Keynote "Artificial Intelligence: Where are we at and where are we heading?"

In time 18:55 - It is not Miro style, it is Kandinsky.

but, what is left for us to do? what is not a process? hm....

[@UserName________]

Nothing is impossible.

[@Brainbuster]

A round square is impossible.

[@Brainbuster]

@@heribertojuarez6046 Maybe one day you'll find a round square if you keep believing.

[@hakusansaku8800]

So now every company on the planet will hire people who can use TensorFlow, etc. ? To gain an edge? Even though AI software is open source, you still need several millions - billions to facilitate that power. As he also said that you need access to big data and to high performance computing.

[@kadabra8268]

right black hat hackers will love ai tec, like you said the cost to power that is un real

[@SoCalFreelance]

Yes, AI has made tremendous gains, however you show a computer a picture and it can't figure out what it's looking at. So still much progress needs to be made before AI can understand its environment.

[@seonteeaika]

The systems are getting better still, but AI has been better than humans at image recognition for over 2 years now. Look up article from feb 2015 with google term: "Microsoft, Google Beat Humans at Image Recognition". I think the reason humans can make errors is because they don't know names for all different types of plants and animals, that's where they will lose to a computer at the very least.

[@SoCalFreelance]

While I would agree the use of hundreds of thousands of internet based images to conduct pattern matching (as seen in ImageNet Challenge) is a positive step, ultimately an AI needs orders of magnitude more computational power to truly understand what's happening in the picture or video beyond just object recognition.

[@artemkovera5785]

Hello. if you're interested in artificial intelligence, you can check out my ebook “HOW TO CREATE MACHINE SUPERINTELLIGENCE”. This e-book is available for FREE in Amazon Kindle Store until 2 April 2017. It includes the following topics: Intelligence as information-processing system Classical and quantum computations Artificial Intelligence and Machine learning Artificial Neural networks

[@entyropy3262]

12:57 "teached" probably the most common german mistake I tend to do myself aswell :D

[@bernios3446]

I understand: "I teach the students...." - he seems to use the present tense. He sounds good enough to use "taught".

[@rubenontiveros6516]

Enty Ropy ,

[@christopherdiedrich40]

@@bernios3446 I think you ought not to be taught, lest it has been brought up before. Do you know what I mean?

[@orca2162]

The host is out of touch?

[@MonaGee98]

When the Speaker shown the chart of History of Technology and said the history of human kind, Asia and Asians are no where to be seen. What credibility does the speaker has?

[@i.C.Infinity]

It's really funny that at the beginning of his talk he said people can't get a grip on the exponential growth curve, then he finished with a prediction that totally ignored the exponential growth curve.

[@brodrock]

Not ignored actually. The chart is designed with an exponential Y-scale so the advancing straight line actually represents exponential increase.

[@i.C.Infinity]

In that case I stand corrected. Even though i'm not going back to look for it.

[@Brainbuster]

You should've marked the time in your first comment.

[@maxsvensons4438]

Just look for rice on a chessboard story. It says it all. But take in consideration that the chessboard ends after 64 fields, but does the development do so? We just can't get in and most people aren't even aware of the problem or do not see where this leads. We are still living in the past and not adopting. Very dangerous.

[@tottisempre]

If you paste these values in Excel and add a scatter chart you will see it is exponential: 2016 1.00E+09 2018 1.00E+10 2020 1.00E+11 2022 1.00E+12 2024 1.00E+13 2026 1.00E+14 2028 1.00E+15 2030 1.00E+16

[@rhondamay1772]

#FakeLeftIsAltRight ;-)