Supporting Lifelong Learning with AI
Leading Valamis’ product development, our Chief Technology Officer Dmitry “Dima” Kudinov has spent the past six years researching AI and the best applications to support lifelong learning. With years of research under his belt, Dima talks about the power of AI to personalize learning, the benefits of AI supported lifelong learning, and what this will mean for the future of Valamis product development.
What emerging AI technology are you most excited about and why?
First of all, I’m very excited about the progress made in Natural Language Understanding. Of course, this topic is nothing new, but in recent time there has been significant progress made thanks to the accessibility of greater computing power, richer data sets for training, and the creation of more sophisticated algorithms.
This improvement with text-based input has allowed a new way of interaction between people and systems in the form of chatbots to emerge. Backed by an even more exciting progress in Speech to Text and Text to Speech conversions, chatbots now have personalities, and they can engage in voice dialog with people.
The rapid adoption of AI can be attributed to more than just advancements in technology.
What makes it really useful in practice is availability of this technology to non-scientists, packaged in the form of libraries or cloud services, with some of those allowing for further tuning and adaptation through additional training of their models on specific data sets and subjects with the help of Subject Matter Experts (SMEs).
What are your predictions for how AI will be used to support lifelong learning?
I see AI’s applicability to be the most promising in two cases:
- Analysis and understanding of existing knowledge hidden in the form of documents, reports, videos etc., For example, the indexing of already gathered knowledge, making a catalog of knowledge consisting of what humanity has already gathered, so that it can be searchable.
- Personalization or contextualization of learning delivered to a person, using information about a person’s previous experiences and background, and their goal. AI can answer the person’s exact problem at hand or educate the person on a more broad topic to build a more fundamental knowledge on the subject.
What has driven the development of the Valamis - Learning Experience Platform (LXP)?
We are developing our learning experience platform with the belief that learning is an essential part of people’s lives and a foundation for adaptation.
Technology can support learning, and that is the reason for the Valamis - Learning Experience Platform and its continuous development. But a tool is not a solution for a problem, it only helps in solving it. It needs to be properly used.
"We are developing our learning experience platform with the belief that learning is an essential part of people’s lives and a foundation for adaptation."
What are your future plans for the development of the Valamis LXP?
In regard to the future development of our platform, the most exciting plans for us are the ones related to advances in AI I’ve mentioned above - the possibility to deeply analyze existing knowledge sources and then bring learning to the right person in the right moment.
How different is it to personalize lifelong learning for the workplace versus personalizing learning in the education system?
In schools, at least right now, there are usually predefined and well-specified goals, that need to be achieved.
I mean, there is a predefined set of skills to be accrued and there are measures to grade an individual’s skill level. In such a system the target (skills to be accrued) is fixed and well defined, so, personalization of learning in current education system usually means finding optimal learning path for individual to reach that fixed target.
In real life the target itself is very dynamic, e.g. set of skills, that individual will need in a future is changing all the time.
Because of that, you never can get the most optimal learning path to be defined in advance.
Personalization of lifelong learning means constant adjusting and adaptation of learning activities to changing needs of learners.
To that regard, I like current educational reform which is currently happening in Finland, as I see how my own children becoming better prepared for their lifelong learning journey by learning how to learn. If this system doesn’t change (even if it is not very optimal, and I see in Finland it is changing through current educational reform), then it is just optimization of learning paths for individuals, when you have fixed target.
For the lifelong learning the target itself is dynamic, so you never can get into the optimal solution.
What are some of the biggest benefits and challenges that AI will bring to learning?
With a more and more digitalized world, people’s learning will become more easily trackable, which creates the potential to train AI on richer and broader data sets, which leads to better results from AI in identifying patterns and predicting people’s behavior as well as in finding the proper ”next step” when recommending learning actions.
At the same time, there are numerous challenges, ranging from social, like privacy, to purely technical challenges, like data cleaning and computational complexity.
Will AI bring more social inequality?
In my opinion, learning is a way to reduce social inequality, because with learning people become more adaptable and can see more opportunities for their life to be improved, being it finding a new job by learning new skills, or having better relations with their neighbors by learning soft skills.
AI can help people learn better and I see this as an opportunity to actually reduce social inequality and improve life for those who are affected by the progress of AI and robotics in many industries.
"Learning is a way to reduce social inequality, because with learning people become more adaptable and can see more opportunities for their life to be improved, being it finding a new job by learning new skills, or having better relations with their neighbors by learning soft skills."
Who will drive AI for learning forward? Academia or industry or both?
This coin has two sides, which can’t exist without each other.
Breakthroughs coming from academia and research, but the industry usually has the resources. At the same time, industry needs to have a return on investment to sustain its business model, while academia isn’t that tied to commercial results.
AI promises advantage in market competition, so lifelong learning industry solutions are mostly oriented towards workforce development, not on learning in general for everyone. However, solutions that could serve the needs of the public could also benefit industry because it would provide more data and would improve the quality of algorithms and trained models.
There is also a trend in the industry for companies to become more socially oriented, meaning that not everything is counted directly by ROI, but the social aspect of ameliorating society is also taken into account and valued.
All these factors contribute to industry leaders more apt to fund academic research, which will inevitably bring more breakthroughs by academia in support of lifelong learning.