Adapting to thrive: How AI rewrote Learnosity’s DNA for a new era of learning

Learnosity CEO Gavin Cooney and CPO Neil McGough dive into the company’s bold approach to AI in assessment. They discuss Learnosity’s swift adaptation to the evolving AI landscape, their customer-centric development process, and their vision for a future where AI unlocks authentic learning experiences for all.

Q. Looking back on Learnosity’s journey over the last year or so, how has your view on generative AI evolved since your first “a-ha” experience with the technology?

GC: GenAI has improved much quicker than expected. It’s improving more than any other technology ever. Google Search, Gmail, Excel and other normal technology doesn’t get that much better over the years, but AI is getting so good so quickly. 

Looking back a year ago, we were trying to parse files and it was difficult, but now it’s easy. We were trying to build worked solutions for math and it was next to impossible, but now it just works. 

It’s unbelievable how much better AI has gotten in such a short space of time.

NM: As technology has improved, people’s understanding of how to get the most value out of state-of-the-art tools has increased. In the early stages, everyone was just like, “Well, I sent this question to GPT-3 or Claude, but I didn’t quite get the answer I was expecting, so I can’t use it for that particular use case.”

But now, it’s more about using multiple elements together—whether it’s for red teaming or collaboration—to get the desired outcome. It could also involve layering more traditional machine learning techniques, like classification, to decide which is the best LLM to use for a specific task, such as feedback and scoring.

We’ve developed a deeper understanding of the nuances, which means not only has the technology improved, but we’re also able to leverage much more of its power effectively.

Q. You’ve attended many major conferences over the last few years where AI has been the hot topic. Since GenAI burst onto the scene in late 2022, how have expectations around AI in the edtech community shifted?

NM: In recent years at conferences, I’ve noticed that everything AI-related could be grouped into two distinct buckets: evolution and revolution. Those in the evolution bucket are companies that have a great product for the next 12 months because it does what everybody else is already doing, but faster, using AI.

On the other hand, the revolution bucket consists of companies whose products make you go, “wow, that’s amazing.” These products completely change how people assess or learn, but no one is in a position to adopt them for the next 24 months.

What you’re starting to see is that, while the hype is still there, there’s now a healthy dose of skepticism among people evaluating vendors in this space. While people want to learn about AI, they’ll only license it if they can understand how it’ll improve efficacy in a cost-effective way.

I have a fundamental hatred for the concept of AI washing, which is just putting the letters ‘AI’ on top of everything.

Q. Would it be fair to say that AI has in some way altered the DNA of Learnosity as a company? In what ways?

GC: AI has massively changed how we act internally and how we see ourselves. You can notice the difference in where we are, how we talk about ourselves, how we’re innovating, how we’re moving forward, and how excited we are about the whole thing. It’s a huge opportunity for us to add value to our clients and to drive innovation.

We put in resources and really tried to back ourselves. Halfway through the year, we had seven or eight people working on stuff in silos.

We unveiled our roadmap at the ASU+GSV Summit this year and it was an absolute smash. There was stuff there that people were clamoring over themselves to get their hands on. This was a lightbulb moment for us, where we saw all the value for our customers.

I had originally planned to sleep on the flight home from San Diego, but my mind was racing. Rather than resting to prepare for a kid who hasn’t seen me in a week and wants to play, I stayed up all night. I was so fired up by the reaction at the conference that I spent those eight hours writing a memo to the team and the board announcing that we’re opening AI Labs, hiring 20 more people, and going all in on AI.

We brought over Kate Hake, one of our product managers, from California to join our Dublin-based AI team. With her experience, and our team of AI experts, we can deliver real results.

AI has brought us back to our roots in terms of the startup mentality of building stuff, but with the maturity of the existing infrastructure and knowledge of the industry. It’s completely changed the DNA of the company—how we think, talk, and operate.

NM: There’s been a demonstrable effect on how we operate as a business, especially at the product level. We were a very mature, stable company and a lot of how we operated in that respect was bringing the customer early in the ideation and information gathering process.

But with AI we need to be able to paint a longer picture because our customers expect us to think long-term. We would usually only show things when they were almost ready, towards the end of the process.

AI forces you to realize that to do true innovation, you need that customer feedback and involvement super early to understand if you’re going in the right direction.

We’re bringing that way of working, focusing on early-stage proof of concepts, and involving customers, into the larger company.

Q. We’ve seen companies in many industries, including edtech, rushing to get AI products to market over the last year or so. What’s your criteria for evaluating product quality or market fit?

GC: There’s a part where we talk to people and say, “Hey, I have this idea, what do you think?” and we do the market research to see which bits are hitting with our customers. I always felt that a lot of people in edtech were selling vaporware—basically selling PowerPoints. That’s why I always like to give real demos, not just a picture of what it’s going to look like.

