Impact of Generative AI on EdTech
Interview with Apratim Purakayastha (AP), CPO and CTO @ Skillsoft
Apratim Purakayastha (AP) is the Chief Product and Technology Officer at Skillsoft, where he built the industry-leading Percipio learning platform that was instrumental in Skillsoft's June 2021 IPO. Previously he was the Group President at ACI Worldwide where he led a large SaaS payments business. AP has a customer-centric approach to drive innovation in product, process, and organization.
In today’s episode, we discuss:
The transformative potential of generative AI on the education industry
What are the potential benefits and risks
How Skillsoft is taking advantage of this new paradigm
New products Skillsoft is experimenting with
How to overcome adoption challenges
The role of leadership in driving change
What follows below is a condensed and lightly edited version of our interview.
AP let's start with a broad point of view. How do you view generative AI? Is this a new paradigm or more of an extension of AI in general?
Apratim Purakayastha: Mustafa thanks for having me here today.
Personally, I believe that this is a game-changing paradigm. Perhaps as big as your personal computer, mobile computing, or smartphone; may be even bigger.
It has the potential to change every aspect of how we work, how we live, how we create. It has the potential to change every aspect of what we do.
There are, of course, risks and concerns with any new technology, like with any new disruption. But it is also a massive opportunity. Applications like Chat GPT unlock more and more incentives for innovation. And what we are seeing today is just the tip of the iceberg.
I could not agree with you more. This is a completely new paradigm. And if we take on that point of view, what do you think is going to be the impact on your industry? And more importantly impact on Skillsoft?
AP: I think the impact on the education sector, and by that I mean all the way from K-12 to corporate learning, can be encapsulated in a few concepts. Generative AI will change what we teach, how we teach, and how we work.
So for example, at Skillsoft, we are changing what we teach. In addition to releasing new courses on generative AI, we are incorporating generative AI concepts throughout our entire catalog. From creative writing to marketing to customer service or even sales, there are so many potential use cases for generative AI that people will need to learn about for their roles.
Next, how we teach will also change. At Skillsoft we have a platform for our corporate customers. Generative AI now allows us to make that platform even more interactive and intelligent. We’re exploring new generative AI implementations in areas such as automatically creating course descriptions, personalizing content packages, automating the generation of assessment questions to test learner knowledge, and coaching and mentoring scenarios.
And as for how we work, I think that is where this technology is most disruptive. With AI, humans, and machines work closely together in partnership and become more intermingled. For example, in customer support, AI may answer an initial service request and then bump it to a human if needed. Then the human could bump it back to AI after adjusting a prompt, and so on. This example works in other functions too, like marketing, sales, product management, programming, and cyber security. We are beginning to experience some of that right now at Skillsoft.
I like the way you paraphrase it into three buckets - what you learn, how you learn, and how you work. Makes absolute sense. And I agree the biggest disruption will happen in the third bucket of how you work. Which will then drive changes to the other two.
With that being said, I am curious about how your company is looking at generative AI. Is it a threat or an opportunity?
AP: It is both (smiling).
If we act upon it then it is an opportunity. If we do nothing, then it is a threat.
Personally, I see this as a huge opportunity. First, it is going to help us increase the production volume, speed, and scale of content. Now we can use generative AI to build the first draft. We can expand to different formats as well. For example, videos are hard to produce. They take time and money. With generative videos, we can be fast and save money for a large class of courses. Generative AI is going to help produce and check the quality of code faster, it will help product managers write epics and stories faster.
And that is just the tip of the iceberg. We see the opportunity of delivering a much more fulfilling experience for the learners. An experience that threads instruction with interaction much more finely and seamlessly.
And what about the risks?
AP: I am glad you asked that question because there are quite a few.
First, we can't over-index on AI. Just like automation AI also needs humans. There needs to be an escalation path from generative AI to a human. For example, for a lot of our learners, AI can be the first line of defense for Q&A. But eventually, they might want to talk to a human mentor. Too many ideas I see today are about removing humans. And that is a big risk. Humans will co-exist will generative AI, which can empower them to perform higher-level tasks at unprecedented speed and scale.
The second is around security, which applies not just to edtech but all industries. Specifically around data privacy, including questions like what information is being shared and is confidential, how is the data being used or stored, etc. These are big questions and they must be addressed before generative AI is broadly adopted.
Third, is bias. We see it in our regular AI models already – but will those biases carry over when we are interacting with Generative AI? After all, the large language model is a reflection of our society. And unfortunately, we as humans all have biases. The question becomes how do you counteract this bias?
And finally, ethics. Are we using AI in an ethical way? Or is it nefarious? From deep fakes to creating new ransomware applications to its impact on cyber warfare, in the future, we may use AI to fight AI. That is where ethics plays such an important role.
I am curious AP, where are you at Skillsoft experimenting with AI? Are you building new products or trying to integrate them into your existing product line?
AP: A little bit of both actually.
I briefly mentioned this before, but one of the first things we are doing is creating generative AI courses and learning journeys for our customers. We see this as the first step. Our customers need this and they are asking for it.
Second, we are looking at our catalog and thinking about how generative AI might transform the curriculum entirely. For example, this could be a brand new professional writing course that teaches our students how to leverage generative AI for their work. The same could apply to courses for engineers, product managers, sales, etc.
The third is to enhance existing offerings like coaching and mentoring and take them to a new level of interactivity and personalization. We can take advantage of Generative AI to produce assessments at scale so that one-day assessments are personalized stressing areas of improvement for each learner individually.
And this is just a sampling of the experiments we are running. Talk to me in a month or so and that list is going to double (laughing).
Now that is within your existing markets. Do you see expansion opportunities, or going after new markets?
AP: Skillsoft’s greatest strength today is that it produces original instructional content as opposed to aggregating content in a marketplace. This allows for authentic, sound instructional design with normative assessments. With generative AI, original creative content can approach the speed and scale of a marketplace. Marketplace creators will not stand still but after a certain critical volume differences become smaller.
Second, is geographical expansion. Now with this new technology, we can get very sophisticated with localizing our content. This allows us to take our core corpus of content and make it available quickly in longtail languages. And then it's up to a distribution model that is scalable to address the longtail markets, which when you sum up is actually pretty large.
Now AP, as you upgrade your existing products or go into new markets, where are your biggest challenges?
AP: A few come to mind.
The first is internal awareness and education. While many employees are aware of generative AI – and there are a few pioneers who are actively using it – not everybody is living and breathing generative AI. They’re still focused on their day jobs. So for me, the challenge is how do I make this more personal to them and their roles, so that they can benefit from it, and in turn, we can improve our productivity. At the organizational level, we’re hosting things like Town Halls, hackathons, and demonstrations focused on generative AI, to further drive employee awareness and participation.
The second is around change management. Once people are aware and educated, we need to change the way we work. And that means new processes, operating models, workflow, etc. Basically everything. And that is a massive amount of change across all our departments. Change management includes training in the areas of safe use of private and confidential data, ethical use, protection of copyrights, etc.
The third is mobilization. And this is more personal to me. Which is how we integrate this into our existing development roadmap. We all have plans and deadlines that we’re committed to meeting. If we throw in new items, how does this impact everything else? This is where I am spending most of my time. Ensure that the team is realigned and that we are focusing on building the right thing.
AP this has been great. Thank you for being such a great guest. We will have to reconnect once you are ready to showcase your new product.
AP: Mustafa most definitely. Thank you for having me. This was a lot of fun.
If you enjoyed this, please turn that ♡ into ❤️. This will help others find this newsletter. Thank you so much 😁