Trade digitisation: Do not rush, spend Kairos time with the cyborg-anthropologist

TXF: We're in the Swift business forum in New York and I'm joined by Amber Case who is going to talk to us about how technology is impacting finance, and to a certain extent global trade and in our day to day lives. Can you explain what your central premise is? 

Amber Case (AC): The principles of calm technology came out of the Xerox PARC [Palo Alto Research Center] in the 1990s and what I try to do is dredge up things that people forgot about. Usually when they're created they're created out of phase with the era that they're supposed to be in and so calm technology was really ahead of its time. Calm technology said that at some point in the future technology would be cheap but attention would be expensive.

At that point how technology engaged with our attention would make or break it. So if you look at people managing portfolios and funds they don't have attention to give the right customer service to everybody. There's people staring at lots of trading screens. There are now robo advisers, there are people trying to send a wire transfer and all this is still stuck with lots of paperwork to fill out. It's not necessarily as quick as it could be. Credit card fraud is still an issue. Everything that we had in the 1990s is still true today.

And so it comes down to design. If you design your automation better to aid people not just make them babysit a machine, then people can be more efficient and effective about their jobs. It's all about amplifying the best of what humans can do and what machines can do.

Machines are really good at figuring a little bit of a pattern in long term data and coming up with some anomaly. But it's the human's job to look at that and say "OK check this against real data and actually see if that's a leading indicator of something".

A lot of people expect AI machine learning to tell us the answer, to be wise, but we are the ones whose imaginations and wisdom need to be applied to it in order to give us something back. The problem is it's an all or nothing type scenario in the world today.

There's this idea that either AI will save us or AI is horrible and will destroy us. And the idea that AI is horrible and will destroy us is from films. The idea that AI will save us completely is also from films. Neither of these are true. Automation exists. It exists since the first tools. Once we started evolving outside of ourselves having hammers as an extension of fists and knives as an extension of our teeth.

We were able to trade out those tools and work alongside them and those tools enhanced our capacity. They became extensions of ourselves in the same way that an automobile is an extension of our self and allows us to go faster. If we look at the history of tools - my background in anthropology - this way we can actually make better decisions on financial technology automation. Whatever group of people that we're part of, whatever we need to do to get our job done, we can make a better decision, if the technology is simple to understand and helps aid us. That doesn't mean that we need to understand how to build the technology. But if we can understand the system around it then it can help us along and we can add our imagination to it.

TXF: So put the intelligence into artificial intelligence?

AC: Yes. Or as Mark Weiser, who's one of the founders of calm technology, said "we don't need smarter devices we need smarter humans". How do we have devices that aid our intelligence?

TXF: How do we? What's are the next steps with this?

AC: Some of the principles of calm technology are that technology should take the least amount of attention and only when necessary. We should make a better use of our peripheral attention. Just think about how a car is designed. In a car you don't stare at the foot pedal the whole time. Could you imagine if a car was built like we built technology today? You'd have to stare at every single piece you'd have to wait for something to load, you'd use Ethernet, you'd Bluetooth into the streetlight and have to change. These horrible things. Or you look at something like an escalator. When an escalator breaks it turns into stairs. When a car breaks all sorts of things have to go wrong before it's fully broken. We use a foot pedal and sometimes we use a stick shift. We have the rear view mirrors that we can glance around us. The whole car is about using our periphery to enable us to focus on our task. Whereas with the phone we're dropping all of that and just using our primary attention.

A lot of design methodology is still from the era of the desktop. The '90s era, which is 'you have all the attention in the world, you have all the power in the world and all of the connectivity and you can stare at a perfect screen with lots of information'. But the world is not a desktop we are computing in-between moments and we need to have information and our peripheral attention whether that's 'the server is down' or 'your trade is in trouble' or 'there's a new piece of information' and there's little lights or tones in your environment or a soundtrack that that makes you aware of changes.

I know that particle physicists do this. They were sick of staring at the screens of all the different particle data events and particle accelerators. So much information. They started to personify the information so you could just sit there as a particle physicist at CERN and kind of listen and if an interesting particle event happened it would be a very specific sound and you'd say 'Ah now I can look at the screen'. They were able to do all of their other things that are interesting intellectual things while listening just like we listen to the radio. That's why podcasts have become so popular. Something that we can listen to while doing something else versus you know looking at Netflix but oftentimes we look at Netflix because we're falling asleep to it. It's like something that we're also using as peripheral attention. There's a tremendous power in peripheral attention that we've forgotten. We forgot the foot pedal - that was the Industrial Revolution, the car we forgot these mirrors, these indicators and these tones and lights. These things all came [in] home appliances, badly. But there's a lot of opportunity.

