Cloud Native Computing Foundation KubeCon + CloudNativeCon North America 2018 (Seattle), 16 Dec 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Continuous Delivery on Platforms is Better for Your Brain (and Business!) - Daniel Jones

Description

Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io

Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects

Continuous Delivery on Platforms is Better for Your Brain (and Business!) - Daniel Jones, EngineerBetter Ltd

Continuous Delivery on a self-service platform like Kubernetes makes it possible to exploit findings of cognitive psychology and neuroscience to increase the productivity and happiness of our teams. This session will highlight fascinating academic research that answers: • How can we improve willpower and decrease technical debt? • Is the present bias real? How can we turn it to our advantage? • Can you increase a team’s effective IQ? • How do DevOps and Product Teams increase empathy, and what impact does empathy have on productivity? Why do you need a self-service platform for this to work? The OECD and Bank Of England are both puzzled over missing productivity. Are we failing to exploit the benefits of modern technology by running our software businesses and teams as if they’re post-war factories, rather than continuously delivering?

To learn more: https://sched.co/GrT5

A

Good afternoon, everybody thank you very much for coming all the way over to the second building and join us, for this is the last session the day right. I am the last thing. Keeping you from party and beer is that correct, okay, well, I'll, try and make sure that I don't ramble too much and keep you any longer than is strictly necessary. So my name is Daniel Jones and I'm. The CTO of a consultancy called engineer better and to keep it brief. We help people deliver value faster, we're all about productivity.

A

We do cloud platforms as well: kubernetes Cloud Foundry, that kind of thing, but for us those are all amines to an end to delivering business value and letting companies do what they're supposed to be doing more efficiently and more effectively. So a couple of years ago, there was a really interesting article on a BBC News website, and it was all about how there it was a lag in the expected productivity gains after the invention of the electric motor and the electric generator.

A

This is a cotton mill from England in the 1800s, and you can see running along the top of that image. There's a really long crankshaft and that's there, because the factory was powered by a big steam engine, so one central steam engine and that choice of technology dictated the shape of the factory.

A

They had to be long and thin because they'd have these big long, crank, shafts to transmit the power to all of the machines, so the choice of technology dictated the shape of the factory which dictated where the machines were put, which dictated how people worked.

A

The electric motor was supposed to change all this, but it took a good 20 years before any of the productivity gains that were expected materialized.

A

Now we're gonna come back to this at the end, so try and remember this bit, but this song is all about findings from cognitive psychology and neuroscience as to why continuous delivery works better with human beings.

A

We're not constrained by steam engines today. But we are constrained by management practices. The relevant methodologies that were inherited from the incumbent enterprises of the day when software engineering became a discipline in the post-war era of mass production. The successful companies were those that were building a million identical widgets, so we copied their playbook. That's not really what software development is about. We shouldn't be making a million identical classes. You make a million identical classes, I'm gonna fire. You ain't, you sure, made a reusable component a long time ago.

A

We're not about doing the same thing repeatedly cheaply. We're about doing novel things as quickly as possible. Continuous delivery on a platform is an important part of that, so you probably all familiar with a situation that looks a little bit like this. An enterprise cut up into silos, specialist teams with one particular function.

A

Only problem is, if you want to deliver any business value. You've got to touch all of those departments. They all need to contribute they're all on the critical path. Now, in these kind of organizations, the kind of organizations we work with, it often feels to the people in one silo, like the people in another silo, don't really care about their problems, but why they all work at the same place. They've all got the same corporate values.

A

They say all got the same mission statement poster at the end of their row of cubicles, surely they're all working together? Could there be something to this feeling that the other themes don't care about them turns out there might be? This is a functional magnetic resonance, imaging scanner. It's used to measure levels of activity in different parts of the brain over time.

A

It does that by looking at oxygenation levels of blood in the brain, professor David Eagleman performed a study where a control group of subjects were placed into one of these scanners and whilst they were in there, they were shown one or two videos, either her hands, being gently tapped with q-tip or painfully stabbed with a syringe that was, the control group and baseline measurements were made of levels of activity in areas of the brain associated with feeling pain and with feeling empathy were measured.

A

So empathy being your ability to put yourself in somebody else's shoes with them yourself into that position, so that gave us the baseline measurements and then the experiment was repeated with a different group of subjects. This group of subjects were asked about their religious beliefs before they entered the scanner. So are you Jew? Are you Muslim? Are you seek? Are you an atheist this time when the videos were shown of people being painfully stabbed with a syringe? Their religious identity was shown on screen when the observer shared the same religious identity.

