| |November 20189CIOReviewOne of the practical challenges faced by most educational organizations is that the legacy content in the sector is not structured and organizedholder across the student, parent, and mentor for an effective result outcome to happen. Therefore the AI algorithms have to distinguish the unique context of the individual student and decide on one or all of the following actions:-- Where self analytics and action would help the student directly (extra learning notes, concept level feedback, additional questions etc)- How to help parents assist the child (where is the intervention required at a behavioural level)- How to create the feedback loop for the faculty/mentor to assist the student.All of above therefore make the AI in education more complex than an adoption in a pre-sales chatbot or a manufacturing process, and thereby impacting the RoI of such a program being driven by any institu-tion. While the business case for AI is not straightforward to define for most organizations, it is all the more difficult in edutech due to the above cited reasons. It takes an organization with an eye on the future to be committed to this investment. One way to over-come the investment barrier is the need of partnerships between tech ecosystem, traditional education players and with academia to mutu-ally invest in delivering AI for the larger good.New age AI on Old World ContentOne of the practical challenges faced by most educational organizations is that the legacy content in the sector is not structured and organized.Given the width of the scientific terminology across subjects, inter-disciplinary concepts inherent in the content, it is significantly more ex-pensive to get the educational con-tent in shape for AI to kick in. One of the leading consulting organiza-tion estimates that it requires more than 5000 labelled data sets per cat-egory for the AI to kick in, and more than 10 Million data sets for the algorithms to be truly world class. Given the vastness of the academic subject discipline, this leads to sig-nificant investment requirements on the content front, where sometimes the payback periods are pushed out to the future, and therefore the natural resistance to adoption of AI in education.The other related aspect which impacts the RoI and adoption be-yond the legacy content, is that the cost of new content generation is even higher - since the content is not user/machine generated, and further requires high degree of accuracy be-fore it is ready for consumption by the end user. Today's millennials consume the content across multi-ple formats such as videos, e-books, question banks, printed books and the consumption journeys can seam-lessly span from one format to the other, leading to high ongoing in-vestments in structured content across formats. Newer startups are solving for the problems faced above, but are challenged since they do not have large data sets which incumbent or-ganizations have, building yet anoth-er case of why collaboration across traditional organizations and newer generation startups makes sense for building strong AI led platforms.The Human Factor in TechnologyIn a world where parents are already struggling with their time with their children, it is very easy to let go and trust an algorithm to deliver `per-sonalized' learning. It is therefore the responsibility of every edtech player to ensure that the technology is built on the right academic mod-els. Wrong biases in the AI models can do more harm than good to the child, and can have far more impact, than in the case of a poor algorithm leading to a missed sale by an AI chatbot.On the other hand, it is impor-tant that the parents are not enam-oured by fancy animations, glossy marketing and choose academically sound, technologically enabled plat-forms, and further understand that they are a critical part of the learning path for their child.Finally we need to be conscious of how hierarchies between student, parents and faculties would change as AI delivers higher knowledge di-rectly to the new age students/par-ents, while the traditional faculties need to reinvent themselves to stay relevant in the shifting teacher-stu-dent equation.
< Page 8 | Page 10 >