From cryptocurrency to AI consultancy

Seeing a need to demystify artificial Iitelligence (AI), Grayson Leversha and Freddy Hopkins co-founded consultancy and software development venture Greenstone in 2023. 

“I started the company after I realised there was a big need for businesses to not only implement AI, but also understand it and the power of what it can do for their business,” Leversha says. 

The duo have bootstrapped Greenstone with funds from previous ventures – crypto included –  and both consult and build products for customers with a mantra of getting something out quickly, and building with the customer in mind. 

“We are obsessed with the customer and push for a strong feedback loop. We incentivise our customers to give us strong, honest, constructive criticism on the product,” Leversha says.

‍

Leversha and Hopkins sit down with Caffeine to discuss how they use AI in Greenstone and where they see it going. 

‍

How is AI being used in your business?

Grayson Leversha: Internally, we use AI in many different ways. The most important thing to know is that we have built and are continuing to build an extremely automated software development process. We are building, have already built, and continue to grow our own machine-learning models that write code for us and help us build our products. 

Every day, we try to think about how we can remove a human from the software development process and how that results in quicker times and lower costs. That is, obviously, for our clients and our own products that we release in-house.

‍

What do you say to those who feel threatened by AI? 

GL: With every removal of someone due to AI, there will also be an increase in jobs that need to manage the bots and the AI models. Technology changes have always happened over time. The combine harvester didn’t kill farming. It meant the farmers could become more efficient.

Freddy Hopkins: To get to the stage we are now with human evolution, there have always been disruptors. AI lifts the level so that everyday people can be better at what they do using these tools.

‍

How are you fostering AI literacy within your organisation?

GL: It’s always about that compounding improvement of learning and knowledge. We love what we do, so learning more about AI is not a chore; it’s a hobby. We like to read and watch everything on every medium. We then ask each other: “What do you think about this video? What do you think about the repercussions of that?”

FH: We’ll tag each other in articles. We’re sending things over to debate, especially when new bits of technology come out, to see if we could integrate that ourselves. A lot of new stuff is coming out in the space, and being first in something can be very fruitful.

‍

What has been your biggest challenge with AI?

FH: Informing our customers about the benefits that AI can bring to the organisation, whether it’s increasing efficiency, productivity, or customer support, without wanting to make them think that it’s going to replace them.

I didn’t think there would be this much fear-mongering. I thought everyone would be excited about AI because it helps them become better. 

‍

How do you manage privacy concerns?

GL: The models we build and all the data they’re trained upon are internal. If we are building an engine for a client that is trained on their internal data, then obviously, that model is only for that client. We don’t make models based on one company’s data, train them on another company’s data, and then sell them to another business. 

Cloud providers such as Amazon Web Services and Google Cloud need to be held responsible for ensuring good data privacy because, respectfully, there’s only so much businesses can do when they host data on cloud services if there is a leak. 

That can come through legislation. In New Zealand, there needs to be alignment between the government, legislators, AI experts, and the New Zealand public to ensure that the legislation is not only correct but also secure. 

‍

What AI trends are you watching? 

FH: In the last couple of months, GPT4 has come out. The technology can now take voice, video, all these outputs, these mediums that were once barriers for computers to understand, and make them completely understandable. It’s so good for tutoring, for example.

GL: I’m interested in how AI can be used for education and helping people who may have difficulty with human communication or learning disabilities. I have members of my family with some learning difficulties, and they have always been the best in our family at working with technology.

We’re also quite fascinated by Elon Musk’s Neuralink. It’s – well, it’s terrifying – but I’m always an optimist. I see the benefits, and then I say, ‘How can we use AI machine learning to help bring the best out in people?’.