A

As the person being hurt, they showed a larger empathic response when the observer did not share the religious identity of the person being hurt, they showed a smaller empathic response smaller than when there had been no labels.

A

This showed that people care less about out groups as soon as you start cutting people up into my tribe in your tribe, my team and your team. We've become less able or less inclined to put ourselves in other people's shoes. So, let's think back to our organization Wow, it's no wonder they don't care about each other's problems start up they're, not even working together, they're in silos. They don't cooperate properly.

A

It solve a service cues and awfulness like that, but even when they do work together, they're going to be less inclined to empathize with each other's problems because of their different identities.

A

Luckily, this is where platforms like Cuba Nettie's can come in and help us a good platform allows a smaller problem space. You don't need to have such broad and deep knowledge in so many areas. So you don't need all of those silos to be on the critical path.

A

What do I mean by a problem space if these are all the silos that we had all the functions that we needed to deliver business value before? Well, it adds up to a problem space like this like this is all the stuff that you hats care about, to deploy an app to get an idea into prod if you're using a platform. Half of this stuff goes away right. That's all done by the platform team they're no longer on the critical path.

A

You still need those people, but a lot of that stuff isn't exposed to app developers anymore. Even things like databases, if you're using service brokers and Service Catalog, you don't need to know how to stand these things up anymore.

A

You just need to know how to consume them, so the size of stuff, the size and the problem that we need to consider is a fraction of what it used to be and when we don't need such broad and deep knowledge in so many different areas, we can focus all of our people in one cross-functional product scene with a shared identity.

A

One group of engineers working on solving business problems, writing apps, deploying them through an automated platform that shared identity gives them a common purpose.

A

Once we've got that cross-functional product team, we can co-locate them extra benefits to a trio of Norwegian. Scientists performed a study looking to explore the correlation between productivity and human capital, so the education and experience of people working at a variety of consultancies. They also looked into the correlation between productivity and social capital. So how much people talked with in these businesses? How regularly they communicated and with how many other people, in all three cases, social capital positively correlated with productivity in two out of three cases, the human capital?

A

So the education and experience of people working at these consultancies had no relationship to productivity whatsoever, so all of that money, they're paying for education and experience- didn't really make a difference. It's their bottom line. Now, if you think of a business like a network, this makes perfect sense right. If this is the telephone network- and there are two telephones- and you can only make one call- it's not very productive right.

A

It's not very useful, not very much utility from this, but if we have more nodes- and we have more connections between those nodes, then it becomes a more productive network. That's why cities are an awesome place to put your business. That's why conferences like this are awesome because you can intermingle with so many other people.

A

Now we can make this network more productive by making sure that the communications are reliable and that their intent is understood first time. If these nodes of people, then we can do that by exploiting some quite cool abilities of the human brain.

A

Now, when you saw that towards the end, maybe you might have raised your eyebrows a little bit widened your eyes if you did, that was a microexpression and that could have been caused by something called mirror. Neurons now mirror neurons are a brain cell that shows a particular type of behavior. It's not physically different.

A

You can't look at one ago, I last summer in Iran, but it fires both when an action is performed and when it's perceived they're proven to exist in monkeys but not in humans, because not many humans are willing to undergo the invasive brain surgery to prove their existence.

A

So instead we talk about the mirror system in humans, and that gives her a something called the mirror hypothesis, which is that when you see somebody else's facial expressions, your mirror system fires, these neurons trigger sending signals to your facial muscles, causing you to mimic their expression in a microexpression, then through a phenomenon called embodied cognition. Your physiological state is assimilated through the front to your brain, the orbitofrontal cortex there's a fascinating case.

A

By about an engineer, a US engineer called Tammy Myers who damaged that part of her brain and she's, unable now to make what we would call gut. Decisions like buying yogurt is a baffling ordeal for her. Like does she pick the yogurt with the smiling lady on front on one with the classy branding? It's all like a semantic decision-making exercise for her.

A

So you see somebody else's face. Mirror system fires, mimic their expression embodied cognition, gives you a physiological state into your brain help. Your decision-making process give you that sense of guts intuition and there's some really quite cool evidence for this.

