The Artificial Intelligence Channel
Recorded June 29th, 2018
Shimon Ullman is a professor of computer science at the Weizmann Institute of Science, Israel. Ullman’s main research area is the study of vision processing by both humans and machines.
https://www.weizmann.ac.il
Source
Why is it that as soon as one acquires the title "professor", a requirement to clearly define ones terms ceases to be expected? This seems to be common practice throughout both the scientific and academic communities. Surely this is an indictment on the students also who are lacking something, not least, the wherewithal, or perhaps even a spine, to challenge authority.
מרתק!
Beautiful Presentation… thanks for sharing!!
"As people we certainly don't need to look at millions of pictures"
I doubt that. If you consider the amount of samples per second we get with our eyes it would probably add to a huge number before a newborn learns to recognize something and not confuse with things that looks similar. Except maybe some hardwired things, like faces.
It's profound that humanities highest IQ's (But not a lick of common sense) have been, are now, and always will be hellbent on inventing machines and weapons that will ultimately destroy them. They even know and do it anyhow, not unlike the lower IQ variety that smoke cigarettes that have "These stinking things will kill you" written on the package for the past forty years.
Human beings are the most advanced robots ever invented, exercose,diet,training, and educating them is far superior than wasting time and resources trying replace them work gives people value,they dont need to be replaced.
Back in 2006, I hacked Google Translates public page and used the translation service to create a real time language agnostic chat (2 or more persons).
I was just doing it because I wanted to find a way to chat with my then girlfriend in her native language without her needing to type English.
It worked but was slow….today I imagine the neural models for language translation would allow what I was doing then (which used Google's old statistical language translation methods) to be far more efficient.
I got really excited about the wider possibilities and sought to gain a commercial license from Google …at least ask about it (Peter Norvig is a Facebook friend) but there was no commercialization plan when I built my hack and Google wouldn't release a formal translate API until a few years later.
I still think freedom from language barriers will be a massive social state shift for human society. I wrote several posts about why during those years….this time is coming now again thanks to machine learning translation models. The AgilEntity framework ( http://apriority.dyndns.biz/aesplash.html ) that I built this TOAD algorithm into (translations on agent demand) is proprietary now but I am engaging the process of open sourcing the platform. It could easily use the existing google translate API (which now uses the neural models for translation) enabling any one implementing solutions with the frame work to plug in this type of advanced many to many language translation.
http://sent2null.blogspot.com/2008/09/will-internet-make-more-languages.html
http://sent2null.blogspot.com/2010/05/story-of-language.html
http://sent2null.blogspot.com/2010/06/communication-as-promoter-of-human.html
http://sent2null.blogspot.com/2011/07/diversity-saves.html
China are absolutely serious and committed on A.I and the tech industry to push itself on the world stage. Quantum computing and G.A.I are going to be the key drivers for any country serious in economic growth.
Great presentation
Why so much name dropping by these experts in these presentations? I’m fed up with hearing about Einstein &Co, Turing, etc. Let’s move on shall we and the presentation can save us some time!
AI is great, and it is a very interesting subject !
I just want to tell how I feel about AI in general.. at least about AI that exists in its current form.
if we don't care that AI is not self-aware or really understands anything then AI is absolutely great!
Intelligent, what does that mean? Does intelligence mean that you are aware that you exist? or that you can think?
A computer just does what it is instructed to do but really fast. But does it really 'know' what it does?
it reads an adress in memory, executes the command, writes it back to another part of memory but really really fast.
If a computer 'speaks' does it really know it talks? ……it just sends data in a particular order to a speaker at 44100 times a second.
It isn't even aware it exists. it isn't alive. it doesn't know why it does the things it does.
It is an illusion of human like behavior, but it doesn't really know that it is doing something funny or clever or fast.
We assign or project those human qualities onto a computer when it does something. We think that its amazing, funny, clever, sad or whatever.
But nevertheless, A.I. is still amazing in my opinion, but it isn't really like I think AI should be.
Meaning that a machine can just as easily calculate 1+1=2 or kill someone. it does not understand that there is a difference between calculating and killing … if it is programmed to do one or the other, it is just doing it, without hesitation, without any feeling whatsoever and not even realising it is doing it.
If you think it does know then a hammer gets a headache if you use it to slam a nail into a piece of wood.
The point is that it does not have any concept about what it does is good or bad, right or wrong, funny or sad.
It just does things not even realising it is doing it. it doesn't think. it isn't aware of anything.
Even Siri, no matter how human-like it sounds, it doesn't truely understand anything. it is an illusion. A good one and entertaining, I agree.
Anyway, maybe I expect too much of A.I. I just want it to be aware, understanding, knowing about what its doing.
Boycott, Divest, Sanction (BDS) israel for their apartheid practices, war crimes, economic expansion into stolen land, and especially for their policy of murdering academics including most recently Fadi Al-Batesh.
Well was not Turing but Weiner that proposed the modern computer design.
However they are very much alike ferengi.