I’m not here to talk to you about how amazing A.I. is, what Deepmind is working on, or speculate about robotic overlords. I do do that, sometimes. Today, I want to focus on the most simple and boring type of A.I. that is Machine Learning without Neural Networks.

Why? Because it will change the way software is created forever.

Wait, isn’t all A.I. just Neural Networks?

Okay, let’s get a couple things out of the way definitions wise. While it may seem that Neural Networks, Deep Learning, Machine Learning and Artificial Intelligence are all the same things, they all have their own history and origin, as well as hierarchy. The reason you might be hard-pressed to see that distinction is because of all the research and media attention around the last decade of advances specifically in Deep Learning.

I’ve done two courses on Artificial Intelligence, one with M.I.T. and the other with Toronto University and Geoff Hinton. Geoff Hinton goes pretty much straight into Neural Networks and then into Deep Learning, as do many other courses on these topics. Luckily for me though, I had done the M.I.T. course which had one out of 20 lectures on Neural Networks, the rest covering all other aspects and history of Artificial Intelligence. So let’s break it down some.

A.I. = M.L.

The good thing here is that most of the terminology actually has logic to it. To put it simply, Artificial Intelligence is any system that can make its own decisions. For all intents and purposes, given the research and advances of the last three decades, you can safely interchange these two terms. You’re pretty much only ruling out rule-based “expert systems” that airlines used in the 80’s. Other than that, everything interesting in A.I. relates to Machine Learning.

Machine Learning covers a lot

Luckily again, Machine Learning is self-explanatory. Instead of you telling the machine what decisions and rules to make, you teach it. A machine that learns. So that leaves the methods of teaching and learning pretty wide open. So what can you teach a machine, and what can it learn?

This is the current landscape. It all sounds very fancy and complicated, and it actually is. To simplify, here is what you can do with the main methods:

  • Classification algorithms can be taught to split existing data into classes, like say names of animals. Then when you give it new data, it will tell you which class it belongs to, like say this is a chicken and not a dog.
  • Regression algorithms basically try to learn the function of a dataset, by predicting future data based on past data. Exactly like the “regression line” you had in Excel, but multipurpose.
  • Unsupervised Learning can be used if you’ve got lots of data and you can’t make sense of it, so you teach the machine to try and make sense of it instead.
  • Reinforcement Learning is how to beat every human on Earth in games like GO and Chess, or drive autonomous cars and drones. If you’re not doing those things, you don’t need to know about it yet.

While the last two get a lot of the media attention, the first two are the moneymakers today. So we’re going to focus on them. Regression is trying to understand how the dots in your plot relate to each other. Classification is the opposite, and tries to separate the dots in your plot into groups. There are a lot of ways to do each of those things, and Neural Networks is just one of them. So let’s get it out of the way, before we get into the practical stuff.

Neural Networks are a special flavor of ML

Neural Networks and the associated learning algorithms hold a special place, because they’re inspired by the brain. We know that neurons are connected in vast networks inside our brain, and that electrical signals go from neuron to neuron to produce all of our conscious experience. Seeing. Hearing. Thinking. Speaking. All neural networks in action.

What’s inside? Well, a bunch of neurons, organized into inputs, “hidden” layers, and outputs. Really the function of the layers is to introduce additional complexity. More layers bring more complexity. Otherwise, you could only do really simple things like add numbers together. But when you make all those spiderweb connections across hundreds or even thousands of neurons in several layers, it turns into magic.

Am I kidding about magic? Yes and no. It’s magical in how powerful such a seemingly simple thing is. It can learn almost anything with a learning process called “backpropagation”, which starts by comparing how far the prediction is from the intended outcome. Then it makes a series of minute but carefully calculated changes across that whole network, and tries again, to see if it got better or worse. The real explanation goes beyond high-school math pretty quick, and involves working out the partial derivatives from the output all the way back to the input.

What magic can it do? It can read handwriting. It can recognize objects in pictures. Play chess, even. All at human level, or beyond. The magic also means we’re not 100% sure what’s going on in there. It’s so complex. Change just one value on one of those connected lines, and the whole output can change. Cat becomes dog. Why does it work? When does it work? How do we find the best and fastest way to train the network? Work in progress, let’s say.

Can’t we just always use magic?

In theory, yes. In practice. Not so much. Let us introduce our other contestants to demonstrate why.

On the left, you see three datasets with a white background. Going from left to right, each column represents a type of Machine Learning algorithm trying to separate the blue dots from the red dots. This is called Classification. Remember, we’ve told each algorithm already which color each dot is. That’s called training data. It’s just trying to create a rule for which area blue dots go in, and which area red dots go in. As you can see, results may vary!

Something you may notice is that the Neural Net, the fourth from the right, is doing something funny. For each dataset, it’s doing something totally different. How does that happen?

To really make this point hit home, the above picture is just one Neural Network with three different datasets. This time, the columns represent changing one setting, called “hyperparameters”, of the network. Even then, you get wildly different outcomes.

Neural Networks are unpredictable by nature. It’s why they’re so powerful. So the tradeoff is big. So why can’t you just fiddle around a bunch to make it work?

Reasons you shouldn’t use Neural Networks every time:

  1. They’re complex, and making informed decisions for their design requires serious math skills most people don’t yet have.
  2. They’re unpredictable, so you have to fiddle around to make it work at all, even if you know what you’re doing.
  3. It’s hard to say if you’ve done the right thing, unless you try a lot of different things.
  4. Even if there are many ways to measure how good your network is, it can be difficult to understand how to fix any problems.
  5. Making up your mind about the above can take a lot of tries, and each try can take a lot of time and money. Think hours or days of waiting for each batch of training to be completed.

The less popular sidekicks to Neural Networks

As you saw earlier, there are many alternatives. I’ll focus on the two which give you simple and predictable outcomes with two very different approaches. Why? Because most often, one of these will quickly solve your problem. Both can be used for Regression and Classification, depending on your problem. Again, I’ll choose to focus on Classification for reasons I’ll explain later.

