Jobs, as we have known them since the industrial revolution, are now under several threats: Low wage countries, of course, with the lever of global trade and technology.

But global trade and technology is indeed what will create huge opportunities in the next years, even in industrialized countries:
Imagine a small company that want to expand internationally. Today you have two options: create subsidaries, or close deals with local companies to distribute your products or services. None of them is adapted to modern economy that requires high speed and efficiency.

Here comes the API economy: What if we expose our business also through APIs? This will allow people all over the world to efficiently provide our services to their customers. For example, a person in Israel could provide our repair service in Hebrew, and a new business is born:
- we sell our service to people that we would never reach (albeit with a smaller margin)
- our affiliate in Israel builds a local customer base, and make extra money
- we keep things simple, in the end an “affiliate user” just needs to accept our terms of service and we provide him with tools

And what makes this kind of wealth creation possible is an API at the intersection of a company’s know how and an affiliate’s market reach.

In API economy takes off, we will see less traditional jobs and more people engaged in one or several API businesses. Implications on society and economy can be huge.

Of course, I keep track of many metrics, the obvious ones (new customers per day, monthly revenue) and the domain-specific ones (like diagnostic accuracy or repair success rates split by camera models).

But you only can track what users do, and the most important information is about what people doesn’t do and why, from people that will not even become users, or from users that give up because your product sucks. But at the end, there is no easy way for you to know why your product doesn’t get noticed or sucks.

Therefore, the information that you collect tends to be heavily biased towards confirming what you already know, rather than telling you what you ignore or what needs to be fixed.

One example: Localization
Our Repair Service is currently only available in English. In Aero Quartet we have also native French and Spanish people, so we would like to make the service available in those languages, and also in German (we’ll need to hire for that). Now I need to make numbers and figure out how much growth I can expect from localization in FR, ES and DE.

I know that exactly 7.4% of my customers are in french-speaking countries or regions. Easy.
But the really interesting figure is how many would-be customers are turned away because the service is not in French. And that figure is not something I can measure, because people searching for “réparer fichier XDCAM” will never find us or contact us in the first place.

Therefore, to take the right decisions, I will try to look at it from different perspectives, and take calculated risks, just like a blind person compensating with other senses.

And in my kid’s mind, all those people who were creating this “magic”, all those distant places in America like Cupertino, were sort of legendary and I was avidly reading about them through the computers magazines of the day, like “L’Ordinateur Individuel” (to my surprise it’s still alive!).

He was not even 30 years old, and he was already THE legend.
Not a mainstream person, like today, but for people who cared about computers, he was already the child prodigy.

Your humble Aero Quartet CEO will never have a chance to meet Steve Jobs, but owes him a lot.

I have started Aero Quartet because I knew that an individual can make significant contributions. And I knew that from him.

When you feel you can’t make a difference because you are at the wrong place doing the wrong thing, then you must make hard decisions and switch gears.
At my low level, my “jobsian” moment was in 2008 when I decided to leave HP and go full-time on Aero Quartet, and I don’t regret it!

Just watch Jobs Stanford commencement speech (YouTube) if you haven’t done it already, and thank him for the pearls of wisdom.

RIP Steve

But a few questions remain unanswered:

• How do you tell your repair kits what regions of the disk map should be repaired?
• What happens with disputed cells like A6?
• Which repair kit use?

DeepMediaScan “Clips”

Your Disk Map is a cell-based representation of the contents of the disk. Now we want to group cells together into “Clips”.

Looking at the Disk Map images, it’s easy to determine that biking footage occupies cells A1 to A6.
Similarly, birthday footage is in A7-A8 and B6-B7. As A6 and B5 are disputed, let’s include them also here: A5-A8 and B5-B7.
And we do the same for party and concert images:

Therefore, you just have to define 5 clips. For each clip, just select the images and drag them into Treasured.

Birthday footage is split over Clip2-A and Clip2-B, because the file was recorded fragmented. We cannot define it in one clip, because Treasured wants the images inside a clip to be consecutive. We will see later how this affects repair.

Repair Kits

Once clips are defined, we want to repair them.
As explained in part II, the clips have been recorded in 2 formats and need 2 different repair kits.

We know that Clip1 and Clip3 are in HD format, because the corresponding Disk Map cells have an HD resolution of 1280×720. The other clips are in SD, their display resolution in 640×480.

This means that we know exactly what kit to apply to what clip.

The diagram below shows that this time, we can recover all the footage.

• Once repaired with HD kit, Clip1 and Clip3 contain 100% of the original footage for biking and party.
• Once repaired, Clip2-A and Clip2-B contain 100% of the original footage for birthday. The two files can be merged together with a video editor.
• 100% or concert footage is recovered from Clip4 with SD kit. Note that it’s done in one shot, the “line-break” in the Disk Map does not divide it into two clips.

Conclusion

We have tools to recover every bit of footage present in the present, even if there is no visible file, or if files have funny names and end up not containing what is expected.

If DeepMediaScan doesn’t find it, it doesn’t exist.
If DeepMediaScan finds it, we can repair it.

18 months ago I quit my day job at HP and went full-time, then a few months later I hired the first employee, then the second one, and so on.
As you can imagine, teen years are busy, fast-growth years, and our identity is building accordingly.

That’s why we have designed a new logotype:

This is a crane bird during migration. It let you imagine the company values that we want to convey through it.

