Should we be scared of AI?

The most frequently asked question from a recent session on AI was this-Should we be scared of it?

The short answer is yes. An AI has the potential to self learn and if we develop it without setting it’s limitations then we are setting ourselves up for some really bad consequences.

Machines can be better than us in whatever we do. Once they reach that stage and no longer think it’s sensible to take orders from us old homo sapiens. It starts to take it’s own decisions. These decisions can be good/bad for us depending on how we trained the AI.

Now the next question that pops up right here is- why be scared when it’s in our hands to train the AI whichever way we want? This in itself the biggest reason to be scared. We don’t know what way of training should be given in order of us (humans) to be safe from them (machines).

We humans are masters at making mistakes. Though we learn from our mistakes, don’t we? But AI is such technology where 1 mistake in training it has the potential to decide the fate of humanity on this planet. Even if we learn from such a mistake, it would be too late to make any amends.

Yes we should be scared of AI. Because it’s only when we are scared that we will analyse each step into training an AI with utmost scrutiny and that’s exactly what’s desired.

This is the primary reason why prominent voices of Elon Musk and Bill Gates support regulation of AI development.

AI could be a technology where humans are crafting the end of their ownselves.

If you ever want to work in the AI development domain. Always keep in mind that the fate of humanity depends on this.

1 mistake and we’re all gone. Literally gone.

-ChoubeySahab