Anecdotal evidence from observing winning entries at data science competitions (like Kaggle) suggests that structured data is best analyzed by tools like XGBoost and Random Forests. Use of Deep Learning in winning entries is limited to analysis of images or text. — J.P. Morgan Global Quantitative & Derivatives Strategy

The difference between Neural Networks, and all other Machine Learning methods is how they learn. As we saw earlier, Neural Networks kind of guess their way to the best solution. Kind of. The other methods actually calculate the best solution. They consider the data you give them, and use a large variety of mathematical optimization methods to simply find a best answer. Another benefit? These methods are fast to train, and fast to execute. Minutes of training rather than days. So no need for cloud computing or special hardware. So let’s look at them.

Linear is straightforward

The most logical and simple way to try to separate a dataset is to draw a straight line through it with a ruler. That’s what a human would do. That’s also what Support Vector Machines (“SVM”) do, despite the kind of awesome and complicated sounded name. The algorithm tries to find the best single straight line to separate your datasets, and then sets a buffer around that line to separate the datasets as far as possible.

You might be interested to find out that the original SVM algorithm was invented already in 1963, decades before A.I. was cool. Many variations incl. non-linear solvers that can draw polynomial separation lines or even radial areas, i.e. not straight lines, but we really want to keep it simple and understandable. So linear it is for now.

Trees are your friends

Decision Trees choose which variables and values most predict the outcome based on your dataset. Slice and dice. It tries to “cut” your data points by separating variables at certain ranges within their values. Once it makes a cut, it moves to the remaining available variables and tries to do the same, while trying to do as few cuts as possible to keep things simple.

The result is like fitting rectangular Tetris blocks on your data. This sounds like a bad idea, but because of this crude approach the tree has a huge party trick that sets it apart in all of Machine Learning.

Decision trees can explain themselves. Yes, you read that right. All those media articles about how Neural Networks are doing unpredictable and even things? Not a problem here.

Even better than that, there is a free tool called graphviz that generates a visual representation of the resulting algorithm. You can actually check the logic, and be 100% sure you know what it does and when. Get a weird result? Look it up, and you’ll see exactly why.

A useful variation of the Decision Tree is a Random Forest, which runs a bunch of individual tree solutions on subsets of your input data, and gives you an average. Compare them side-by-side in the big Classifier comparison chart above, and you see the idea. There’s also a whole group of super-efficient boosted tree algorithms, if you had to get real fancy. But, you probably don’t.

How software is currently created

So, back to the big picture. We now have some cool new tools to play around with. So what? I’m just creating an app or website. This doesn’t apply to me. I’m not trying to beat chess masters here.

Wrong. First, let’s establish how most software is created today. Software is rules-based. Meaning you define a set of rules on how things work, and then the software just does the same exact thing over and over and over.

This is a typical structure commonly used in modern software. You have three kinds of code. One that shows things (view), one that defines things (model), and one that decides what happens between the two (controller). In this kind of structure, there are two ways explicit rules are imposed: the model itself, and the “business logic” of the controller. Business logic is a fancy word for “if this happens then do that”.

So what goes in the model box? A fixed model with fixed relationships. This is why software is slow and hard to create, because you have to map it all out. The further you get, the harder it is to change anything. Innovation slows down over the iterations and versions, as the degrees of freedom are reduced to zero.

How (simple) Machine Learning can help you create better software

The terms A.I. and M.L. have become so overused that most people now scoff at anyone who says they use either. I used to be that guy. But having been a practitioner for a while, I’m starting to see the light. There is, in fact, a legitimate way to sprinkle a little A.I. into any software.

Teach logic to your software

What if rather than have to decide on how everything has to work at the beginning, you could just teach your software what to do? That way, if you had to change it later, you could just teach it again? While you may be picturing Tony Stark and Jarvis, you can do it too, today.

This is where we get back to Classification, specifically. What is logic? What is decision making? It’s connecting a number of inputs to a number of outputs. Also called Multiclass Classification. What’s a great algorithm for this purpose? Something that allows you to train on data rather than define the code, but is simple and explainable? Decision Tree. How does it work?

To train any classifier with scikit-learn, you need two lines of code. Yes, two.

classifier = sklearn.tree.DecisionTreeClassifier()

classifier = clf.fit(inputs, outputs)

The best part is that it can replace complex logic and modeling work with one line of code. Yes, you read that right. Once you train a model, this is how it works:

output = model.predict(inputs)

Alternatively, you might want to get a probability distribution across all possible outputs for a set of inputs. That is much harder, as you can see.

outputs = model.predict_proba(inputs)

I mean, isn’t that just beautiful? If you have new data, or need to replace the model, you have to change one file: the model itself. Job done. No database migrations. No automated integration test suites. Drag & drop.

How do I get data tho?

So what kind of data can you pump into one of these classifiers? Here’s one simplified example. Let’s imagine your app is recommending what pet a user should buy based on their preferences. You might ask about characteristics that users would want in a pet, and train a model to produce a recommendation. The output will depend on how much data you have, and how specific you want the recommendations to be. Rather than a database model, which has to return an exact matching dataset using complicated join statements, you could return the top 3 most probable choices in one line of code.

Toy example: few simple inputs, few hundred datapoints

In most cases though, your data won’t be that simple, and the inputs won’t be unified as a mere yes/no which can be turned into 1 and 0. So you may need to adapt your data to be something the classifier can learn. You could do some operations manually to turn the strings into numeric classes, or run automated algorithms to encode your data, such as a One Hot Encoder. Since the training is trying to establish relationships in your data, making the numbers easier to relate will help get a better result, as long as you can still interpret that result!

Simple example: many inputs of various formats, thousands of datapoints

So you may have a question of how to generate such training data. I mean, who is qualified to say what is the right behavior? What if you have inputs but no output labels? This of course depends entirely on the problem you’re solving, but answers could range from creating and labeling your own data, to finding existing (open) research data, or even scraping existing databases or websites like Wikipedia.

An interesting opportunity this approach creates is that of expert opinion. What if you crowdsourced the training data from a panel of experts in that specific field? Maybe doctors, zoologists, engineers, or even lawyers. Well, maybe not lawyers.

