Dear Loretta,

You are right thinking that everything with AI is programmed. More and more complex such AI programs are being developed, they are complex in relation with the task they are meant to accomplish.

The downside of complexity is that the more complex they become, the more lengthy and complex their training becomes, because they need to be trained into providing the right solution before being put in service. Simple 1 or 2 layer neural networks are easy to program, but starting at 3 layers, heavy training is required. New very efficient deep learning techniques have now been developed to help the process.

To get an idea of these techniques, you can have a look at this book:

http://www.deeplearningbook.org/

Contrary to urban legend fed by science fiction stories, there is no chance whatsoever that any AI program ever becomes sentient, or autonomous, for the simple reason that they are not alive.

Looks like they will forever miss the emotions required to become sentient, because it doesn't seem possible to simulate a limbic system such as is causing us to be sentient. This recent paper explains the function of the limbic system in causing us to be sentient:

http://www.scirp.org/pdf/CE_2019022016190620.pdf

Best Regards, André

Hi.,

Artificial Intelligence (AI) began in antiquity, with myths, stories and rumors of ... This work culminated in the invention of the programmable digital computer in the 1940s, a machine based on the ...... AI researchers began to develop and use sophisticated mathematical tools more than they ever had in the past.

kindly check this link:

https://www.nap.edu/read/6323/chapter/11

your idea is very correct!

Dear Loretta,

" If so will A.I. or segments of it be able to become autonomous."

We have already autonomous facilities like coffee machines. A.I. pretends to be replacement of humans.

Preprint ARTIFICIAL INTELLIGENCE EVOLUTION

Please read the article

Preprint THE PHILOSOPHICAL STUDY OF INTELLECT EVOLUTION

Best regards,

Konstantin M. Golubev