What’s one thing you wish more people understood about AI?
There are different paths to answer that. Humans naturally think linearly, and we need tools like scientific methods and mathematical models to help us go beyond that. We struggle to think exponentially or in high dimensions, so we often use analogies and first principles analysis to understand complex concepts, including AI. Therefore, there’s a tendency to put AI into something that we’re familiar with so that we can understand it. To anthropomorphize it and think it will behave like a human if it mimics human intelligence. This can cause fear, especially when we project human traits like dominance onto AI.
But it’s important not to conflate intelligence and dominance. I believe dominance is an outcome of evolution that humans, an evolutionary trait linked to survival. That’s why we want to create a democratic society, to protect and put guardrails around that. And we sometimes assume AI will naturally develop similar traits. However, Al can only possess capabilities that we code into it. Human intelligence has a natural tendency towards dominance, but it takes a human to codify it into the AI.
Looking at AI in another way. I send you back to 1924 and your job is to explain Zoom on a mobile phone.
So you need to now tell somebody, hey, there’s something called a smartphone? You and I can Zoom call on it. And how does that work? Well, actually, you know, it’s a Zoom call, so what’s a Zoom call, and how does it run? It runs on a mobile phone. Wait a second, what’s a mobile phone? Mobile is just a mobile computer. Wait a second, what’s a computer? Oh, a computer is like a very powerful calculator. But wait a second, you have two computers. How can they talk? Well, it’s something called the Internet. Wait a second, what is the Internet? You know, like the telegraph. Yeah, but a telegraph does not transport any images. Now how do you explain to a 1924 individual how you and I can Zoom on a mobile phone?
There’s no analogy in 1924 that will work. I think that’s closer to the truth of AI, and what it will become.
Why do I pick 1924? Because from 1994 to the Internet, I guess like maybe 2010, 2014, that’s probably the biggest 90 years. But then think about a 90-year jump compressed into ten years or five years. And that’s the key difference about AI.
So that’s one idea I want to share. The other is that AI doesn’t truly reason. While it can appear to reason by predicting outcomes based on statistical approximations, it’s not genuine reasoning. Closer to accurate is that AI reasons by analogy, which is different from first-principle reasoning. For instance, reasoning by analogy in 1924 wouldn’t lead to the concept of the Internet.
True breakthroughs, like the amazing people who looked into the physics of how to actually look into the signal processing, which gave birth to radio, gave birth to now remote communication, gave birth to the Turing machine model, is something AI is not yet capable of. That’s a different type of reasoning that actually AI does not possess today. We are nowhere near to having reasoning and understanding, not to mention bringing that product into the world.
How did you identify the need or opportunity within the home building industry specifically for OpenHouse.ai?
My background is always in predictive models, but the industries where I started out had been digitized or heavily digitized, which makes sense because if you don’t have data; you can’t really build machine machine-learning models. After telecomms and hedge fund banking, I came to the realization I want to do something important that has a big impact, and a good positive impact to many, to the public.
Home building is very cool in a way that is very tangible, it is very close to human need. It’s very physical. It’s where the digital meets the physical world and real world physical impact. It’s important because everyone needs shelter, and it’s super different because actually it’s a fast industry although for actually for anyone on the tech side they ask why I would work in such a legacy industry. But to me it’s very cool to work in an industry so different from the machine learning world. And it offers an opportunity to do something important, something significant.
At the beginning, I considered whether I had the skill and skill set for a tech startup. I had joined a startup as CTA to build an R and D team, so I knew how to run a big team. But it’s a completely different game to go and start to build a team from the ground up. I’d never done that. And I realized just like it takes a village to raise a child, it takes a community to raise a startup that never existed before in the category.
Now the question was, can I form a community? I’m a nerdy data scientist who had big corporate jobs. Now I had to figure out a built team from the ground up. So I thought, how about I start a Google Cloud developer group? So first I helped out the GDG YYC and then I actually went further to start a developer group in Edmonton remotely. At the same time I was thinking of going to talk to some home builders, and we were very fortunate to find a few home builders in the beginning to actually walk along the journey with us. I’m so grateful for them, and they opened up and said ‘this is how we run our business.’ And that’s when I realize, no one on the tech side had actually really spent the time to appreciate how complicated the home building business is.
They are simultaneously running a factory, like manufacturing, and doing direct to consumer. But they also need to manage risks and capital investments, and all the policy they need to navigate. So it’s like it’s a risk management firm. All of this is compressed into single company along with massive project management.
It’s a massively complicated business and no tech company was actually rolling up their sleeves to understand the intricacy of the business. That was the opportunity we saw.
How does your work at OpenHouse.ai deliver on your promise to “transform homebuilding”? And what does this mean for the average person looking to purchase a home, especially in a market that’s viewed as unaffordable for many?
Now, I wish we had an answer that could immediately solve the problem, but it’s not so easy.
There are important hypotheses that I can respond with. One is, do you believe in the effectiveness of the market, at least to a certain degree? We know the market is not super efficient, but do we believe in competition? In the market that’s actually the main driving force to drive the profit eventually to benefit consumers. But if we believe that an efficient market will eventually propagate profit through competition, we believe what we do will eventually trickle down to the public.
in a competitive market a business needs to cater to the customer and meet them where they are. So the business who knows and understands the customer’s wants and needs the best will be the winning home builder of the future.
By helping builders understand buyers needs and wants, they will be able to build better homes for their customers. If they build better homes, they will be more profitable. If you are able to build better homes for the market and through competition, eventually, competition will actually drive all these benefits back to the consumer. And that’s our belief. Now, of course, I recognize the big assumption is, if the competition will actually be able to continue to propel the home building industry to improve building better homes for consumers in a more efficient and cost-efficient way. If we’re able to do so, then yes, then we are able to benefit to the consumers.
What are you focused on these days? What’s taking up your attention?
Really to make sure that we are prepared. We anticipate very fast growth in the months to come. We need to be able to capture the opportunity for ourselves. At the same time, we need to make sure every single customer that we are onboarding enjoys the experience.
We know how things should be, and while we’re the team that started the business, we’re also the team that is growing the business. And that’s my focus right now.
Is there a book or a podcast you’d recommend to readers?
The first is called 7 Powers: The Foundations of Business Strategy. A lot of people talk about strategy, but what does strategy actually mean? So it’s a very interesting definition of strategy, which is essentially your pathway to build power in your business. And then Thinking, Fast and Slow.
For podcasts. There are so many good ones. One is Lex Fridman, a research scientist at MIT.