A.I. is becoming mobile friendly

Traditionally, one of the challenges in adopting A.I. was that you needed to run these models in the backend. So first of all, you needed an actual backend server, which often meant learning a different programming language, and the hassles and costs of hosting and so forth. Secondly, it meant those models could only be run when connected to the server. So if it was a core feature of your app for example, it would only work online. Boo.

Apple has been first to tackle the offline issue by introducing the CoreML SDK as part of iOS11. It works like a charm. All you need to do is convert your existing model into CoreML format, and you can literally drag & drop it into your XCode project. From there, the model will generate a class API for you that you can call as follows:

guard let marsHabitatPricerOutput = try? model.prediction(solarPanels: solarPanels, greenhouses: greenhouses, size: size) else {
 fatalError("Unexpected runtime error.")

The future here is that several companies including Apple are rumored to be working on dedicated A.I. chips for their next generation devices. That would enable fast execution of complex neural nets on your own device.

How to get started

  1. Scikit-learn tutorials are a great place to start. It’s all in Python, which is the easiest language to pick up, so don’t be intimidated. Start with this one, and get through it line-by-line. There are few things to get your head around in terms of preparing data, so just do it.
  2. How to run different classifiers and visualize the results in 2D.
  3. How to add Machine Learning to your iOS app using CoreML SDK.
  4. If you want to start with a book, this hands-on guide by Aurelien Geron on scikit-learn and Tensorflow is recommended.

Try it, it’s really not that hard if you know how to code at all!

Source: LinkedIn

A now widely circulated and famous article by Inc. magazine titled The Psychological Price of Entrepreneurship unveils the truth behind the typically idealized, romanticized picture of the heroic startup founder. VC’s. Unicorns. IPO’s. Admiration of the masses and glossy magazine covers. It could be me.

Or… Losing your life savings. Losing your friends. Even family. Personal debt and bankruptcy. Sacrificing your health. Burnout. Mental illness. Yes, even suicide. The forgotten founders you don’t read about. Those who fell on the startup sword.

Here’s to all you failed founders, who reached for the stars, but didn’t quite make it.

It’s true. No doubt about it. Creating something from nothing often entails more than a fair share of risk, stress and sleepless nights. Obstacles. Failure. Rejection. Rinse & Repeat. There may be entrepreneurs out there who had calm winds and smooth sailing from Day 1, but that’s not the experience you should prepare for. Somebody also won the lottery, but that shouldn’t be your plan for life.

Most people’s first startup experience will be like sailing across a stormy ocean alone, with no experience in yachting, no life vest, and water leaking in.

But that gloomy reality is only your’s, and maybe your spouse’s. You can’t talk about it. If you do, you’re manifesting failure. Who in their right mind would complain to their employees, customers, board, advisors, or investors that it’s actually really REALLY hard? You’re trying to convince everybody that it’s going to be amazing. That things are going better than expected, and massive success is just around the corner. Get on board before we’re famous! Yet, one false step and the whole house of cards collapses.

Fake it till you make it is a core belief in any person trying to hack together a business from sheer will to succeed.

While there is certainly great value in examining both ends of the spectrum, I want to focus on a golden middle that is actionable on a daily basis. Glory or gloom may be around the corner, but what you do today, tomorrow, and each day counts more. If you can live with yourself through the ups and downs, you can survive to fight another day.

Depending on your experience and chosen industry, your success factors around product/market fit, financing, and scalability will vary, so I won’t go there. That’s what accelerators are for. Instead, I’ll talk about ways in which you can battle the day-to-day struggles that all founders face: rejection, stress, and loneliness.

These methods aren’t mine, but I’ve been using them the past two years, so far pretty effectively, in juggling two startups in Bambu and Missionready.

#1: Be in it for the right reasons

Statistics would point that startups are the worst get-rich-quick scheme ever invented. The media bias towards success stories makes it seem the opposite, unfortunately. If you’re in it for the dough, the rough patches will eat you up. Every obstacle will make you doubt whether it will ever generate that sweet cash you so desire. Then again, if you have a real passion for what you’re doing and/or what you’re trying to achieve, it will give you a sense of purpose. Trust me, that’ll come in real handy.

#2: Blasting through daily rejection

Something you can read through the lines of startup blogs and founder interviews is the acceptance of rejection and failure. There is a selection of inspirational posters available to remind you of the gift that is daily rejection. Nothing beats some rejection with a good cup of coffee to kickstart your mornings!

It’s mostly a facade though. Humans hate rejection by design. Evolution has given us these massive brains to play the social game, and win. Rejection in evolutionary terms means dying alone in a dark cave, talking to your stick friends. It isn’t healthy.

So be grateful.

Not for the rejection, but for everything else. Things that make you happy in your life. Things that are going well. The rocks that keep you rooted in place, big or small. Some days it’ll be the little things, like a memorable conversation or something that made you smile. A random compliment you received. Some days it can be the appreciation of good relationships in your life, the beauty of your favorite season outside, or the good health of your children.

Putting your mind in a place of gratitude when you open your eyes, and when you go to bed has been scientifically proven to change your epigenetics. Meaning the structure of your brains and DNA change if you think good thoughts. Think about it. If sleep is a reboot and regeneration mechanism of the brain, do you want to enter that mode for hours each day with a brain filled with thoughts of revenge, despair, and bitterness? Similarly, as you wake up, is it a good idea to base your day’s thought processes around potential ways to fail today, that nasty email you got last night, or perhaps take a moment to appreciate things that will give you the mental capacity to blast through the insignificant bumps that come along each day?

Another great way to get similar perspective is through social support. People who value you for you, not for your achievements or status. It can be friends or family, but not acquaintances or random people at networking events. Importantly though, do not use these people as garbage dumps for your troubles. You should look to them for comfort and perspective, to get away from your tiny bubble of trifling troubles. Don’t tell your kids about your latest series of rejections from investors, ask them about school and the latest happenings at the neighborhood sandbox. Detach from your reality just for a little while, and immerse in theirs.

#3: Putting a cork on stress

Stress and startups go together like two peas in a pod. Can’t have one without the other. No wait, of course, you could just have less stress if you didn’t have the startup at all. That works. Other way? Not so much. Just a matter of how much, how often, and how you cope with it.