A

In another study, a control group of subjects were asked to correctly identify the emotions being expressed in photographs of people's faces. That was the control group. Then repeated this time. The subjects were Botox users. Now, given we're not in LA I'll, explain that Botox it's a lethal neurotoxin, which in teeny tiny amounts, is just enough to paralyze muscles stopping them from firing. So you don't get wrinkles.

A

On average Botox users were less able to correctly identify emotions being expressed by other humans.

A

I'm not suggesting that Botox users are emotionally dead, narcissistic narcissistic, sociopaths I'm, not saying that, and that's not what the research suggested either. Instead, it suggested that that chain was broken, like Botox users know just as well that smiling humans are happy right, but mirror system was firing, sending signals those facial muscles, which then weren't responding depriving their brains of this physiological information to help tip the decision-making process.

A

So, let's add that up? If we can put people face to face, then we can exploit these abilities of the human brain and make them more able to empathize with each other. Empathy is the lubricant of sociability. If you understand somebody's intent, it's much easier to get along with them and sociability is positively correlated with productivity.

A

So think about every time that you send a Skype for business, instant message to one of your colleagues or send an email instead of walking over to their desk or you have a voice call instead of a video call. We all know narrowband communication suck right. Otherwise we wouldn't have emojis I, have a nine-year-old daughter and I would much rather live in a world without emojis, but that ship has already sailed. Unfortunately, so just think about the things that you're giving up by using narrow bands comms we co-locate people. We can exploit this.

A

So let's go back to our cross-functional product team, yeah they're, all in the same place, they're communicating synchronously face to face. They've got a shared identity. They can see each other's facial expressions. They can understand each other's intends better, but in the old world it didn't used to be like that communication was asynchronous, it was slow. If you were particularly unlucky, it was via service tickets and service queues. That meant that there was a high transaction cost to getting anything done.

A

If you wanted two weeks of the test, if new team's time, but it might take you six weeks to get that time, so people ended up bunching up features together instead of working on one feature at a time. They'd work on aspects of a feature. So this is our project plan. Then it looks a bit like this, where we're asking different functions of the business, to do three things at once.

A

Rather than doing one thing at a time because of these high transaction costs in projects like this stuff off to the bottom right tends to fail. Right, load testing normally comes at the end. It normally goes badly, but why we're doing waterfall? We've got an exhaustive requirements document that explicitly states in the 31st chapter that we should be able to sustain a thousand requests. A second did you not read the whole requirements document?

A

Why is it the stuff? The end tends to go wrong. It could be because of the present bias. The present bias is a well observed psychological phenomenon, first recorded in 1950s. That shows that humans value things more highly when they're close to them, both in terms of space and in time. This is why buy now pay later. Sofas are still a thing, because that value is discounted because it's further away in time, but it's well observed right. So this isn't really.

A

You know it's 2018 right, nearly 2019 go and why am I telling you about things from the 50s? More recently, it's been shown that the present bias may have a neural or underpinning some lucky rhesus monkeys in an experiment, we're strapped down and given a choice between a little drop of sugar, water now or a slightly bigger drop of sugar water in a few seconds time and scientists were able to map out the preferences of these monkeys.

A

So if you are a rhesus monkey, how much more sugar water do you need to make it worth waiting another 10 seconds for it, so these preferences were tested and plotted out the lucky lucky rhesus monkeys then had electrodes inserted into the intraparietal area of their brains and the firing rates of specific neurons were observed.

A

These neurons seem to be encoding the value of the choices they're, making the faster. They were popping the more valuable the choice, the slower they were firing, the less valuable with the choice when the two were plotted together, there was in near one-to-one mapping between the two of them. Findings like this in neuroscience are extremely rare. The brain is made of countless complex adaptive networks. It's heightened tension working against each other. This was something remarkably straightforward in this graph.

A

The black lines are the neuronal firing rates and the red lines red line is the observed preference. So if the present bias is biologically underpinned, that means we can't get away from it. It's unavoidable. We can be aware of it and we can rationalize about it, but that takes self-control which, as we'll find in a minute, is a limited resource. So if we can't get away from the present bias, how can we exploit it? Instead? Can we turn it to our advantage?

A

Let's go back to our project plan. What hang on a minute like we're at cube convent where we're not doing all this stuff manually, that would that would be rather daft. We've got a platform to deal with all this stuff, so all these things go off the critical path, and we already said we're: gonna have a cross-functional team who I would really hope in 2018 are doing TDD rather than having a separate testing phase, at which point it doesn't make sense to be doing three things at once anymore.

