Comments
Machine Learning's Greatest Weakness is Humans
View Comments: Newest First | Oldest First | Threaded View
Re: Moral and ethical decsions
  • 6/28/2017 3:35:35 PM
NO RATINGS

Probably not unlike asking your car mechanic to make your dinner.

Re: Moral and ethical decsions
  • 6/28/2017 3:26:02 PM
NO RATINGS

What we may discover by programing A.I. that what we thought was illogical or irrational behavior in humans actually does have a purpose.  

I think much of what is illogical or irrational in humans comes down to a lack of programing or experience in a certain situation.    If you have a robot programmed as a mechanic and then ask it to make you a salad it might behave in ways that seem downright weird. 

Re: Moral and ethical decsions
  • 6/16/2017 11:37:56 AM
NO RATINGS

Either we make computers act more like humans, or we will have to improve humans.

Re: Moral and ethical decsions
  • 6/16/2017 10:28:11 AM
NO RATINGS

It almost seems that there will have to be built into machines those good and bad days, or at least some randomness of active to simulate how human minds are flawed and either not rational at times or at least do maybe predictable behaviors but we just haven't figured out yet to predict them!

Re: Moral and ethical decsions
  • 6/14/2017 12:13:36 PM
NO RATINGS

To Lisa's point, the technology is not intrinsically nefarious it can become nefarious in the hands of those that use it for nefarious purposes. Very much the same with the hacking incidents that we have seen that use harness technology for the purposes of bringing down a system. AI  will grow up and have its challenges but I don't think it's as scary as many are predicting it will be.

Re: Moral and ethical decsions
  • 6/13/2017 5:09:14 PM
NO RATINGS

..

Lisa writes

There's also the dark underbelly, meaning the use of AI for nefarious purposes.  I don't think that issue is being addressed adequately either.  

Definitely. I think we can be pretty certain that in today's world, with contestants ranging from the powerful military establishments of technologically advanced nations to fanatically obsessed and dedicated terrorist movements, some pretty darn unsavory implementations of AI will be arriving soon. Actually, I think they're already arriving ...

..

Re: Moral and ethical decsions
  • 6/12/2017 3:21:48 PM
NO RATINGS

One of the advantages would be that machines typically don't have good days or bad ones.

Re: Moral and ethical decsions
  • 6/8/2017 7:02:02 PM
NO RATINGS

AI is being built into so many things.  Vendors and some of the big consulting firms are moving extremely quickly, confident that everything will work as planned.  Frankly, I think we're getting too confident too soon because we're running so fast.

There's also the dark underbelly, meaning the use of AI for nefarious purposes.  I don't think that issue is being addressed adequately either.  

Generally, I think more critical thinking needs to be done.  Looking at the potential benefits only is naive, in my opinion.  Naysaying isn't going to stop the train.  

As my favorite saying goes, "Think.  It's not illegal yet."  :)

Moral and ethical decsions
  • 6/8/2017 6:20:42 PM
NO RATINGS

"From an analytical perspective, I find all of this fascinating because human behavior is linear and non-linear, rational and irrational, logical and illogical. If you study us at various levels of aggregation, it's possible to see patterns in the way humans behave as a species, why we fall into certain groups and why behave the way we do as individuals."

 

This is truly the issue for AI. AI will be very successful in linear decision making and repetitive actions that are predictable but when it comes to getting judgement or moral decisions like driving a car into a school bus or a tree, AI will still stumble. It has its place but not everywhere.



INFORMATION RESOURCES
ANALYTICS IN ACTION
CARTERTOONS
VIEW ALL +
QUICK POLL
VIEW ALL +