Archive for 2014

Entrepreneur-in-residence (EIR) has its upside.

You just shouldn’t go in to it blind.

Understand why you are accepting the EIR position.

Before you get to the point where you no longer need money:

20% of your time spent on raising money.

20% of your time spent on hiring people.

60% of your time spent on product development.

The interesting thing about VR is not what we are starting to do with it by creating experiences we’ve never had before but what we will do with it once we can do virtual porn and create experiences most software developers will never get to experience.

Sex.

Having sex with another person.

I’m talking about sex.

The one thing most entrepreneurs forget when they rush off to create a chatbot interface is that they didn’t take in to account my elderly mother-in-law who just spent twenty minutes “chatting to that nice man at the bank.”

Your technical co-founder will attempt to take pride in their work.

And you will at every step, unconsciously, attempt to make them produce work they cannot be proud of.

And that’s okay, because you needed a new developer anyway.

The old one had become too slow.

P.S. This is sarcasm.

Your technical co-founder will quit when it becomes too “challenging.”

And that’s okay, because you will need to learn that technical co-founders don’t quit because a project is technically challenging but because a project is managerially challenging.

Deep learning, which we called…

Machine learning, which we called…

Artificial Intelligence, which we called…

A bunch crappy research code that works sometimes, in a very narrowly defined problem domain.

Entrepreneurs are told “take risks” but they are never taught “how to fail.”

When mentors say “take risks” we just assume, naively, the entrepreneur will just figure out.

And if you have been through a modern Western (mostly American) education system it is doubtful you were exposed to much failure.

You will set your technical co-founder up to fail many times over the course of a project by having an idea of when a feature should be ready.

And that’s good because you will need to learn how to reconcile the desire for something with the time available to implement it.

Your technical co-founder will want to build software that is maintainable.

And you will want software that works.

And that’s a lesson where you learn that the future version of your current technical co-founder hates every fibre of your being for all the technical debt you accumulated with your half-baked ideas from the other technical co-founder.

Most (not very astute) entrepreneurs view a web application as a one off purchase.

Build it.

And then launch it.

And then let the customers use it.

Realistically this never happens.

Stuff breaks the moment someone (anyone) starts using it.

It’s like your fancy new product you have just released starts rusting the moment you let a customer touch the shiny new surface.

I think I once owned a motorcycle like that.

The difference between “things with computers in them” and IoT devices, as I see, is that the IoT device needs to be smart about the environment it inhabits and the context of how it inhabits that environment.

A thermostat adjusts the temperature.

A smart thermostat can adjust the temperature a little more efficiently from your smartphone.

An IoT thermostat is aware, on a fundamental level, “I am a thermostat” and can reason, in a meaningful way, what adjusting the temperature by five degrees down in the afternoon on a sunny day means and versus adjusting the temperature five degrees down at 4AM in the morning in the dead of winter actually means to instead.

It goes beyond some fuzzy logic with an historical record and becomes a new way to work with available data, usage patterns, and the interactions that humans and other IoT devices have with itself.

Your biggest obstacle, not to success, but to just getting started, will be your inner critic.

That little voice that just won’t shut up.

One of the major problems with chat interfaces is that if you don’t speak the language, you don’t get to use the app.

I might not understand French, but I can muddle through the GUI of the local metro app for timetables to get me to my destination.

In a chat interface, I would be lost, and then it would be as though the app never existed, and I would have to resort to the old fashioned way of using paper timetables.

Assuming that paper timetables still exist and weren’t part of a cost cutting measure now that the chat-interface timetable app exists.

Fixing diversity (and not just gender diversity or even diversity of colour and culture but also age and physical disability) is valuable in attracting employees from diverse backgrounds.

We can agree on that.

But fixing diversity is also valuable in retaining those employees from diverse backgrounds.

Your inbound funnel of diverse potential new hires is important, but if they are inside your organisation only to wake up one day and realise that after two years they are still being held back or are being actively suppressed, retention of good (diverse) employees will be difficult.

Your diversity programme doesn’t end at the termination of the hiring funnel.

I have always liked jobs that pay the bills but don’t require all of my time.

I am happy working 15 hours a week for $200/hr than 40 hours a week for $200/hr.

As long as you have enough to cover your expenses, and a little more besides, I see no point in pouring every waking hour in to a job just so you can accumulate more money, but only by the number of hours you put in.

The problem with chat interfaces, as we found out in the 70’s and 80’s when building text adventure games (The original Zork trilogy is my all time favourite) is that the player’s vocabulary is (potentially) infinite but the designer’s anticipation of that vocabulary is limited.

I would say 9 out of 10 people who are capable of building a product are not actually capable of selling a product.

Which is slightly better odds than the 10 out of 10 people capable of selling a product are not actually capable of building one.

I’ve thought up more neat and viable ideas in a month than I can implement in an entire lifetime of work so why would I be interested in stealing yours?

The one benefit of a chat interface that has not been addressed is that the UI becomes a customer survey.

At every interaction you will learn new ways of how your customer wants to interact with your app or service because they have a limitless vocabulary to do so with.

When designers set the vocabulary, it must be, by it’s very nature, limited to a pre-defined set of interactions.

A lot of people write about Millenials and what they want.

But the one thing I have observed with Millenials above all other generations, even my own, is the desire for personal growth.

The one thing that the internet has shown the Millenial generation is that there is a world of opportunity and experiences out there that other people are having, so why aren’t you having some of that?

Me?

I live for growth.

If I have ever felt a job or the culture at a company (when I directly worked for companies) is holding me back, if there is no opportunity to grow and learn and experience, then I wish the company adios and move on.

Imagine that feeling instilled in an entire generation.

A lot of people have wanted this for themselves, but they have buckled under the social pressure.

The internet shows the new generations “I’m doing it, why not you?”

I’d rather make the people that matter to me happy, than make happy the people that don’t matter.

If you build a product for your users where the user interface is a log file, your product-market fit might be a bit too hyper-focused.

Log files are good for figuring out “what happened” but not so much for figuring out “what we should do?”

And they are lousy for answering the “how do I do?” question that new users will inevitably have.

At some point you will talk to a different developer because you didn’t like the answer that your technical co-founder gave you, even if your technical co-founder stated the truth.

A part of growing as an entrepreneur is learning to accept truthful answers backed up with data, even if we don’t like the answer.