a year
For instance, imagine you are a cashier at a grocery store. A customer comes up with ten items. You scan the first item and then put it in the bag. The next nine items you have to scan each item individually as they go into the bag. That's 9,000 times you have to look at that item and make sure it is going into the right bag. Now imagine doing this for 200 customers a day. It's easy to see how humans can be better than AIs at small details like this.
AIs are getting better but not yet good enough on social interactions. For instance, if two people come into a bank and one says "I am here for my appointment with so-and-so" and the other says "I want to open a checking account", an AI might get confused which customer to help first because of these conflicting statements. A human being would immediately figure out what was going on and help both customers appropriately.
For instance, try playing Jenga against an AI and see who wins (or loses). Or try solving a Rubik's Cube against an AI (there is one in competition mode right now on Amazon Echo). Or try playing chess with any computer program ever made compared to Gary Kasparov or Magnus Carlsen or any human grandmaster ever made. Or try writing out a check using a pen vs using an automated check writer machine (which I do every month). Or try writing out numbers using just your fingers compared to an abacus expert or calculator expert or even any adult who went through elementary school math classes (I tried this recently and was shocked by how bad I am at it). Physical dexterity is hard for computers but easy for humans, especially when we learn as kids how to use our hands well.
For instance, take two images from Google Images and write a short story about how they relate or don't relate (try it!). This is very hard for computers but very easy for humans since we were taught how to write stories in elementary school when we were learning how to read and write all those words that make up stories ("The quick brown fox jumps over the lazy dog"), etc.. Or take some random objects from around your house and make up a story about them ("This glass of water my grandmother gave me before she died has been here ever since waiting patiently until I finally realize its true love - another glass of water just like it"). This is also something computers will never be able to do well unless they start reading millions of books first in order to learn what creativity looks like in literature so they can mimic it in other areas like art or design or music or whatever else they want creativity in their products/services/experiences/etc.. But creativity isn't just imagination - there's also logic involved which makes creativity even harder than just imagination alone.
In chess, there are 400 possible moves after every move by either side (each player gets 25 moves), 1900 possible moves after both sides have made all their moves, etc.. In Go there are more possible moves than atoms in the universe (over 10^170) but each player only gets 180 turns so there's still some uncertainty about what will happen since not all possibilities will be explored by both players during those 180 turns given such huge possibilities compared to chess where each turn only has 25 possibilities (and many games end before even half of those possibilities occur). If you want AIs that can play these games well then you need algorithms that
Comments (0)
No comments.