Like Drake, stress can go from 0 to 100 real quick. Once it’s up there, you have to double your efforts just to maintain your levels and not go ballistic. Smaller and smaller things start getting under your skin. Even the good things in your life can start taking a different, negative hue.

Here are a few ways I’ve found I can limit the downside of stress before it becomes a rampant and self-feeding spiral of destruction. Pills and potions not needed.

Daily routines: One seemingly arbitrary place to start, is to simply control the amount of change in your life. Your startup will provide ample amounts of excitement and change, so stabilize everything else. Eat at the same place. Take the same bus. Maintain the same hobbies. Watch the same shows. Keep the same circle of friends. Don’t go looking for extracurricular adventures as sources of extra anxiety. Save all that emotional capacity for the office, and embrace your inner Ned Flanders when you leave.

Breathing: For the last year, I’ve maintained a daily practice of breathing. I started from the nice built-in Breathe app on the Apple Watch, which reminds you to take a few deep breaths every hour of the day. I noticed how dramatic the resetting effect could be, right in the middle of that frantic noon email rampage, or in the last minute taxi ride to make it to your next high stakes meeting. A similar experience for your phone is the Oak app.

I’ve since moved on to more powerful methods like Wim Hof, which opens a whole world of exploration within something as simple as taking a breath. The science really backs this one up, as breathing is the only function in your body that is controlled in parallel by both your voluntary and autonomic nervous systems. By controlling your breathing, you can gain some control over other autonomous functions like your heart, immune system, and even digestion. No, this doesn’t qualify you to wear a long leather jacket like Neo from The Matrix.

Mindfulness: This is currently way more on-trend than breathing, but at the same time it’s trickier to make it effective as it requires more focus and persistence. For high-pace guys in the startup biz, it can be difficult to sit down and chillax to some mellow ambient tunes without sketching out new battle plans to beat your competition to a pulp inside your head. Luckily there’s a slew of apps out there, like OakCalm, and Headspace to work your way up to zen-mode. Personally, I find a good 10min breathing session to give me the same effect, but with more action involved to keep my rushing brain in check.

Workouts: In many ways, working out can combine a lot of the benefits of the two categories above. Additionally, it can be used as the ultimate physical purge of all negative and destructive energy built up over another week of bruising startup life. Whether you hit the trails, track, treadmill, or actually punch a bag, it helps to go hard once or twice a week. You’ll be amazed at the sense of emotional rebirth 30mins of physical exertion and pain can bring. You could try Missionready for this.

#4: Co-founders are your co-mmiserators

Having a cofounder that you trust enough to share vulnerabilities with, is one of the best-kept secrets of startups. This is another person, perhaps the only person in the whole world, who not only cares as much as you about making it work, but actually lives that same reality as you do. The same ups that friends shrug off casually, before moving the conversation back to their favorite Netflix shows. The same downs that you keep from the family, so they wouldn’t worry needlessly.

Founders are humans too. Except Elon, I’m not sure about that guy.

As a personal story my cofounder at Bambu, Ned Phillips, has become a huge asset in my life. We laugh at our wins together, and often marvel at the random paths of our successes. We try to laugh at the failures too, while searching for meaning and silver linings. Most importantly, we try to enjoy the minute experiences that make up the journey. We have mutual interests outside of work to keep the relationship grounded, despite our 15 year age difference.

People say it gets lonely on the top. Well, it’s also pretty lonely on the bottom. The worries of the world can seem to pile up with no end in sight, and having someone there to share the burden can make all the difference. Co-founders can often bootstrap each other emotionally to remain afloat through the waves of a typical startup journey.

Hope that helps someone out there, and if it does, I would love to hear about it.

Are you a current or aspiring founder? What works for you?

Source: LinkedIn

Guys, it’s happening.

Did you miss the news? You probably did, because the announcement was tucked in between demos of the latest iMessage stickers. Millennials… sigh.

At the annual developer conference WWDC, Apple announced their next move into the world of Financial Services. If you thought Apple Pay was as far as they would go, you were mistaken. Sorely mistaken!

Enter the “Apple Pay Cash Card”

According to the limited information available, Apple will automatically issue a virtual payment card to all iOS users, allowing them to receive and hold money in Apple Wallet.

Pause. Let that statement sink in.

Read it again.

This isn’t your weird cousin’s Google Wallet

This isn’t Venmo. Or even Paypal. We aren’t linking up accounts and plastic cards here. We are issuing cards. Virtual cards. To all iOS users.

One billion iOS users. One billion new cards.

Apple cards, in Apple Wallet, with Apple Pay.

Apple is brutal in implementing its closed ecosystem strategy of always owning the full stack. It rarely leaves space for middlemen, and anyone left kicking pays big time to participate. Banks paid millions in “marketing fees” just to get on Apple Pay.

Want to receive money? Buy an iPhone.

This development conveniently closes the loop on Apple Pay. You can naturally use your Apple Wallet money to pay for anything, with a flick of the Apple Watch on your wrist. Same for online purchases with the recent introduction of Apple Pay for websites. All you need is Apple Dollars now. Don’t tempt them.

Does this make Apple the world’s biggest financial institution?

How much cash will Apple be holding in users wallets in one year? Will nine zeros be enough? Ten? Twelve? That’s a trillion. Only Apple makes numbers like that seem feasible.

Customer acquisition cost for one billion accounts? $0. How do you like them Apples?

How soon will they start to pay interest? What about something like Apple Wealth? The scale and impact really boggle the mind. We are breaking new ground here. Only WeChat and Alipay are operating on this scale, but let’s face it, they are still one trick ponies operating in a single market.

Who can compete with them at this point? Facebook has no hardware. Google is too open to achieve this level of integration. Maybe Amazon…?

Venmo, you’re out.

Paypal, watch out.

Banks, freak out.

Source: LinkedIn

My first startup Bambu has now made it well clear of the first 12 months, so we are now only at 90% risk of immediate and catastrophic failure. Phew! I can already sleep with the other eyelid closed.

Do I know what to do right? Am I suddenly declaring myself an expert? Most definitely not, but I have observed a lot of startups and heard their stories. I have second-guessed a lot of our own decisions. We’ve talked at length about potential outcomes. A lot of what we’ve done right was partly happenstance, but I can see how an alternative outcome could’ve slowed us down or even stopped the whole show.