The logotype will appear in products in website as soon as we find time to refresh them.

It’s a bit as if a layer of fresh ice appears and becomes every day thicker. After a couple of months, you start refraining yourself from posting, because what you have to say doesn’t look “important enough” to break the ice.

The longer the silence, the harder it is to break it.

So today I’m taking my little hammer and with this modest post I reopen the channel.

The website had been the atom the the Internet for a decade: You would shop on Amazon, find flights on an airline website, sell on eBay, look for a job on Monster.

Now for many people the Web is Facebook. This where the services are, where their people are hanging around, where their leisure time is spent. This recent article, Facebook wants to be our One True Login, and the habit of using Google Search as a replacement of the URL bar, show that most people don’t get how the Internet works.

Where do you want to live?

And the success of platforms like Facebook is not surprising: it’s a place where people doesn’t feel confused, a curated, safe environment where they can do their stuff. Without messing with URLs, multiple logins and the dangers of the wild Internet.

And developers follow users, because users is where the money is or will be. This leads to a fragmentation, happening now under our eyes.
If you are a start-up or a software developer starting a project, one of the strategic decisions is: On what platform?
A few years back, the answer was evident: Web.
But powerful platforms have emerged: Apple’s iTunes / App Store and Facebook are the most impressive examples. Palm purchase by HP may end up creating a third powerful alternative.

Platforms?

Those platforms bring customers by the millions, take care of the most annoying aspects of web services, like login, marketing, payments, among others. Why would then developers want to open their web shop “in the wild” and fight for visibility, customers, and reputation, if they can have a comfy land inside the walled domain of their Lord?

Therefore, the Internet is now shrinking, not in size, in number of users, but in ambition.
Ambition is heading down for two reasons:
Innovation is more likely to happen in the wild than in walled domains with strict rules to respect. And if those new platforms are draining a big amount of the innovation power, there is less for the Web.

Why does it matter?

For developers, because they are in front of an ineludible strategy decision: What platforms for my products?
For users, because they are giving away their personal information, consuming habits, payment information, time and money to a single vendor.
And bet all your assets on a single vendor is risky.

This is why the Web should ultimately win, because nobody controls it, because it’s multi-vendor by essence, and because you are not tied to a single vendor that ends up controlling you.

The last cycle embraces roughly the last 15 years, and it’s been an amazing journey:
I remember back in 1994, sitting in my university dorm room (or what it my neighbor’s room? I don’t remember owning a PC at that time) and trying the new stuff: Linux with X-windows and a little program called Mosaic to see graphical pages on the “World Wide Web”. Or was it running Windows 95 beta with Netscape 2 ? Or maybe both, I’m not sure.

There were actually very few sites to visit. I remember being told by a friend (there was no portal, no search at that time…) to go to the Louvre website, one of the few to have high-quality pictures. It was amazing.

Porn sites didn’t exist either. To get hot stuff, you had to download text files from Usenet and build images out of them with complex tools like uudecode.

As incredible as it may seem today, you couldn’t search either. Just look at the web archive and you will see that Google did not exist until late 1998.

You couldn’t find stuff. No music. Computers at that time were not powerful enough to decode mp3 music, and did not have enough storage anyway!

So the web at that time was very small, and had very little “functionality” or use for day to day activities.

Fast forward to 2005.
Web 2.0 is all the rage. The Internet has become a marketplace, an application platform, people starts blogging en masse. Wikipedia shows that collaborative efforts can.
Maybe circa 2006-2007 we have witnessed the high-water of the web, ambition-wise.

We never have been so close the see the vision of the 21th century web fulfilled: an universal, open, ubiquitous software and services platform.
Despite impressive efforts, like HTML5 standard, Webkit, Cappuccino, and the demise of poisons like Flash or Internet Explorer 6, the web is under new threats.

Next post will explain why the Internet is shrinking in ambition, and why it matters.

For a one-man-business that has grown slowly over the years, adding a second person comes as a shock: you discover that you have no processes in place, no office hours, no office space, no collaboration tools.
Suddenly, you have new legal liabilities. Worse, your business apps do not “support” the new split-up of tasks, you have to modify them. And having to share the information is a new habit to learn.

Therefore, one should consider hiring the first employee as an important business change, that must be carefully prepared and implemented. The benefits will come down the road, but during a few weeks, be prepared for your daily dose of frustration:

• your business will never be the same again. It’s no longer your “baby”.
• things you were doing routinely will require training. The “rookie” will not have your productivity, at least not in the first weeks.
• paperwork and set-up (see below) will consume a lot of time.

Hiring is really a daunting task here in Europe. Governments say they do their best to ease economic reactivation, with incentives and the like. But honestly, that’s bullshit. We are still in an over-regulated mess. Twentieth century stuff with applications threefold, and don’t forget the company stamp in the corner or we reject it, mandatory yearly office inspections, …

I had to do a ton of paperwork. None of it for free. Set-up costs are important, in particular when you hire the first person. It can work for middle and big organizations, where one-time costs are diluted. But very small operations suffer the full weight (money and time) of this bureaucracy. Not surprising that we have 20% unemployment rate in Spain, if hiring your co-worker is so difficult.

Despite all this, hiring is a turning point for Aero Quartet:
We believe that the movie repair business can scale, and we are taking the first actions to make it happen!