Let's say healthcare there's this thing called the alert fatigue and doctors and nurses hear between 1000 and 10,000 alerts a day. That's one of the leading causes of mistakes in hospitals.

A new nurse will answer an urgent request in seven minutes. A veteran nurse who's been there for a couple of years answers in 40 because they'll be fatigued, because no longer do the alerts mean anything. Plus they're in a really bad frequency, they annoy people and they keep patients up.

These kinds of principles can be applied to almost every industry. It's all about getting your mental space back so that you can do real things again, so that you can have what the [ancient] Greeks called Kairos [καιρός] time. There's Cronos [χρόνος] and Kairos. Cronus is the industrial time. We're having a meeting right now it lasts for 30 minutes. It's very formal, right? And in that meeting we're supposed to be doing these things and when it's done it's done. This is oftentimes in a corporation especially out of the industrial revolution, corporations are about repeating the same thing again and again. Cronos time. But all of the advances, innovations, weird stuff comes from Kairos time, comes from you know the CEO going on vacation and in the middle of the night saying 'oh my gosh our pricing models are all wrong'. They could never come up with that unless they had that 'squishy' time. I mean it starts with Archimedes sitting in the bathtub learning: 'Oh this is how we get density, eureka'. He runs through town and tells the king this is how we can figure out whether gold is false or not, your crown is fake. It's all about taking a break and doing something else.

So most of our companies don't have any Kairos time built in and most of our technology is in Cronos time so we just end up getting in this situation where we're thinking exactly the same way. We're not being very innovative and all of the alerts just show up in that time. If we can start to take these alerts and put them into your peripheral attention and get back some of that Kairos time we're not going to make the same bad decisions again and again. This is why big companies buy startups because startups have more Kairos time. Startups have that in-between time, weird time, they're very excited about things and then when they get bought the Cronos time gets loaded upon them and within a year you know they're usually out of there.

TXF: TXF has still got our foosball table so we're hoping to get the peripheral thinking, the alpha waves, the things that happen when you're not really thinking about 'the thing'.

AC: And the foosball table is often a way to say we have Kairos time even if you don't really have Kairos time right it's 'hey look there. There could be some time.' But then everybody feels guilty to do it as it's often in a very obvious location.

TXF: Right next to my desk.

AC: If you use it everybody hears it in the office. It's very obvious who is using it you know and why would you use it during daytime. You need to be busy, busy, busy you know. This is a very American thing. I like going to Sweden and some of these other places because you go home at a certain time and so in some places like Germany or in France where you don't have to answer your work email, you're not required to. The decisions that you make within that time limitation are better. You're not sitting there in the meeting saying OK well we'll say that we'll do this on Monday. We have to get it done because we have to leave the office but we can't just sit around. And so having less time and saying what's the most efficient thing or the clever thing to do, or 'we don't have time to train on a new system so how do we fix the existing one'. These are more boring management methodologies but they end up saving a lot of time.

TXF: Traditional trade finance is very much paper based, very much seen as a backwater in certain respects but it's being digitized in small pockets. Having said that are there lessons that the digitisation of process and of technology can learn from making things a little bit easier, calmer, smarter?

AC: Absolutely. A lot of it is contract automation. There's a lot of room to grow there. There are programming languages coming out that are legal programming languages so you can speak in legalese and then turn that into code. That is pretty interesting. The question is how is the digitisation done?

I had a lot of friends in law school and they said look 'to fund our bar exam, we're going to digitize some medical records and started a medical record digitisation company. They went to all their family and friends who are doctors and lawyers and they say 'we'll digitize your practice'. And you know it was interesting because on the one hand you can access all these things online but if the system goes down, is there a backup copy. Another issue was they had a secondary thing that they had to work on. If if you're dealing with paper and you're faxing and mailing things you have a record, you know which fax was sent where, you have phone numbers you're calling people up. There's an analogue social network around that. Once you digitise it a lot of these things start to disappear. So you know less about the person that you're sending an item to. Plus if they're stored in a central repository you have a security issue. Not only are you doing your primary job of finance. You have a secondary job of security. And not every company, even security companies, can do both jobs well.

There's another way to look at it which is decentralisation, not like a blockchain but just everyone has their own database. And when something changes all the new information gets written to your database or your company's database. And that way when somebody hacks your system, because they will, they're only getting what's shared the last two hours on that database. So it's less you're always going to get hacked but it's not putting the onus of security on your individual company.

Because security is expensive and it's hard to get funding to do security before there's a breach and then afterwards you're out of money as you have to pay out.

Just setting up systems in the opposite way is interesting. I'd like to see that for medical records. I go in, I bring my medical records with me, I get them written onto, I take them home with me. Because it's personal data, it should be near me.