Then there’s a process of ideation and iteration of that idea. We see what the customers want, pick a few ideas that we really want to go with, build a POC, and we’re not afraid to throw away code or ideas if they’re not working.

Things are moving really fast in the AI space, so you might build something and then have to undo that or adopt more underlying technology. You need that agility.

NM: I have three criteria for evaluating products:

• Criterion number one is: does the product solve a real problem? There’s a danger of being a solution looking for a problem. Just because you do a lot of cool stuff with the technology, doesn’t mean it’s valuable to someone.
• Criterion number two is: can it solve that problem in a cost-effective way?
• Criterion number three is: does it improve efficacy, add value, or reduce costs to our customers? Can we stand behind the fact that it does what it’s supposed to?

Our biggest achievement is that we completely changed Learnosity in a year.

Q. On a related note, what do you think has been Learnosity’s biggest achievement with AI in the past year? 

GC: How we adapted so quickly, effectively, and holistically—I’m really proud of that. It’s hard to say whether I earned my salary on any one day, but I look back at 2024 and I’m honestly really proud of the massive change that we’ve driven in the company. We’ve pivoted to this space and we’re still fundamentally doing the same thing: building technology to deliver value to our customers.

15% of the company is now working solely on AI and innovation. I’m proud that we’ve done that. When we go to our clients and we show them our products, there are these a-ha moments and excitement.

Our biggest achievement is that we completely changed Learnosity in a year.

NM: I have a fundamental hatred for the concept of AI washing, which is just putting the letters “AI” on top of everything. What we’ve done is taken that technology and effectively applied it in a way that permeates the business’s operations at multiple levels.

We’ve managed to get the concept of invention and innovation using AI to a broader level. For example, we have things that were first done as a really quick and dirty POC in January, that made it to a more polished POC to show to customers, and are just now being released into the product.

Q. What kind of feedback have you received from customers about our AI-led products? How has it shaped our development plans?

GC: The feedback has been outstanding and exciting. Our customers’ customers are asking what they’re doing with AI, and we’re ideally positioned to deliver that value.

When we show them what we’re working on, they’re gleeful. When we get massive excitement, we know we’re doing the right thing.

I’m reminded of when we hit upon the idea of building an API back in the early days of the business. It felt like we’d finally achieved market fit, everything just clicked. A bit like the feeling you get when you meet the person you’re going to marry: when you know, you know. 

That same feeling returned when I spoke with customers about what we can do with AI.

Q. We recently announced the creation of a brand new team dedicated to AI research and development. Tell us a little about your vision and goals for this.

GC: We felt that just doing AI alongside everything else wasn’t going to cut it. We needed to have dedicated resources. Putting a few people aside to work on stuff is fine, but we decided to go all in and hire 20 people.

We want AI Labs to move as fast as a startup, not a big company. We felt that watercooler moments and whiteboarding would be important, so the team needed to interact in person. We’re seeing the fruits of that decision already.

We based the team in Ireland because that’s where Learnosity was born and there’s a lot of talent in the AI space here.

NM: We’re an Irish company and we’re proud of that. We want to move faster and we have some of the people who’d worked for us in Sydney and the States now working in Dublin. We have people who are able to deliver real products at scale, who know what an API is, and how to deliver it. They understand our customers, the use case, and they have delivery experience.

We’re combining that experience with the nimble work of a startup. We’re also bringing in people with expertise in a bunch of different industries.

We get the halo effect of having deep expertise in the edtech space and knowing how to apply technology to get outcomes, along with learnings from other businesses.

Q. Can you give us one bold prediction you have for the future of learning? What role do you see Learnosity playing in that future?

NM: I’ve attended conferences for 10 years where everyone always talks about more authentic learning, but they always go back to the same old tests. 

Why? Because they don’t actually know how to deliver authentic assessments. 

My prediction is that AI, and what we build, will act as the key to unlock the shift in education that our customers have been struggling to achieve for a long time.

GC: Learnosity has always been at the forefront of innovation and we’re particularly proud of that. We’ve made a real difference over the last decade and we will for the next decade. 

I always give the example of the NCTM conference in Chicago: when I look around, there are 15 companies there powered by Learnosity. If you’re the 16th company and you’ve chosen not to license from us, that’s fine—but then you’ve got to build assessment technology that’s at least as good as ours. 

That’s the power of Learnosity: we single-handedly raised the bar of what’s acceptable in learning products.

Things change surprisingly little in one year and surprisingly a lot in 10 years. A lot of change is coming over the next 10 years—much more than we saw in the last 10. And when it comes to assessment, we plan to lead that change.

Want to add AI essay grading to your learning platform? Watch this on-demand webinar, hosted by CEO Gavin Cooney and Product Manager Monet Slinowsky, to see the future of AI for learning 👇