Keynote "What AI can (and can't) do"

Share it with your friends Like

Thanks! Share it with your friends!


CeBIT Global Conferences - 23 March 2017: Keynote "What AI can (and can't) do" / Toby Walsh, UNSW Australia, Data61, TU Berlin


Garantia COBIA says:

In regards of artificial intelligence the most we can aspire to have is: "Intelligence simulators", don't expect automatons to develop real intelligence to solve complex problems humans can't, develop self awareness or consciousness nor to invent new things as humans do, you can't program anything you do not understand!

JamesPossible says:

this is beautiful…"ALL ALGOS WILL HAVE A BIAS!" people must understand this.

Dennis Kibler says:

Here are few things that AI has had no success at. Writing poetry, worthwhile poetry not some nonsense. Proving an important mathematical theorem. Teaching anything that requires more than a rote answer. Currently I think AI has reached the level of intelligence of a bee, but that may be an overestimate.

Teh Jokur says:

AI will be able to do anything. It's just a matter of time. And there will be the day when an AI "wakes up" which will be the day mankind will understand what conciosness actually means and I hope I will live until this day comes.

ghoulfish says:

17:54 Shows SKYNET logo, claims AI won't ever take over the world.
19:02 Next slide. Shows something looking very much like a prototype of a T1 Battlefield Robot and talks about not opening the Pandoras Box

Wayne Lee Ramsden says:

i think employment and there for money will be on its way out once ai takes over.then it will be a resource based economy..isn't that the venus project in a nut shell.other wise it will be a very unequal society.even worse than we have today.

NotImportantJoshHarris* says:

Is technology morally neutral?

NotImportantJoshHarris* says:

"Good ends, not for bad ends." What is good?

NotImportantJoshHarris* says:

History lesson 101:

British Scientist says, "Not possible at all."

It happens the next day.

Write a comment