A

If this team has self-service, they can do everything they need to do when they need to do it without having to wait and pick up another task, while they're waiting for something else to be done, and that makes it economical to have a feature based workflow, where you're continuously delivering one thing at a time, and that means that all the things are important to that feature are important now, because we're going to do it today, not in three months time by another team that I've never met we're going to do it today.

A

That allows us to exploit the present bias, but you can only do that if your teams have self-service if you're wrapping up your cube in service queues and requests for access and their DevOps team, if you've got a DevOps team you're doing it wrong by the way, if you're, making people request things and waits for them, you don't get this benefit. You can't do this you're still having those high transaction costs. You give people self-service who service brokers in the Service Catalog like hey, I'll, spin up a Redis.

A

You can exploit the present bias rather than falling victim to it.

A

Now the automation that a platform gives you can help with other things like technical debt in 2013, send or Mullah Nathan and elder Shaffir started looking into the psychological effects of knowing that you don't have enough of something. That's important to you in your life in a exemplary pre-civil piece of research, they asked shoppers in a New, Jersey mall, a hypothetical question. You've been involved in a car accident. Your insurer is not going to pay for the cost of repair to get your car back on the road.

A

How are you going to find the money to get your car back on the road and respondents were split into two categories: corner income, either rich or poor? Now each respondent was only asked the question once, but they were given different versions of it: either they had to find three hundred dollars to repair their car or three thousand dollars. So they're asked the question that you know Mull it over. What kind of sacrifices would you make?

A

Where would you find the money all that kind of stuff, and then afterwards they were given a fluid intelligence test. This is Ravens. Progressive, matrices and I would hope. With a conference full of engineers, you can probably guess what the next one in the pattern is, and it's used as a proxy for IQ.

A

So when the rich people were given this IQ tests, it didn't matter whether they were asked to cover three hundred or three thousand dollars. Their scores were about where you would expect them to be controlling for demographics.

A

When the poor people were asked to find $300, their scores were about where you would expect them to be, but when the poor people were asked to find three thousand dollars in this hypothetical exercise, their effective IQ was thirteen fourteen points lower than you would expect, just by being reminded by hypothetical exercise, though it didn't have enough of something that was important to them in their real lives.

A

Fascinated by this, the researchers went on to look for other effects of scarcity. One of the tools they used was one of these a Stroop test. You've probably seen these before you have to say the word rather than the color. It's printed in this exercises executive control, so you have to quieten down responses from different parts of your brain and go now I'm, going to listen to that bit time and time again, different demographics, different plants of the world different circumstances.

A

The researchers found that scarcity reduces both people's self-control and their ability to remember stuff in the future. Like most people think that humans are really bad at memorizing stuff, like we've, got really good memories, it's remembering to do stuff in the future. Your prospective memory, that's hard, like I, need to leave the car keys on the side, so my girlfriend can use the car later on. That kind of stuff is what lets us down and that's the kind of thing that scarcity can impact, but scarcity. The scarcity effect isn't triggered just by money.

A

It can be triggered by more abstract things as well. Like social contact turns out lonely people were demonstrating these effects, but importantly for engineers the scarcity effect can also be triggered by time.

A

Now you have think about what happens every time you go to a group of Engineers, they say hey end of the Sprint or Friday. Are you you made commitments to get these stories? Some are they? Are they gonna be done on Friday, because otherwise, we're gonna have to like replant this sprint and it's gonna go into the next. Remove sprints and the PM is not gonna, be happy.

A

Scrum masters not gonna, be pleased every time, you're reminding people of deadlines, you're, reminding them of the scarcity they're under you're, making them less able to solve abstract problems. Whit's last time, I checked was what we hire engineers to do, but, worse than that, we're reducing their willpower, their self-control, so they're gonna be more likely to cut corners to meet your deadlines and since of it, all off they're gonna be less able to remember to patch the corn as they cut to meet your deadline, and why do we end up doing this?

A

Thank you if you're doing waterfall, maybe this happens once every three months? It's no biggie, like the cost of it, is they're not that bad. If you're practicing scrum and people are fretting over the commitments they've made, then you're doing it's every two weeks, a perpetual state of scarcity for your engineers.

