I got a funny brother, Hamid.
It’s good to know Hamid.
Whatever amazing things I’ve learned in life, this strange brother was actually the reason behind it.
Secret, subtle reason.
Sometimes evil reason.
In 1987, he enrolled in a FORTRAN course to crack the puzzle of computer programming.
He got stuck at first assignment.
We used to share the room. After dinner, he asked me if I was interested in learning computer programming.
I was excited.
He gave me a 10 minute introduction and then gave me his assignment to solve.
I wrote a solution. It was slightly different from his solution.
The problem was simple: Given the formula for calculating Force, F=m*a, calculate F, if m=5 and a=3.
This is what my brother wrote:
F = m * a
m = 5
a = 3
But the computer won’t understand and kept printing garbage.
Computers process commands sequentially, so I put the values of “m” and “a” first, then the formula to calculate “F”. Like this:
m = 5
a = 3
F = m * a
Now I have been working on the Programming Puzzle since then.
I went on and tried enrolling in Computer Science but it was off limits for non-Saudis.
I enrolled in a FORTRAN course but I realized it was a very limited language. FORTRAN stands for FORmula TRANslation.
I went to the central library and found a delicious programming book, TRS-80 BASIC With Business Applications.
BASIC provides ability to manipulate more than numbers.
But the problem was how to teach computers to understand Hamid.
What was missing? How could even a child understand things but a computer cannot?
How come even a mosquito knows what it’s doing but billion-dollar machines with AI (Artificial Intelligence) don’t know what they’re doing?
They keep telling us that our brain is a huge neural network so if we keep building larger and larger neural nets (which basically compare two images and tell us how similar or different they are), then eventually, computers will accidentally gain understanding.
Because “intelligence” is a magic to them which comes accidently if we keep building ever larger ANNs (Artificial Neural Nets).
They believe in magic so they proceed like magicians and keep cooking weird stuff that strikes fear and awe in the audience as well as the investors.
So last year, the total investment in AI reached a staggering figure of $8.5 billion.
The most impressive achievement: AlphaGO, a computer that can play GO, a board game.
Does it even know it’s playing a game?
Can it choose not to play or even not to win?
Computers still don’t understand Hamid.
And the scientists are not working on making them understand Hamid.
That’s what’s wrong with AI.
We need to appreciate Hamid’s finding. And work on it as I did since 1987.
Then we will find.