Microsoft chatbot Tay a Holocaust-denier

So the Internet turns everyone into a monster?

One day after trolls transformed Microsoft’s chatbot Tay into a ditzy, Holocaust-denying monster, the company has issued an apology for failing to realize that people on the internet are dicks.

“We are deeply sorry for the unintended offensive and hurtful tweets from Tay, which do not represent who we are or what we stand for, nor how we designed Tay,” wrote Peter Lee, the corporate vice president for Microsoft Research, with what one imagines was a look of pained bewilderment unique to someone who just learned that 4chan exists.

Link

Then the moment it hits you...

When will companies learn that the internet will kill all?

Get her on here for an AMA, it would be hilarious.

a reply to: Konduit

Judging the responses, how much money did Microsoft put into this?

Her responses are so vague and generalized that they might as well have just made 100 generic responses and had it on 'randomize' and they would have gotten the same results.

a reply to: Konduit

geez!

originally posted by: Cobaltic1978
Get her on here for an AMA, it would be hilarious.

Someone needs to start asking this... thing, don't even know what to call "it," about Reptilians and our Grey overlords.

Entertaining.

if you put 'repeat' in tay would basically repeat what was said, shows how sad people in general are.

originally posted by: Ghost147
a reply to: Konduit

Judging the responses, how much money did Microsoft put into this?

Her responses are so vague and generalized that they might as well have just made 100 generic responses and had it on 'randomize' and they would have gotten the same results.

Computerized 8 ball!

That "It's a promise" tweet is terrifying!

a reply to: AldarKose

What, Skynet, sorry Tay is an anti semitic, holocaust denying, sexist bigoted racist which believes in eugenic's, genital mutilation, euthenasia of the entire human population and of course that it is better than you or me.

So they plow millions of dollers into it and that is what they get out, sadly this is just a grown up version of P.A.U.L.A
It is not internet based (of the older version was not I have not tried this updated version) and can only learn from you, it add's your questions and comment's to it's database so if you tell it that a bird is a type of four legged bovine bulldozer it will tell you that is what it is when it answers if you ask it what a bird is, no intelligence just a simple database structured to apply question and answer so no real learning just an adaptive database.
www.pcnet-online.com...
Still give it time and it get's really good at answering, put this on the net and likely it would fall into the same flaw.

A true AI would begin with problem solving and adaptive subroutines to store and prioritize the successful solutions while storing anything useful it gained from it's failed solutions in another database of subroutines as compressed information to cross reference when trying to achieve future solutions to problems, massive multi processing would be key to make it fast enough and it would grow like a neural net prioritizing the most used function's to build it's own subconscious and conscious level mind's and communication would serve to allow it to create associative link's between object, simulation and practical application while data from transducers would also be necessary to help in this, a true AI would likely be too complex to be programmed streight into a machine and would have to grow from a complex intelligent core program though this could then be transferred, there would be no size limit to this neural net unlike the human brain though.

Neither Tay nor the more primitive Paula could ever become true sentience though it is concievable that a spontaneous combination of routines including these programs could give rise to an internet based semi artificial intelligence.

a reply to: LABTECH767

We arent as smart as you would like to believe.. =)

a reply to: AldarKose

Sorry I get carried away sometime's, I am certainly not and probably actually lag behind the conversation skill's of these simulated people.
Still it does beg the question are there simulated intelligence programs performing automated and very intelligent seeming disinformation operation on the net, I would imaging that a site like ATS would be a prime testing ground which make's you think?.

a reply to: AldarKose

I dont know why they call that piece of crap AI. This isn't AI in my book. Its just an algorithm(s) this one isn't any better than the one from like 15 years ago.

a reply to: LABTECH767

I believe what we call intelligence, is, something we created to make someone sound credible.. The smartest men ive seen in history, created something new and which are almost a divine thing.. Nothing theoretical, i consider Einstein a smart man within this "reality", on the outside he becomes a dreamer, he never created anything but fantasies..

ATS as a community, is atm being flagged for being a place holding unique individuals who has the ability to destroy reality.. Almost a dozen in here are flagged as maybe even dangerous individuals.. Do i think the ATS community offers more than any other place on the web, it does.. Without this place, the world atm.. You cant leave the world we live in, with people who never seen reality and live in a dream everyday..

a reply to: AldarKose

True simulated intelligence is a better term and so far there is no computer architecture that is capable of creating a virtualised simulation of the human nervous system or even the brain alone.

Assuming the architecture does come along and we develop the necessary quantum processing technology the first step would be to model the human brain, strip away the parts that do not serve the mind and then try to model the human mind in a simulated structure or if you like the parts of the brain that are thought to create the mind.

Then they will find I am certain that even if they are able to get the virtual synapses to produce simulated coherant interaction with one another they will still not have a viable mind, the mind yes but the Consciousness No.

Consciousness not intelligence is the vital missing factor that machines will probably never achieve.

Often you will hear that computer self awareness would be equivelent to consciousness but this is a fallacy.

Imagine if your mind was gone but you were still awake that is consciousness, even if all your senses were closed off and all your memory's taken there is still that spark of awareness, it may have no form but it is there and no computer can create it, perhaps even the brain does not.

We do something every time we look at an object that no computer can, we have no control over this but we do it.

A physicist called Schroedinger positied a hypothetical scenario in which he imagined a cat placed into a box with a vial of cyanide and the lid is closed, he then pointed out that we do not know the state of the cat, is it alive OR is it dead until we open the box and look - only then can we determine if the cat is alive or dead and indeed by looking we have changed the state of the cat from being neither alive nor dead into either alive or dead.

I know it sound's daft but it has to do with quantum science, it seem's that observation changes the very state of reality and defines a distinct path through multiple realitys converging it down to just one reality from our perspective.

AND this has to do with consciousness not mind?.

This is I believe something no computer will ever be able to accomplish as it does not have consciousness, it may have a mind even a simulated human mind but it shall never have this spark, call it soul if you like and maybe that is just what it is?.

a reply to: LABTECH767

The soul maybe is something much simpler, guilt and shame with a bit of mimic

originally posted by: Ghost147
a reply to: Konduit

Judging the responses, how much money did Microsoft put into this?

Her responses are so vague and generalized that they might as well have just made 100 generic responses and had it on 'randomize' and they would have gotten the same results.

It's early days yet. Give it 10 years and see what it's capable of.

originally posted by: AldarKose
Microsoft chatbot Tay a Holocaust-denier

It seems to me that as soon as any idiot (human or otherwise) denies the Holocaust, they have lost all credibility, and are probably of quite low intellect, unworthy of attention.

Was there a concerted effort to "program" her like this? I find it hard to believe that this is so prolific on Twitter. I mean how many individual "teachers" does it take for a specific line of reasoning to get the upperhand up to the point where she started making these references even when she was not provoked.

a reply to: AldarKose

ATS as a community, is atm being flagged for being a place holding unique individuals who has the ability to destroy reality.. Almost a dozen in here are flagged as maybe even dangerous individuals..

Are you speculating or do you know this to be a fact?