A

Why do we have to do that commitments? The business doesn't trust us, yes, boss, I, won't get the world gum boss now I promise. It will definitely be done by Friday, because I can predict the future yeah right, there's a better way. Continuous delivery can make deadlines meaningless.

A

If we continuously deliver the most important thing, then we're always working on the businesses number-one priority. So let's go back to our project planning. No one's gonna win a prince to award for this. Are they like one bar one bar in Microsoft Project? But if we rotate that 90 degrees through the power of keynote animations, we get something that looks like an ordered backlog prioritized by business stakeholders.

A

Engineers only work on the thing at the top of this list. They don't come in and go our world Rachael's better. At that I'll wait till she's in or I don't want to do. I'm gonna skip to the task, I think fun. They always work on the feature at the top, and that means there is trust between the business and the engineering team. They're, always working on number-one priority and never goofing off.

A

If we continuously deliver the most important thing in the simplest way, then we're never over engineering. So if we have requirements come in that we need to be the business needs to be able to cut something. We're not gonna build this. We're gonna build this, we're taking the shortest route to the number-one priority, we're trimming all the fat off the task. If we're doing things in the simplest way, then the work will get done when the work gets done.

A

Sticking a date on a calendar then becomes a bit like shouting at the earth to make it go around the Sun faster like do. We have any astrophysicists in the room or just physicists like the equations for velocity, do not depend on dates on a calendar thanks its kinetic energy and mass. You can add more energy, more resource, your team up to a point or you could reduce the mass, take some complexity out of the task, but those are the only two things you can do.

A

It will get there when it gets there and anything else we do. We are deluding ourselves into a false sense of certainty by making commitments trying to predict the future. Like I've, been in the IT industry 14 years, I can't predict the future I, don't know exactly how long a task is going to take. If I could predict the future, I'll be out buying lottery tickets and not practicing scrum.

A

Continuously, delivering the most important thing in the simplest way, frees us from that scarcity effect, and you can only do that when you've got the automation, you need to get stuff into production, because if you want to know how fast the team goes, you go look in production, you don't project forward for things. We can't know because we're not clairvoyance, you go look in production and what we delivered in the last two weeks, the features that are in the hands of users, generating data and revenue.

A

That is how you find out how fast a team goes, and you can only do that if you've, automated all the way through to production. Now continuous delivery can help us with technical debt. In another way, you've probably heard about a series of studies that was done in the 1970s. Looking at the outcomes of us parole, board hearings, and it was found that if you had your parole hearing at the end of the day, you stood a 20% chance of success, a 20% chance of being granted parole.

A

But if you had your hearing just after lunch, then you've stood a 65% chance of being granted parole and for a long time, people thought that this had to do with the parole board. Having eaten lunch, they had ingested some food digested, some glucose metabolized it into lysine, which is used in the pre, medial, frontal cortex, which is the area of the brain responsible for executive control that we talked about earlier.

A

But in a finding that's been repeated a few times this decade. It looks like there's something more to it.

A

Subjects were given an exhaustive Stroop test, so they're asked to do this stuff until they're repeatedly failing it, then they were asked to gargle one of two drinks: either a sugary drink or an artificially sweetened drink same but in the mouth spit it out and then immediately afterwards they have to squeeze a handgrip exercisers heart and for as long as they can to demonstrate their willpower.

A

So who did better the people who had the sugary drink? Well, they artificially sweetened one.

A

It was the people that had the sugary drink, but why right they're gargling it they're, spitting it back out, even if they cheated like there would have been no time for them to digest that sugar metabolize it into the right chemical and get it to the right path. Their brain couldn't have had an effect. Instead, the conclusion is, but it's that the sugar was binding through chemical receptors on the tongue, activating the brain's reward pathways, and it's this sense of reward that replenishes self-control.

A

Now you think about that anecdotally. It makes perfect sense right. That's why bad days tends to stay bad if you're, on a diet where you're trying not to drink or try, not smoke, and you let yourself down, you, don't feel very rewarded and then your will. Power goes out sort. It I'll have three pints, then we're as good days. You feel rewarded it's easier to resist the temptation.

A

So we tie this back to continuous delivery, then think about us working on one feature at a time we're getting things out of the door and importantly into production, where we can see that we've done a good job on a daily basis, not being pushed out three months later by a team that we don't know so, hey Rachel, you know a thing we worked on. Apparently it's live and what thing I, don't even remember sorry: I've got a deadline on Friday, Comic, Con told you.