These models aren't perfect. We're going from a mechanical state where things are very well known like on the web or with a fax or with analogue paperwork. This is exactly what it is right now. This paper won't ever change. Digital means corrupted documents as you know, but it also means speed. It doesn't mean improvement of process, sometimes it makes for more process because people are dealing with more things. So it really depends.

I would look at different industries and ask 50 people what was your digitisation process like. What did you regret? What did you like? What would you avoid so that when you do your own process you can avoid all those things? It's like that for anything. I wanted to write a book so I asked 50 people who wrote books so that you can learn from other people's mistakes because there's other industries that are very heavily digitized like the credit card fraud industry.

They know exactly what percentage they can send to an automation to check for fraud and they know exactly what percentage to send to a human being. They have it down. They know that they'll never be able to fully automate. And that's a good thing because they know that they've gotten to the point where they understand that there are some things that are better done by humans. The company that bought our startup back in 2012 had very well paid admins and admins were the stewards of the information in the company and they were around for 10 or 20 years.

They really understood all the paperwork and the bureaucracy in the company and they hand held people to walk them through the paperwork and that ended up saving weeks and weeks of time for everybody and because they were the experts.

And I think a lot of automation shifts the idea that we should be experts onto us. So instead of calling a travel agent who's really good at the MTI matrix system or whatever they used to book their flights you go through Google Flights. So you're becoming an expert in travel. But you're becoming an expert in everything. So when you go to the grocery checkout no longer somebody clicking something and feeling proud about it, you're the one doing a digital checkout. It ends up taking the same amount of time because you can't scan the barcode. Somebody has to come over and check your ID if you're buying alcohol. And so it becomes this issue where the labour gets shoved onto the end consumer and consumers are never going to be perfect with these devices because every supermarket has their own digital system and the whole point of having a worker is to know the system well and then provide customer service on top.

Or like when people get wealthy they don't buy automated things, they buy handcrafted Italian leather shoes, they buy local foods. They vacation in Italy where they can eat food nearby.

TXF: Vinyl is the new digital. I guess these go back to source. Does that actually have implications for small companies let's say banks have to a certain extent offloaded some of these things because they're not taking the risks themselves and are partnering with fintechs, etc, to take some of these risks at a more individual level. Is there still a role for these banks in a way? We're here at Swift after all. 

AC: Yes, there's a tremendous role for banks. There's a few different ways banks can play. They can partner with companies like Simple who have a very good user interface, a good digital user interface and just make it very simple. Banks are generally not going to be able to make a good user interface.

So they could partner with a company that does. And then just be the bank behind the scenes, handle all the regulation that's hard to do. In Portland Oregon there's credit unions there's Umpqua bank. There's these friendly banks you know in Spain there's a very friendly sensory bank that gets people to go in and hang out and you want to go to their bank. I mean for a lot of people who are busy they want mobile deposits they want mobile fraud detection. Through Google Hangouts I get fraud detection. I get notification of every single thing that I've spent and I can see immediately and then I can contact. I don't want to have to call customer service. I want my bot to do that. I want to send out a bot that says hey there's fraud. Can you handle this? And then they send thing back and says it's been done. So there's just more user experience stuff so that I won't have to call somebody out.

I want very easy transfers so that I can do my small business from the road from any country. I want it to be very easy to have the documents that I can get paid around the world. I want all of the VAT and different taxes automatically handled and calculated in real time depending on what it is. All the currencies, etc. There's a lot of room to grow on the design side  bringing in capacity, turning people to superhumans. I want a whole bank in my phone. Plus the customer service. Simple has done a really good job with that. There's other apps like Robin Hood, free trades it's great. Got it. Been using it for four years. I try to use E-Trade and it's just like why should I have to use a desktop to bank or to invest, I'm going to be late on my trades.

TXF: You want the API to be embedded?

AC: I think so. Banks can provide a tremendous capacity and power but there needs to be a partnership with good design. And once that good design is there you know all that the banks behind the scenes guarantee and security, etc. But it should also be a friendly face. So the more automation that happens the more important customer service becomes. I remember getting money from our investors and putting it in Silicon Valley Bank. There was no bank branch that we went to. But this guy showed up at our office and did all the banking with us just this guy that we knew. Hi I'm your representative in Silicon Valley Bank. Let me know what you want. Wow. Not kidding.

I don't know if millennials would want somebody to come to their house to handle all that stuff but I'm sure they will want somebody to help them set up a 401k and explain it to them. It's transitioning but banking is ugly, unapproachable and it's miserable for a lot of people a lot of people are afraid to check their bank accounts. You know the fact that every single bank statement is texted to me after every single transaction is really nice I know exactly where I am with my finances.