So here’s my two cents. Take it or leave it, but comment below anyway.

#1: We’re focusing on R&D

If you haven’t launched anything during your first year, the chances are you’re doing something wrong. The market may have moved on already, while you were busy defining a legal framework that will allow you to draft initial requirements for a potential go-to-market strategy, to protect IP you haven’t yet built. Under strict NDA, of course. Being in the market is all that matters. Build on results, not spreadsheet projections of results.

Who is your customer? What problem are you solving for them? Why will they pay you? Why will they stick with you? Don’t write a single line of code, until you can summarize these in 30 seconds to a stranger.

Here’s a novel approach for startups. Put the above on a Powerpoint. Pitch it until you find a customer. Build a prototype. Pitch it until you find another customer. Build an MVP. Pitch and sell. Build the next version. Pitch and sell. Repeat until exit. At least then you can afford to pay your intellectual property lawyer now.

Mo revenue, less problems. – Notorious B.I.G.

There aren’t many things that more revenue won’t solve. Can’t afford to hire the right people? Revenue. Can’t raise money? Revenue. Can’t get partners? Revenue. Can’t get Twitter followers? Revenue. Can’t find your car keys in the morning? Revenue.

#2: We’re in Stealth Mode

Recently we polled a group of our B2B startup peers, and out of 20 companies zero had a marketing expert on the team. Meanwhile, many of our peers wonder how Bambu is seemingly everywhere at the same time. Online. Offline. Events. News. Hackathons. Blogs. Panels.

Our first employee was our head of marketing. Don’t underestimate the value of “top of mind”. It matters with customers, investors, partners and hires. Momentum builds, cross-pollinates, and creates new momentum babies. If you’re in “Stealth Mode” see point #1, I’m guessing it applies to your situation. Thank me later.

#3: Business plan paralysis

If you’re starting from zero, as most startups tend to be, you aren’t going to be able to forecast anything with any relevant amount of accuracy. So don’t bother. The time spent on projecting your IRR on Q4 of Year 3 could have been spent hustling.

Hustle > Planning

How does one hustle? What are the conditions for peak hustle? Bambu CEO Ned Phillips often tells the story of the dentist. The dentist is a specialist, that comes in once the receptionist has taken in the details of the customer, and the dental hygienist has set up the space chair and tools. Enter dentist. Comes in, does his thing, and out in 5. A whole orchestration happens before and after this most valuable piece. Be the dentist. Build a team around you to orchestrate, and allow you to maximize the volume of your high-value pitches. Hire that team early, not once you can afford it. This is not a chicken and egg situation.

NOTE: Networking does not qualify as hustling.

#4: #machinelearning

This was driving me nuts two years ago. Now it’s two levels above ridiculous. I’m pretty sure the word is ridonculous.

If you could design a vacuum cleaner with machine learning, you would get funded. I mean, say that it has machine learning. You don’t actually need to have it. Nor understand it. Nor intend to ever understand it.

#deeplearning > #machinelearning

It’s almost like saying cloud. By now everyone assumes you have it anyway, so saying you DON’T have it would be weird. No harm done, amirite?

Don’t build your house on a deck of cards. Build tech that addresses a use-case, and if that works, you can spend time sprinkling some magic dust on top. Do Steve Jobs magic, not actual illusions.

#5: Bagging

Everyone knows a few bags. Douche. Sleaze. Scum. Tea? There is a persistent myth around the founder who makes his own luck by elbowing their way into success. One cold-called 5,000 investors to get funded. Another stole the key employees from his competitor. Some guy didn’t pay freelancers on his first deal. Some lady faked data to impress investors. Lots of fist pumping action. Until they strike out big time, that is. Think Bernie Madoff. Does it have to be like that? Isn’t there an alternative?

Good guys win. Not as in when you arrive at the pearly gates, but as in next quarter. You may douche your way into your first deal, but that streak ain’t running long. That customer will know you pushed them, and will not renew. Investors will spread the word if you pull a fast one. If you remain humble and open, you will win the second, third, fourth, fifth and every deal after. Being a good guy never goes out of fashion. People do business with people. That goes for customers, investors, partners and hires. Being a good gal/guy matters.

How to do keep going

If you’re passionate about something you understand, you will find a way to win.

It really is that simple. You will find a way.

Source: LinkedIn

You know that annoying “Uber of X” thing that’s going around?

Well, guess who’s the Uber of Fintech? Indulge me.

Uber. That’s who.

Uber is the Uber of Fintech.

Feel free to quote me on that one.

Über Retirement > Regular Retirement

Some months ago the company announced an interesting partnership with the posterboy of online wealth, Betterment.

To get started, drivers using the Uber platform can sign up for traditional or Roth Individual Retirement Accounts (IRAs) directly through the Uber app. In addition to creating a retirement account, drivers also have the option to create taxable accounts for other investment goals, such as major purchases (e.g., vehicles of their own) and safety net funds. -Betterment

I mean, how is that not great for the drivers? C’mon now. All you need to do to start building your financial security is press a button inside the app that provides your livelihood already. Embedded, son. This is what people need, instead of some sleazy fund hustler with a cheap suit. They need douchey startup guys wearing hoodies. Way better.

Embedded Finance

Anyhoo… This is the future of Finance. Payments inside ChatSaving inside eCommerceBanking inside Telco. Instead of forcing tedious banking relationships on you, finance will become a natural part of your life, like it was always meant to.

What kind of difference will this make? Take a look at this here graphic, buddy. Look for the red line. Million’s the unit here, by the way. As in 100,000 million.

Coming up with these use-cases is the challenge. Millennials don’t care about your 0.1% interest rate, they’d rather get paid in data minutes. Or PokéCoins. Whichever requires the least amount of swiping.

Global != Local (nerd for not equal)

So Betterment got the sweetest Fintech deal out there. Hooray for them. What about the rest of us? Do we settle for Lyft? Well, there’s good news. You see, Uber also just announced a partnership with Moneyfarm in the UK. Whaat, you ask?