A

This kind of stuff improves people's sense of reward, helping them resist. Technical debt improves their self-control. It's also an intrinsic reward, rather than extrinsic reward. You've people, bonuses, financial incentives and all goes horribly wrong. There are loads of studies to show that I'll tell you about in the bar later, if you're interested. So what let's tie us back together? What's this got to do with platforms? If you've got manual, processes and you've got to hand things off between teams, then you've got those that siloed situation that we had at the beginning.

A

Having a platform means that we can have cross-functional products teams with a shared identity who will care about each other's problems because they're on a mission together, we can then co-locate them and make them better at empathizing, with each other, make them more sociable and more productive. If we give that team self service- and they don't have to wait on other teams to get the things they need to do done, then it becomes economical to work on one feature at a time that which means that everything that's important.

A

That feature is important now, allowing us to exploit the present bias, which we call escape from continuously delivering. The most important thing in the simplest way allows us to escape that scarcity effect. The work will get done when the work gets done continuously delivering and seeing that we've done a good job and that features are getting into production straightaway will increase people's sense of reward, improving their self-control, making less likely to incur technical debt.

A

So why did we see a 20 year lag in the productivity gains from the advent of electric motors and electric generators, because there was loads of Krofft that needed to change? There was an old way of doing things. The new technology see where I'm going with new technology right that should be changing the world. If you wrap it up in the dressings of the old way of doing stuff, it has no effect.

A

It doesn't matter how shiny or yeah meal is and all the operators we install and all the lovely things or look at the CRTs or they're, so lovely. Look at Miami. Look at my gear tops if we're still asking people to do things in manual ways. According to the old playbook, we're not going to see the productivity gains, we're just going to take VMs and replace them with containers. Great yay, IT industry, if we're not making people more productive than this is all for nothing. So we're not constrained by steam engines.

A

I've worked with some enterprise databases that do make me wonder whether actually we are sometimes but but we're, also not in a world of certainty. We're not in an era of mass production where you can do identical things as cheaply as possible, and that will lead to business success when I started engineer better with another chap called Dan just confused everybody. We spent some time trying to work out what we have in common.

A

What we believed- and this was one of the observations that fundamentally businesses boil down to these two things: the hard-earned money of shareholders and the easily spent time for employees for easily spent.

A

You can read that as easily wasted and it's easy to be glib about the hard-earned money of shareholders when we live in increasingly unequal times, but Sam Ramsey, who was at Google cloud for a while, before moving on to Autodesk, once told being a bit of speaker training before a conference that time isn't measured in seconds time is measured in heartbeats you're, all spending heartbeats now listening to me, spending heartbeats with maybe a Ravi drinking beer in the other building, but you're not going to get those back.

A

I really hope you don't ask for a refund, because I can't give that to you all of the old-fashioned processes that we have in our enterprises are doing the same thing. They are wasting people's heartbeats, but we also know that time is money.

A

Therefore, money is equal to heartbeats. So of those two things that businesses are made of people's time and shareholders money, it makes no difference which one we're wasting we're wasting people's time on the planet, that they're not going to get back your massive enterprise that maybe doesn't care about people and has its archaic ways of doing things. I talked to a few people from banks earlier who had some particularly interesting processes. It's easy to like forget about that wastage of money, but at some point somebody went out to work.

A

They left their family to go to work to spend.

A

However, many hours earning a wage to save up, to pass down to their children, to invest in a pension that bought the stocks and shares and the enterprises doing old-fashioned things and wasting that time and money, and we know better so I leave you with this proposition that, if you're building a platform make sure you don't dress it up in the processes of yesteryear, make sure that you're giving people self-service that you're making your business is more productive, because, if you're not what are we doing other than navel-gazing about our lovely Amal.

A

All of the research that I cited is up on a very dry blog post, engineer, brain all of the journals and things are on there, including some other material. That's not in this talk, there's other versions on YouTube and if you want to troll your favorite scrum master, there's a great bit of LinkedIn, clickbait, I, wrote, called scrum makes you dumb, which just talks about the scarcity effect with that I think I'm over time and I did promise not to hold you from the beer for too long.

A

If you want to ask any questions, grab me afterwards, but otherwise thank you very much and have a great cube gun.