We’re all becoming Gods.

When my eyes were operated on, they replaced my damaged lenses with new ones. My vision is even better than before the accident.

In the flash of a scalpel, I was transformed into a superhuman.

But in other ways, we are all becoming enhanced humans, turning into Cyborgs.

It’s already happening in your brain.

The phone in your hand plugs your mind into a cloud computer containing exabytes of structured data running trillions of cycles of machine learning and supercharging your synapses and making you brighter.

With bio-data inputs, it knows how you feel and predicts what you need. It recognizes your biases, corrects your hormones and emotions. It is making you wiser, too.

Welcome to the Alpha version of your super brain. An augmented Cortex, if you will. None of this is my idea, of course. Yuval Harari shows how we are achieving godhood in his 2015 book, Homo Deus. I read it in 2017 and still love it.

But not all human capabilities can be enhanced right away. A real-time Babelfish, predicting the movies you will like, or mimicking Shakespeare is trivial compared to more intricate problems like autonomous driving.

Harder still?  Making a prediction if a complex physical security project (worth $3 million) can be delivered successfully and profitably. Until now, no computer has been able to do that. Virtually no person or company has been able to, either.

Large systems integration projects are considered very risky. Risk managers shape solution design more than subject matter experts. It is more art than engineering. Over 1/3 of them fail in one way or another. And of the ones that are handed over, less than 1/2 are profitable.

But I became one of the only people on the planet who could predict which projects would succeed.

Let me explain.

After a long career in IP physical security and eleven years of learning, I accepted a new job. It was January 2014, and I became part of one of the four Siemens Centres of Competence in Dubai. Every week, I would receive project proposals from our solution sales teams located in the Middle East, Asia, and Oceania. Only those large complex projects over $3 Million passed my desk.

I would have days (but sometimes only hours) to review the technical proposal and judge whether it ‘would succeed’ and whether it would ‘be profitable.’ To decide whether submitting the project to the customer made both technical and commercial sense.

How did I do? Well, at first, I failed, of course.

How could any person make a subjective judgment – without data – about something so complex and have any chance of success?

Could you check hundred and one things on the never-ending list:

• product selection
• design of network
• Sizing of all servers, storage, workstation
• SW configuration option
• RFP and all addendum compliance
• procurement plan (with all alternative solutions collected)
• project plan
• resource plan
• …

I hope you got it. It is a very long list of things to check. It’s a good thing that I suffered from the Dunning-Kruger effect. In my ignorance, I was too stubborn and too egotistical to admit that I could be wrong.

Very soon, however, I began to learn. 

Because I came face to face with the results of my choices, one hundred projects later, I was capable of measuring their success and profitability. This is how evolution looks like in real life.

I had the inputs – a lot of data; RFIs, RFPs, gigabytes of project plans; systems solutions, OEM vendor proposals, technical and user requirements. Intent descriptions and compliance requirements.

I had the feedback cycles. Effectively, I was machine learning myself.

So I started seeing the errors that were hiding in plain sight. The mistakes that would guarantee failure or project overruns, the markers of failures.

And from the infinite messiness, I began to see the patterns. I started recognizing the shapes, the golden ratios, and the similarities of the projects that became a success.

Jan Berka had become a “human AI”.  And with over 1,000 project cycles, I had been trained to become very, very effective.

This gave me a much more challenging problem to solve; how could I build my “human API” to deliver that experience to anyone who needed it.

• Could I package my skills, experience, and learning of thousands of projects to enable other people to get all the benefits?
• Could I systematize the project design process to ensure that none of the crucial inputs could be forgotten?
• Could I simplify the design process by abstracting all the common elements into reusable templates?
• Could I extend the existing digital twin models to include the whole project lifecycle?

The answer was yes. And that’s how I came to invent Cortex.

Till next time, Jan.