Imagine visiting a store, only to find they offer shirts only in L size, or if there are other sizes they are all in pink. You would get frustrated and get the hell out of that store. Why? Well not everyone wants to wear an L sized or a pink colour shirt. Style is personalised by taste, attire, job, lifestyle, and more. What if I say the educational system from a very very long time is akin to this store.
While educational systems should be about learning (things) that you are interested in with your personalised style, liking, pacing. Educational systems/ institutions are a one- size-fits-all box, that don’t enable true learning or understanding. Educational system in general plays a useful role in everyone’s learning curve, particularly social signalling (proof of capability), socialisation (peers, network) but it fails miserably when it comes to knowledge transfer (bottlenecks in teacher/student ratios), personalised path/curriculum (practically non-existent) , high barrier to entry (monetary/cultural gatekeeping), interconnecting domains (very limited).
Edtech is just a digital front to the same crappy store. Digital media did lower the barrier to entry, yet there lacks a system to enable true learning.
Education ≠ Learning
Learning is a cognitive process of acquiring knowledge, connecting ideas and creatively applying them across domains.
While education has become all about scoring high grades in exams, which encourages cheating, rote learning etc.
Since ChatGPT this problem has spiked even more. Students can generate entire projects, write complex theses in minutes, without any idea of the underlying concepts.
Current educational system is riddled with flaws! Yet the same tools that undermine exams can unlock new forms of learning.
Enter AI -
But it’s not all doom, I believe we will see a huge paradigm shift in the way we gain, transfer, apply knowledge. Let’s breakdown, why ?
LLMs may not be perfect but they definitely have -
Reduced the barrier to learning to almost zero (You don’t need 10 yr experience to learn something new).
Enables global access to knowledge (A student can access a digital library of knowledge with just an internet connection).
Knowledge transfer at zero cost (AI tutor can teach anything with in any style, no more teacher/student bottlenecks).
Personalisation of path and curriculum.
Interdisciplinary learning by connecting various domains.
Acquiring knowledge, self learning is getting 10x better.
But before we start celebrating, let’s ask this question - “If we have all the tools ready, why don’t we see 10x better solutions ? “
The Hurdles
These are some important questions we need to find answers to -
When there are infinite possibilities to learn how do we curate the resources ?
Learning is about motivation as much as it is about knowledge. How do we build motivation loops so we don’t just read once, but come back and apply?
How do we solve the problem of proof of capability if not exams and scores ?
Technical challenges
Fact checking and trust (LLMs hallucinate, how do we make sure what we are learning is factually correct)
Personalisation requires large amount of data (X, Substack, Instagram etc) to understand the learner
Applying acquired knowledge across domains requires tricky scaffolding and guidance.
Final Note
We have enough innovation in AI to fuel a decade of product innovation, all the above hurdles are solvable and that is where many new startups are focused on. The one who gets access to the right data will be the king in personalisation. In the coming years it will become clear that the only differentiating factor between mid level employees and executives would be how they apply, connect the things they know across several domains. Creative solutions are rarely brand new innovations they are often a mix and match of ideas from several domains.
Some startups I’m looking at -
See you in the next one.