You can’t just take a U.S. Robo and drop it in rural Indonesia. It isn’t Bear Grylls.

Uber solved this problem twice, and will probably solve it again. The point is that Fintech doesn’t work well as a global platform. Particularly when talking about non-daily finance, say like Wealth or Insurance. Local context, social norms, technological adoption, and regulations make a huge difference. The concept is different. The distribution is different. The marketing is different. The UX is different. Also language.

Source: LinkedIn

Disastrous 2008, the year of financial crisis is also famous for the financial revolution which is sweeping the world today. It is the year that gave birth to the Robo-Advisors; the new breed of wealth managers that emerged to redefine the US$67 trillion wealth management industry. Backed by powerful algorithms they provide rapid automated investment solutions based on risk appetite.

The investment industry is flooded with news of top investment management firms shutting down loss-making businesses ever since 2008. Words like restructuring, relocation, lay-off, strategic reviews still continue to haunt the industry. The deteriorating capability of asset managers to outperform major indices has further fueled investors to shift towards passive investment strategies such as ETF. In fact, investors have pulled $340bn from actively managed funds in the US last year and poured in $505bn into passive equivalents.

Robo-Advisors currently manage around $50bn of AuM, and as per A.T. Kearney, a consulting firm, AuM is estimated to explode by 68% annually to about $2.2 trillion by 2020. Gone are the days when stalwarts used to debate whether Robo-Advisors are a threat to the traditional financial advisors. We are at the juncture where Banks, Asset Management Firms, Financial Advisors and Non-financial firms have started embracing the technology to provide efficient investment services.

Robo-Advisors have now successfully emerged as a silver lining to the stressed Asset Management Industry and for the financial advisors. Still there exists few critical questions – How Robos can outperform actively managed long-term investment funds? Will the development in neural research and A.I. enable the Robos to work in isolation? Or the bonding that now exists between a financial advisor and the Robo continue in future as well? In this case, it is not the time, but technological advancement in A.I. that will dictate the future of Asset Management Industry.

– Rohith Thatchan, Financial Analyst (HK) at Bambu

Recently The Economist published a special report called Lifelong learning is becoming an economic imperative, talking about how the workforce will need to adapt to increases in technology and automation across industries. Short version, resistance is futile. Long version, read on.

Yet as elegantly analyzed by Denis Sproten in his piece AI is rewriting John Keynes equation of employment, it really isn’t that simplistic. In the aftermath of the first industrial revolution, you saw manual laborers go from farms to factories. The cognitive requirements were higher, but most handled the transition in stride. Today, workers are transitioning from factories to offices, again increasing the cognitive requirements, but still most people cope. Can we level up forever, you ask..?

NopeNuh-uh. Technology is improving at an exponential rate, yet our genetics are improving through evolution at a linear rate, if at all. At some point, exponential leaves linear in the dust. We’re at that junction.

Automation combined with Artificial Intelligence means this time there is no new labor environment. We just don’t need you anymore, Pete. Automated software and hardware only need a few specialists to design and operate them. The cognitive requirements for such sophisticated engineering work are simply beyond reach for the majority of the global workforce. Re-education bootcamps and online courses aren’t going to help much.

There are two main forces at play:

A: The automation of work is increasing

Back in the early 2000’s, there was a lot of excitement around industrial automation. Replacing people pushing carts and packing boxes inside factories. Machine vision to instantly examine every single product for defects. Increased production. Fewer stoppages. Higher quality. Lower cost. No brainer, literally. Some workers moved into cushy back-office jobs, others didn’t. These aren’t intelligent machines either; they mechanically repeat a series of carefully (human) choreographed movements to do the job. Like a metallic and slightly terrifying marionette.

Now we’re already seeing the advent of Robotic Process Automation, which is doing the same thing to software and data processing. Instead of having an army of offshore workers process Purchase Orders or process expense reports, you train a software robot how to do it. This will be a huge disruption for the massive offshoring industry in Asia.

The problem is exemplified in the recent news from Infosys releasing 9,000 employees from jobs that can be carried out by just 500 employees empowered with automated software.

This pace of human replacement becomes exponential with the advent of artificial intelligence. An intelligent robot or algorithm isn’t restricted to clearly repetitive manual tasks. Intelligence implies decision making, even creativity. It can learn by itself. You can replace entire business processes and divisions of human workers with artificial intelligence. We are the very beginning of this evolution of automation. Buckle up.

Elon Musk has estimated that up to 12-15% of the global workforce will be displaced through driverless cars. That could happen over the next two decades, even faster in developed markets. That isn’t a lot of time to prepare for this transition.

B: The complexity of work is increasing

Yes, there will still be jobs. The requirements will again take a notch up. Way bigger than last time. Right now the trend seems to be to get into software, but that may be short-sighted too. With visual design tools and A.I., you can certainly replace an entry-level programmer. Pretty soon A.I. will re-write its own code.

A good algo will create fewer bugs, consume less pizza, and work around the clock

Today’s flavor of the month, data scientists, are having a good time in the job market. Everything’s better with a sprinkle of #machinelearning, amirite? Will that be the case in 10 years, though? Machines are getting pretty good at finding patterns in data, too. Betting your career on a specialist skillset in such a time of disruption is a tough proposition. In my mind, I would rather bet on interpersonal skills, a wide personal network and a solid understanding of basic sciences.

Good generalist > great (obsolete) specialist

So in summary, most manual work will be replaced by physical robots, and most simple cognitive tasks will be replaced by software.

Sooo, what are we humans supposed to do then…?

#1: Power

Humans are great at concentrating power. I will take your marbles, because I can. We’ve already seen this in the development of the stock market. Historically speaking, profits are at an unsustainable level. It shouldn’t be possible. But it is, and technology, offshoring and now automation are the reasons. Companies can afford to pay people less, often the biggest cost driver, while continuing to generate more revenue. Slowly they are also reducing headcount, but consistently creating bigger profits for shareholders. That’s how the system is designed to work. Sell more, pay less #bossmove

Today the 8 richest people on Earth own as much wealth as the poorest 50% of the world. If you thought there was already a power imbalance in the world, wait for the next 10 years as automation and artificial intelligence spread across industries. Wealth will increase, but it will fall into far fewer hands.

