Why AI might not be that bad!..for now..
These days, Artificial Intelligence is the buzz word! Everything now revolves around this new term that was ironically coined in the late 1950s.
Artificial Intelligence is about making smart machines (period). This would mean that A.I is just about making machines tackle tasks more efficiently or perhaps even do things that even humans (can but would need time)!
However, recently many individuals have been mentioning, how A.I can be dangerous for humanity. Even people high up in the industry have comments to say about A.I like,
There is great danger in training an AI to lie
— Elon Musk (@elonmusk) December 24, 2022
And that’s true. If A.I reaches a point where it is capable of deception. We’re done for!
But don’t fret just yet! Artificial Intelligence, like I said, is all about making smart machines. Even the most advanced projects such as ChatGPT are still miles away from the real expectation of an A.I. This comes down to how most Machine Learning (another fancy term that brings up hype) projects learn patterns or adapt.
Most M.L models have a simple take to it. It consumes data, performs some computing and then throws out a response which when successful would be useful. This could be considered as the fatal flaw of the system. Since the system is only equipped to learn based on the limited dataset given to it, the model is unable to perform as well as a human since it lacks experiences which a human would’ve during their lifetime.
But again, we are inching closer to the finish line and we’re charging towards creating the single most powerful A.I. and ChatGPT and other A.I. projects will definitely contribute towards that goal.
Hope this was a good read, it was my first!
Cheers