Let's Talk AI with Chris Cook

Meet Navarik’s Chief Technology Officer, Chris Cook.

For this edition of our employee feature, we share Chris’ unique journey from his degrees in Technology (Masters, University of Nottingham) and Biochemistry & Neuroscience (Bachelor of Science, University of Keele), running his own businesses, and a sense of adventure landing him in Vancouver, BC.

In his 10 years at Navarik, Chris has contributed greatly through his expertise, mentorship, and support.
Today, we're excited to chat with Chris about AI and his passion for technology.  

Q: Tell us a bit about young Chris!  

I was born in the Netherlands and my ties to Canada begin with my great-grandad. My family travelled a lot and we've lived in many countries like the Middle East or Indonesia. When I was 17, I took a trip with my family to Calgary and we had the chance to see Banff, Jasper and the Icefields Parkway. I was mind-blown upon returning home to London, and thought to myself how much I'd like to live in Canada. Here I am! 

My love for technology began when I was 6 years old. I was already coding on my old Spectrum, rubber-keyed computer with 16kB RAM and that’s where my love of computing and technology started. I’ve always found it exciting to make something that works and generates impact.  

With my love of computing, the thought of building an artificial brain was quite striking to me. That was what triggered my studying the life sciences (Biochemistry & Neuroscience) to understand more about how human brains worked. 

 I married these two passions together, which lead me to my interest in AI.

Q: What does the typical day look like for a Navarik Chief Tech Officer? 

It’s highly variable, but I enjoy it. It ranges from technical architecture, designing systems, process framework, reviewing and evaluating proposals, people management, project management, and general consulting across the business.  

I especially enjoy the coaching and mentorship aspect for all 28 of my staff! It’s rewarding to add to the growth of my team whether it’s inspiring them to think of a different way to do things or being part of their learning process. 

Q: How have your past experiences prepared you for your role?

I think I'm unique in the sense that I've got knowledge in the technology front, and science from my Biochemistry & Neuroscience degree. At Navarik, we're also working with inspection companies who are conducting scientific testing, so there's a minor overlap in that regard.

Because I've also run my web agency business in the UK, I have business experience from there as well. Early on, when I decided to stay at Navarik, it was because I saw potential from this fairly unique set of experiences. I saw that I could offer my skills and benefits to grow the company. 

Q: We hear that you’ve been working on an AI project of your own. Spill the tea!  

I built a backpropagation neural network with Java that learns using backpropagation algorithm (a learning method for artificial neural networks). This neural network is capable of looking at encoded data and can recognize patterns within it.   

The only end goal was to create it because of my deep-rooted interest in AI.

While there is no use in mind yet, it’s capable of solving linearly separable problems. If there’s any problem with a distinct yes or no answer, it should be able to solve it. For example, if I were to train it with enough data, it would be able to identify an inspection report with reasonable certainty out of all other types of documentation. With additional work, it could handle more complex, non-linearly separable problems as well 

Q: When we talk about training AI, would you say there is a suggestable volume of data or amount of time for AI to learn?  

Training time varies based on network complexity and data quality. The algorithm works through a process called convergence. You provide input and the first run should produce results that are way off, but then you go back, run it again after adjusting its internal state and it should be closer to the desired output. This process repeats until the network ‘converges’ to give you the correct answer.  

However, there is a bit of a challenge in that AI can be overtrained. For example, your network can become so good at recognizing one type of dog, but fail to recognize other types of dogs. Increasing the network size or capping the training process can help prevent the network from becoming too converged.  

Q: Reflecting on your time in university when AI was still in its infancy, are you surprised by the advancements made today compared to the expectations of what AI was to become? 

Yes. The holy grail of AI research is now known as Artificial General Intelligence (AGI) – almost like a conscious machine. AI has gone through various cycles of excitement, and the current focus on AGI is the most thrilling yet. Although AI has advanced further than I initially thought, it hasn't yet reached this level of AGI.   

AI’s development is remarkable, and when we get to the point of full AGI, this might start introducing significant moral and philosophical questions around the sentience or consciousness of the neural network. 

Q: AI has already trailblazed new paths in multiple industries – how do you see it impacting the commodity trading industry?  

AI is more than one thing. Right now, the excitement is focused on Large Language Models (LLM) like ChatGPT and Gemini. They’re amazing in the way that they can reasonably hold a conversation.  

While AI is currently prone to hallucinations and mistakes, these tools are going to be extremely useful in many respects where we can receive a solid response to any question asked. It is reasonably disruptive in the sense that it’ll remove work off our shoulders, but will it remove humans in the middle? Not for some time to come.  

I expect that in general, we’ll see productivity gains. There’s a further possibility of exploring business intelligence or training AI to give product-specific info but mainly under the guise of an assistant. Right now, my interest is exploring the use of LLMs to help us understand the large amounts of data we collect.  

Q: To what degree do you believe that AI has been influenced by media perception of how AI is presented in sci-fi shows, like Star Trek, versus literature influencing what AI has become today?  

I think there are a bit of two things going on here.

I think as a child who watched these shows, I also dreamt about these things and imagined what an exciting world that would be. I'm sure you can also find an example where someone made a scientific discovery and somehow it's made its way into science-fiction. 

There is an element where art inspires science, but also where science inspires art.

Thank you to Chris for sharing his time, insights, and experience with us for this feature. 
Navarik is privileged to grow with your expertise and impact.