If knowledge is power today, intelligence is power tomorrow

The unfortunate truth about the current global economy is that it is driven by increased shareholder value. Not employee. Not society. Shareholder. Not a shareholder? Sucks to be you.

#2: Creativity

Humans will still own creation. Well, for a whileMaybe not that long, actually. Algorithms can already write code, design hardware, write songs, craft poems, and paint. Still, certain people will always strive to create, whether or not there’s a market for it. We’ll always have some form of art. We’ll play sports. Yet for most commercial purposes, the machines will take over. Their rate of improvement isn’t linear, it’s exponential. They simply aren’t limited by squishy things like DNA.

Relax Bob, go throw a ball or something. Leave the thinking to your buddy Hal.

Eventually, the machines will design better machines, and we won’t even need to worry about that. We won’t know how they work, but how could they explain it in terms we mortals could understand…?

#3: Leisure

What if we just, like… didn’t work. At all.

Think of a vacation that never ends. What would you do? Anything? Everything? It would be amazing initially. But is fun really enough for humans? Weren’t we great explorers once? Would Albert Einstein have settled for fun? Buzz Aldrin? Elon Musk? Would you?

(Un)Luckily, Virtual Reality has you covered. Try the virtual Mars experience, narrated by virtual Elon Musk. Download Zuck’s Virtual Startup and be the next Facebook! Take a break from serious work, and let’s become warlock elves in World of Warcraft VR3D. Why even walk at all, when you can fly on top of Mt. Everest in the comfort of your own sofa?

Think fat people on floating beds doing online shopping all day while sipping smoothies, like in Wall-E. Think people permanently engrossed in Virtual Reality, like in opium dens of old.

Then again, maybe we’ll all just take up fun, rewarding hobbies like yoga and scrapbooking. Well, virtual yoga if we’re feeling lazy that day. Or just watch the A.I. workout, really.

Don’t bite the metallic hand that feeds you

If we’re not working for money, then who’s paying for all this? Bill Gates and others have advocated taxing robotsNot even Wall-E can beat the tax man, hah! While a likely short-term solution, that clearly becomes entirely ambiguous and impossible to police in the case of software automation. Dagnabbit.

Elon Musk, among others, has proposed the concept of Universal Basic Income. Finland is already running trials. You get paid to live. No questions asked. Do whatever you want. Woohoo! Except maybe crime. And strictly no complaining about the machines.

This is the unfortunate end-state of this path. We’ll be relegated to pets, essentially, playing around harmlessly while the machines work tirelessly on our behalf. Hopefully more like a human-size hamster cage with cool slides and stuff, and less like the human battery farm of The Matrix.

Trump, the great socialist, to the rescue!

Ironically, it may be capitalist poster boy, Mr. Trump himself, who comes to the rescue. What we’re already seeing is a new trend of nationalism, focusing on internal affairs first. Companies are suddenly rewarded for creating jobs in the homeland. That obviously doesn’t sit well with global competitiveness. For most businesses, it would be more efficient to offshore and automate. Perhaps Trump’s populist approach is a way to slow down or counter this trend of automation by enforcing domestic employment… maybe he really IS that clever? Hmmm. Umm? Yeahhh.

The real issue that we’ll be facing is the true purpose of human progress. Is it still progress if it only applies to a chosen few?

Even if such issues were dealt with temporarily and locally, the networked global economy will create interesting new conflicts. Most U.S. companies have a majority of their business and operations outside the country. Think Apple, Google, Amazon, GE, J&J, Pfizer, Intel, etc.. Are they liable for outcomes of workforce automation in every country? Will Apple be liable for eventually replacing their planned India manufacturing staff with robots and algorithms?

Who decides that? It’s going to be a big mess.

What can you do, except get fat and wait for the matrix to switch on?

Become a shareholder. Be the guy that automates, not the guy that gets automated. Build passive income. Learn yoga. Start your own business. Enjoy life as we know it.

It ain’t gonna last.

Source: LinkedIn

0 out of 93.

That was the winning record of all teams in NFL history, when facing the situation Tom Brady was in during the final minutes of the Super Bowl last Sunday.

He made it 1 out of 94. Here’s how.

#1: You win games by winning plays

Winning in startups, winning in life, it rarely happens in an instant. Years of struggle can go by, just to taste a brief moment of glory. Then you’re back in the daily grind. Instant gratification is rare in business. Overnight success is built over a lot of long nights.

…that’s why you play to the end – Tom Brady

You focus on every individual play as if it made all the difference. Danny Amendola could’ve eased up at any moment when two defenders tackled him at the goal line. The two point conversion was a long shot anyway. Statistically, their team had already lost. Yet he didn’t think about the next play, nor the game, or about statistics. This was the only play that mattered. He struggled in the moment, but he made it count.

#2: Others don’t need to believe in you

The odds are against you as a startup. It’s hard. Customers say no. That one VC laughed you out of the room. Your family thinks you should quit on your idea and get a real job. But those people aren’t you. They didn’t come up with this idea. They never believed in it. Do you?

Nobody would reasonably expect that in the situation pictured above, the guy in the silver helmet wins that play. He’s sandwiched between three defenders, he’s in mid-air, there’s a leg in his face, and the ball has already bounced off another player’s shoe. Those three defenders are certain he won’t catch it. Even after the play, everybody else thinks he didn’t catch it, so the referee watches the slow motion footage.

Julian Edelman believed he could make the catch, and he did.

#3: You can fail many times, you only need to win once

Half-way through the game, the Patriots were down three touchdowns to zero. The biggest comeback in Super Bowl history had been from 10 points behind. The Patriots were down 21. The opposing team was high-fiving and celebrating at the sideline.

Tom Brady, the Patriots quarterback, wasn’t playing to his best ability. He missed some key throws that he would usually make. Balls were being dropped. They even missed an extra point attempt. Nothing was going according to plan. The game was a constant struggle. There were no easy moments. No gimmes. No gifts. No lucky bounces.

