switch because I can't be bothered to re-download all those apps and learn new habits) and not a platform like my electricity suppler (electric potential is electric potential, switching is a phone call, and I'll give my money to the folks who build those beautiful fields of windmills thankyouverymuch).

I was chatting with my friend in banking (remember I said I have a friend in banking) on Friday.

I said:

Look, what apps can you build on the current account. Thinking about that service in the US that sweeps your small change... there's Digit which is a text-message-only artificial intelligence that helps you save money. Clever. Catchy. And I'm sure I read about one that watches your current account and automatically donates money to charity.

Well there's a service like that charity one in the UK, something offered by Lloyds that handily and easily donates your spare change to charity: Save a few pennies every time you spend with your Lloyds Bank visa debit card. Problem being I've never heard of it.

And of course I haven't. Lloyds has a thousand products. They aren't existentially threatened if this particular one doesn't take off, and it's a marketing expense anyhow so it's barely possible to tell whether it matters. But the charity idea is a good one.

How about this -- spitballing an idea...

What if a consumer bank partnered with an online fundraising service, let's say Just Giving which makes it super easy for charitable causes to set a goal, and allows individuals to spread the word through their social networks.

They partner and set up a new form of charitable giving called "sweep" -- it gives your spare change to some some organisation for 3 months, say. And it's frictionless... one-click for this particular bank if you already have an account, tap and it's connected, enabled by the technology built to meet European PSD2.

Just Giving (or whatever service) is crazy incentivised to market this. It's a competitive advantage, it's existential for them, they live or die by that KPI. And the bank... well, they do what they do best. Only now their name is out there, customers have both the warm fuzzies and a new incentive to stick with this particular account. Sure other banks will catch up, but keep building apps and get partners whose interests are exactly aligned with getting the word out and making the partnership succeed.

Maybe the next one-click integration is Square retail payments.

Maybe buy the Square kit in the UK, you click a button and BOOM you have a small business banking account with insert name of forward-looking UK bank here. Sure you can't withdraw any cash yet due to European anti money laundering regs but hey the same is true with PayPal right, so come into the branch whenever you're ready and we'll do the paperwork. Meanwhile you can run your shop, go ahead, make sales.

Where does this lead?

Perhaps, just perhaps, it leads into a form of customer relationship which is more relevant to people under 40 than in-person meetings and robot voice phone menus.

Why shouldn't I be able to follow my current account on Twitter and get a direct message when my salary hits my bank? And if that same current account asked me, in that same DM, Hey, Matt, look, I can set you up an ISA, you've got a couple hundred a month spare to save into it, you up for this Y/N? Sure I'm going to say yes.

These are simple concepts to implement. But they're also forward looking, aligned with strategic interests, and might actually get some attention.

My point is, for the UK consumer banks, the ideas are ten a penny. The difficulty, as ever, is going to be the organisational change to take the opportunity, and keep taking it.