"It is important to be really clear on the problem": Pearson’s Milena Marinova on AI in publishing

"It is important to be really clear on the problem": Pearson’s Milena Marinova on AI in publishing

This July, Pearson appointed Milena Marinova to lead artificial intelligence (AI) and personalised learning for the company - the first such appointment in the educational publishing space. 

Marinova is a technology executive with over 15 years in business and product development, strategy, marketing and venture capital, and has extensive knowledge of machine learning, neural networks, computer vision and the commercialisation of new technologies. She joins Pearson from Intel, where she led the development for commercial applications of AI; she previously held executive roles at Idealab, a prominent startup incubator, and Hyundai Capital America.

Pearson has already integrated some machine learning capabilities in areas such as authentic assessment (developing feedback ​capabilities ​for marking complex mathematical activities), automated essay scoring, adaptive learning (catering to each student’s individual level of ability) and intelligent tutoring (working with partners to integrate AI capabilities into English language learning in China and incorporate a virtual tutor into products in the US).

Marinova is tasked with expanding these areas, as well as creating innovative new AI-led applications. We sat down with her to discover her insights into AI and publishing so far.

Your background is in tech and venture capital. Why have you decided to move into the education & publishing industry now?

I believe education has the potential to benefit hugely from digital disruption and advanced AI techniques. The opportunity to utilise AI to improve the learning experience, and ultimately to better prepare people for their career in the future, is an incredibly exciting prospect.

What are you working on right now?

Right now, I am working on developing human-centric AI – this means making the learning experience better for students and teachers; enabling lifelong learning through more accessible and affordable products; and building better products and solutions using new technology.

What are the biggest opportunities you see for AI to improve educational publishing?

Technology is changing radically how we live and bringing many benefits. How we learn will be no different. Pearson’s digital transformation is helping more people develop the skills they need to prosper. Pearson is well positioned to engage with millions of learners across the world because of the significant reach of its platforms and its deep expertise in developing world-class content and assessment.

We can take advantage of AI techniques to help students learn better and improve outcomes. We are also reducing mundane tasks, so teachers can spend more time with students. This not only generates better learning for students but also creates more engaging teaching experiences for instructors.

What are the biggest red herrings?

In general, the biggest red herring is having unrealistic expectations of what AI can do. This distracts companies from the valuable areas where AI can be deployed and applied usefully.

What’s the most common mistake companies make when looking to harness AI?

Many companies don’t have a deep enough understanding of how to solve the problems they have. It is important to be really clear on the problem in order to identify the right AI to solve it. In addition, the problem must determine which AI algorithms are used, not the other way around.

At Pearson the problem we are trying to solve is clear: how to make learning experiences better for students and instructors. We will do this by continuing to ensure that Pearson’s products and services deliver improved outcomes for more learners around the globe.

You’ve only been in the role a month - but what do you think the publishing world could learn from the start-up scene?

One key learning is to focus on speeding up the ‘go to market’ process; developing an agile approach to product launches, from prototyping to commercialisation.