We never felt out of the game – Tom Brady

When you see Tom Brady in the picture above, you’re thinking he’s given up already. It isn’t working. Despite winning the Super Bowl several times before, this wasn’t going to be his night. Maybe it’s someone else’s time to win.

But Tom Brady still thinks he can win.

The Patriots played the entire game from a losing position. Until they won, in the first overtime in 50 years of Super Bowl history.

Source: LinkedIn

I was having lunch with Luke Janssen, founder of Tigerspike, the other day talking about history. Specifically the first world war, and how technological progress caught humanity off guard like never before. Luke made an interesting connection to the disruption we’re seeing in Fintech today, so I stole his idea and made this post.

A gentleman’s game

Today we think of war as absolute horror, something that ruins nations and the dreams of entire generations across the world. Yet in 1914, the civilized nations of the western world were experiencing something called war fever. As perverse as it sounds in today’s world, war, in the European context was seen as something of a purification ritual, a romantic adventure for young men to make their names. War was still seen as a necessary part of culture, to keep humanity from descending into decadence. It was a grand game for gentlemen to rattle their sabers at the border. After battle generals would shake hands, exchange swords and the victor would win a province or two.

The iconic image of this virtuous image of war is the French cavalry. Bright shining breastplates, horse hair helmets and of course bright red pants to complete the French Tricolore. Virtually identical to the horsemen of Napoleon’s army one hundred years earlier. Nothing had changed in a century, imagine that. Their weapon of choice was the sword, yes sword, waved for dramatic effect with white gloves. These soldiers literally went into war with white gloves. Regular infantry soldiers had no helmets. They wore wool hats to the first world war. It’s hard to comprehend with the advantage of perfect hindsight.

Banks are the French Cavalry of today. They’re going into a war with white gloves.

Think about it. Banks are like civilized nations battling each other in a real gentleman’s game, where each will have it’s short lived glory. Launch a new product, hurrah! This new campaign will bump us a slot or two in the rankings, jolly good! Why don’t we open a new market, and show those locals a thing or two! Let’s send our best and brightest to duke it out. Tally-ho and ho-hum!

Say hello to my little friend

In earlier wars, the fierce hussars of the cavalry would often overrun infantry positions with aggressive sprints across the battlefield. Cavalry had been the most dangerous weapon on Earth for thousands of years. Think Ghengis Khan. Alexander The Great. Infamous for their deadly cavalry. In previous conflicts of the 19th century, opposing forces could only fire 5-6 rounds per minute in defense with their dainty rifles, by which time the cavalry had stomped their fancy frocks into the mud. Why would this war be any different.

In this new war, waving their shiny sabers and prancing onto the battlefield, the cavalry was met with machine guns. Machine guns are different. They fire 500-600 rounds per minute. In the first minute of the war, the cavalry was made obsolete. Thousands of years of military doctrine was made meaningless in a matter of seconds.

Fintech is the Machine Gun

The startups and tech giants out there aren’t competing on your terms. They’re not playing your game. They aren’t going to shake your hands, in fact, you won’t even know you’re in a war until you’re losing.

So what do you do, as the cavalry facing the machine gun? You throw away the horse and dig trenches. No, you can’t win the battle, but at least you won’t lose! You embrace new banking regulation with open arms because it means the cost-of-entry into your market is ever more difficult. Trenches!

Now you think you’re safe in your trench. Bullets are still whizzing above your head, but you’re safe.

Fintech is the Land Ship and The Zeppelin

Can you imagine what it must have felt like, to be the first soldier to witness a metallic, growling ship slowly emerge from the mists of the battlefield. But you’re not on water! It’s a land… ship. Your bullets are bouncing off. It keeps coming. You can’t stop it. What is it?!

Can you imagine the residents of Paris, seeing a dark cloud gathering over the city. How curious people must have been. Then it starts dropping bombs. You thought the war was being fought hundreds of miles away! Home was meant to be safe. The Hague convention strictly forbids acts of war on civilians! How is this possible?!

It’s a war of attrition. Total war.

This is what banks are facing. Your competition isn’t playing by your rules. They’re making their own rules. They aren’t asking your permission.

What are you going to do about it? Broker a deal and create an alliance? Dig the trenches deeper? Create your own secret weapon?

Source: LinkedIn

Here’s a fun trivia question for all you digital folks:

What digital channel gathers
25,000,000 transactions each month,
but doesn’t fit in your pocket?

Drumroll… the beloved Automated Teller Machine! By the way, that number is just for DBS in Singapore, who operate the world’s most active ATM network.

We recently nabbed a little trophy with DBS, for our work on their upgraded ATM User Experience. While most ATM’s you’ll see across Asia come with a retro two-color text-only experience, we wanted to get as close as possible to the mystical “omnichannel” experience. That means the ATM UI should look more like your website and apps, less like your pocket calculator from high-school.

If you think about it, as long as we need cash, which could be another 10 or 20 years by some estimates, why would you ignore the experience of getting cash? Cash is king, particularly in Asia. While some ATM’s can support up to 50 types of transactions, from credit card applications to buying government bonds, the core transaction is still cash.

By running the numbers on typical cash withdrawal amounts, we were able to provide intelligent and personalized Fast Cash options right on the home screen for a one-click transaction. The outcome is shorter cycle times and ultimately shorter queues, which is what you want to see. With bank branches on the decline globally, ATM’s will not disappear, but will actually see increased usage.

As Über and Amazon do with payments, the best experience is no experience at all

You can also personalize the experience like you might do on a real website, by changing the font size and language from a contextual menu. The first time you try it it’s almost tempting to play around with it, because it feels novel in such an old school context.

Something that really struck me was a senior banking executive telling me the story of how his elderly father never trusted the ATM’s when they came out in the 1960’s. He wanted a human to count his money, surely the machine would get it wrong at some point! You can only trust your money in human hands! In a lot of ways that’s what we’re seeing today with the onset of digital banking, with more and more transactions moving onto mobile devices. Older generations will always prefer the security and comfort of whatever channel they’re used to, be it physical or digital.

Whether it will keep rolling out cash, or serve other purposes like video chat or card issuance in the future, we should not forget this channel. Long live the ATM!

Thanks Alok, Rob & team, that was a fun journey!

Source: LinkedIn