But not a lot of people have that set up. And so they have no clue. I'm speaking from a millennial perspective right now but services like Acorns can come in and say 'hey round up your spending money. So you buy something for $2.25 we'll round it up the nearest dollar and you invest that extra pocket change over time so you don't even notice. That's cool. There's a lot of different opportunities for these kind of in-between solutions.

TXF: It doesn't all have to be blockchain?

AC: No it can be really straightforward. It can also just be pre-filling paperwork for you 'here's your tax documents prefilled' and you just have to check it. It might not be filled completely properly but at least you have a start. And that's where the kind of machine learning or at least just automation, I just like to call it automation, gets you.

And if you know if you're like a bank or something and you want to do automated tellers just go to everybody who's digitised already and say to the customer representatives 'which one do you like? Which service do you like? Which model do you like and which system do you like?'

And you find the system that they love, buy that. You've already had everyone beta testing it. And if you find that everybody hates it and there some service rep that's trying to get you to buy their version of automation with a lot of steak dinners don't do it. And then you get a vendor lock-in and you've trained everybody and you've actually lost money. So in some cases automation loses money because there's so much maintenance. In some cases it really helps.

I've been tracking Square because it enabled after the recession all of those tiny little food truck vendors and small businesses to not spend $40,000 dollars on a POS system with some crazy lock in for two to three years. I mean that ate the market. In exchange Square gets a transaction hit or a nice transaction bonus but they also get to see all of the data on all the small businesses and which ones to send loans to and they can sell that to banks.

Wow, that's amazing right? So again Square is a design interface the banks could partner with. All sorts of different partnerships in the design realm that will really amplify small and large banks, it's just looking in the mundane. What's changed in the landscape that people are actually using, not pie in the sky ideas are not the promise of blockchain or just a little bit more tech and everything solved. And then partner with those companies with the best design. There's a few designers that said we're going to make a $10,000 portfolio 10 years ago and they said we're just going to invest in companies that have exceptional user experience.

So they had Amazon and Netflix, Google and all these different companies and they made a lot. And it was just kind of a silly idea. That's the best thing. And then once you buy a company don't destroy their user experience department but empower the lowest common denominator. Empower all of the customer service representatives. If you have really great employees try to get rid of as many middle managers as you can which is really hard. Try to find out the quiet people who are doing a great job. The ones that you haven't noticed, the ones that aren't drawing any attention to what they've done and reward them. It's really hard to do, it's kind of backwards. That goes for every industry as well because those are the people that have the touch points. I'd like to see customer service reps getting paid $90,000 a year. You have automation helping them. So you've actually saved money and you used some of that money to have the best customer service you will get so many customers. Just look at Virgin Airlines. All they did was change the lighting in the planes that just a few little pieces. All you had to do is some sensory tweaks and then suddenly 'wow this was a great thing'. Same with banks.

TXF: What will this look like in, say five years’ time?

AC: The companies with the best design will win.

TXF: And that will apply to all trade finance companies and banks and people moving money along supply chains too?

AC: Yes, and design doesn't mean how it looks. It means how it works. The companies with the least amount of components.

I was just speaking at ARM who make [silicon] chips. So why did they win as a chip company? Well they made a chip with the least amount of components to get the job done. And then they always worked within those constraints they said let's make it the smallest, easiest to produce, or hard to produce but long term.

Look at the technologies that have been around for 20 years the background behind the scenes. If you are working on blockchain can you make the speediest blockchain? What can you do that will last the test of time that reduces the amount of parts. Do you really need to automate something? Can you have a human do a better job?

The companies that I've seen that focus on rewarding employees are winning and are going to win much more, like Ikea. The companies that have great design user experience turn users into superhumans, they will win. And then the companies with boring background processes that are very fast like Google will win. So it's the same thing as what's been winning already and those big companies that we admire it's like well Apple works about user experience. Google works on very high speed. Virgin worked on treating people well.

They'll all get bought, they'll all get consolidated, some of their spirits will be crushed. But those are the companies that people are investing in right now aren't they. I mean Microsoft's looking to end user experience and they've done a pretty good job and they also have great dividends you know. Those are some things to look out for.

It's pretty much always the same.

TXF: Keep calm and don't get crushed!

AC: Yes stop racing to implement all the newest tech.

And if you do you'd better be a Jeff Bezos type person, you'd better understand 'why?' Don't just automate because automation is cool.

It's not cool it's just a series of issues that you haven't seen yet. If you're going to have the same issues as what you had in your company before it's just new issues and you don't know anything about them. Figure out why you're going to automate and take your time but don't take your time and just like delaying decisions, spend some Kairos time and interview lots of people and see you know that kind of qualitative analysis will save you years. You won't buy the wrong system and listen to the lowest down on the food chain employees because they're the ones with the insights.