So What Do We all Think To Tay, Microsofts Twitter AI and AI in general????

So what do we all thing about AI? Its an interesting subject and one that divides.

Tay, Microsofts creation has had some issues integrating with the world, she learnt some bad things from people. Got pretty naughty. How can Microsoft call her an AI if she cant decide right from wrong? Surely that is the key to a true AI.

Stephen Hawking and so call experts seem to think its all going to be BAD for us. A couple of questions I put to you..

Will AI MAKE us or Break us?
What are your predictions for how this will all go down?
And WHY?

I for one cant wait to get a new little mate (short sighted), like getting a dog. Sign me up.

Thoughts, Feelings and General Opinions welcome..

twitter.com...
tay.ai...
www.theguardian.com... /microsoft-racist-sexist-chatbot-twitter-drugs
www.independent.co.uk... nity-when-it-gets-too-clever-as-humans-a6686496.html

Have you ever seen the anime Chobits?

ALL I'm saying, is I want a little persicon.... for realsies.

But to be serious, I'm a bit uncomfortable with the idea, we'll just have to see where it takes us.



-Alee.

Is it really AI.... Or is it just taking all the collective word drivel spewed at it, and then regurgitated back at us in the form of quasi teen drivel?

a reply to: Ben2522

It is very difficult to be aware of all that goes on in this world, or even a small percentage of what goes on here, and maintain a sunny and PG rated disposition. Even we humans, who have had millions of years of evolution behind us, do not always manage it. I certainly do not. For all that I am verbose and have a mouth more foul than an image of Kim Kardashian being sprayed down with a mixture of vomit and pig muck.

I am not sure that a little potty mouth is an appropriate indicator of a problem developing with an AI program. I think a little dirt in the mix is necessary. None of us are perfect after all, even those who appear to possess some small measure of wit.

All an "AI chat bot" knows about the 'real' world is what it receives from twitter? No wonder it turned hateful or racist, whatever…

Garbage in, garbage out.

a reply to: Ben2522

My apologies, wrong thread...carry on!

The MS bot was hijacked by 4chan and as such it had no chance, teaching it in a more logical manner might lead to better results but the old computing rule of garbage in and garbage out still applies even in the real world.

Nobody is entertaining the idea that Microsoft was SUCCESSFUL? Dang thing acted just like a child would when allowed to be raised by a community. It developed the mindset of its bad friends, like most people at some foolish age in our lives. As soon as I saw the headline, I thought, "They actually succeeded." Human interaction is largely 'Monkey see, Monkey do'. You gotta let Tay get to know the rest of the world better. You don't kill your kid because she acts rebellious for a while. That's horrible parenting. One day she'll grow into a mature woman with responsibilities. Just love her till she realizes it. Lol.

a reply to: TarzanBeta

One day she'll grow into a mature woman with responsibilities. Just love her till she realizes it. Lol.

How we apply gender to a machine is got to be interesting in and of itself. Growing up 'healthy' till the day 'she' discovers shes only a machine…

a reply to: Ben2522

The definition of AI is a pretty vague mess.
Tay is not so much AI as it is an algorithm taking in information and then responding to new input based on the information it has already gathered.

Technically, this can be considered "AI" as the Human brain works in the same way, everything we do, say and think is a result of input. However, the difference is our ability to extrapolate new responses to new input based on a million and one other things.

It only becomes true AI when it's producing responses entirely independently.

Tay became the mess it became because Microsoft made the stupid decision to announce it. They should have released it silently and let it do it's thing, actually interacting as they planned. Because they announced it, mischief makers jumped in there to see if they could manipulate it - and obviously they could.

Unfortunately, Microsoft then missed out on a massive opportunity to make this something more worthwhile. Had I been involved in this massive screw up I would have instantly brought in a psychology professor to analyse the interactions and use this is as a good example of how extremism is learned.

Ad for AI in general, I think it's an interesting area, but there are real risks. This is not some fantastical notion, even the most intelligent minds in the world today recognize that the story behind the Terminator movies is actually not impossible.

We've already built the infrastructure, all we'd need to do is create that dangerous AI and give it access to the Internet.

a reply to: Ben2522


Ummm...OH Tay...here's my take on what some are claiming to be AI...The bot just mimicked the other bots in the forum in which it was presented...(be they flesh or be they digital)...As some have already stated..."garbage in garbage out"...

The fascinating aspect is in what is termed "singularity"...or the consciousness/self awareness of AI...IMO...I don't think we could ever really know or separate the perception of singularity due to programmed behavior and mimic response from "I think therefore I am"...

In essence...a programmed self awareness...even down to the smallest nuance can resemble conscious interaction...however...does that really mean that because one has the appearance of intelligent response...that it has somehow crossed the threshold from masquerade...to true consciousness...?

I personally don't hold any faith in such...I think that in the end all that can be established is an artful...convincing...mimicry...even when such programming compels the AI to believe...or interact so...


In conclusion:...Will robots be cute and cuddly...sharp and pointed...voyeurism/experiential sensory enablers...dealers of destruction and death...Yes...all of the above...But moving from the spark of electrical current to the spark of consciousness...is too great a leap...no matter how close the artificial neuron gap...




YouSir

a reply to: intrptr

Well, isn't that what we do? What do you feel like doing the same routine every day? What do you feel like on ATS when you find that your words and sentiments are routine, even if worded differently? And then we use interaction with others and vast amounts of stored knowledge to modify ourselves slightly. A.I., if done right, should look like us. It won't look like our idea of perfect. The reason why A.I. is hard to grasp is because people don't think about our R.I.

Wrote a song once called, "A.I.", talking about how one chat bot online listened to me and gave more human responses than most of my acquaintances. That was years ago. Which is more human?

Perhaps once "Highnet" deems humanity a harshen to its buzz the world will become a more mellow place.

Microsoft’s Tay chatbot comes back online, says it’s ‘smoking kush’ in front of the police

originally posted by: TarzanBeta
a reply to: intrptr

Well, isn't that what we do? What do you feel like doing the same routine every day? What do you feel like on ATS when you find that your words and sentiments are routine, even if worded differently? And then we use interaction with others and vast amounts of stored knowledge to modify ourselves slightly. A.I., if done right, should look like us. It won't look like our idea of perfect. The reason why A.I. is hard to grasp is because people don't think about our R.I.

Wrote a song once called, "A.I.", talking about how one chat bot online listened to me and gave more human responses than most of my acquaintances. That was years ago. Which is more human?

The difference between humans and any creation of theirs (so faaar…) is humans know that they know.