Are we learning better than machines?

Machines need thousands of data points, if not more, to learn a simple concept, like recognising an animal, for example.
How many data points do we need? — One, sometimes none! Right?
Are we using this amazing intelligence amazingly enough or are we reducing it by trying to imitate machines?

- Food for thought

Learning is broadly 2 kinds, in my view:

1. Learning information as ‘Facts’: Every time there’s new information, a new fact is planted in the brain; and knowledge improves!
2. Learning information as ‘Insights’: Every time there’s new information, reasoning takes the brain to the bottom of the situation/idea in the information and stores an insight instead of planting a fact; and thinking improves!

We definitely need a combination of both, but we often see people being more biased with one approach over another.

It is important to focus on how we are learning. Here’s why:

In the first kind, as we get used to planting facts, we quickly reach a situation of ‘This is right’, ‘This is wrong’ — a very binary world, just like machines.
This is because either a fact is there in our universe of facts, or it isn’t. There’s no other possibility. A fact will only get added via whatever process we have chosen to absorb facts: Books, Eminent personalities, Popular opinions etc.,

In the second kind, the information is thought-about, understood and stored as an insight, we call this generalisation when machines do it. However, a machine needs 1000s of data points to do this for each task/situation/idea, but a human brain doesn’t.
Hence, when given a new situation, humans are ‘able’ to extrapolate their learnings basis some common characteristics/features/people behaviours etc., from one situation to another, and are ‘able’ to adapt. We wouldn’t need to experience or fail each situation/person when we can extrapolate our learnings/insights.

All of us assume that we are in the second category.

But then, for example, do you think ‘Racial discrimination’, ‘Feminism’, and ‘LGBT rights’ are different ideas? Even worse, do you think one of them is okay and others are not?

Aren’t they all about equality and impartiality at their root? Why did we need one or more examples for ‘each’ situation/idea to accept that idea?

If people are able to learn thinking and not facts, then we wouldn’t be needing to fight for the same thing — such as equality — in different forms, generation after generation, and teach each idea (as a new fact) as it comes along.

And here we are, endlessly debating about machines ‘learning to think’, ‘becoming self-aware’, and ‘taking over humanity’; when we are imitating machines in their current state, ‘learning facts’.

Feeding on facts seems simpler. But don’t you think, by introducing a little complexity in our brain to think and reason, might reduce the complexity in our lives and society?

Originally published at https://my-ink-life.com on December 13, 2020.