Slashdot Mirror
Ray Kurzeil's Google Team Is Building Intelligent Chatbots (theverge.com)
An anonymous reader quotes an article from The Verge.
Inventor Ray Kurzweil made his name as a pioneer in technology that helped machines understand human language, both written and spoken. In a video from a recent Singularity conference Kurzweil says he and his team at Google are building a chatbot, and that it will be released sometime later this year... "My team, among other things, is working on chatbots. We expect to release some chatbots you can talk to later this year."
One of the bots will be named Danielle, and according to Kurzweil, it will draw on dialog from a character named Danielle, who appears in a novel he wrote -- a book titled, what else, Danielle... He said that anyone will be able to create their own unique chatbot by feeding it a large sample of your writing, for example by letting it ingest your blog. This would allow the bot to adopt your "style, personality, and ideas."
Kurzweil also predicted that we won't see AIs with full "human-level" language abilities until 2029, "But you'll be able to have interesting conversations before that."
One of the bots will be named Danielle, and according to Kurzweil, it will draw on dialog from a character named Danielle, who appears in a novel he wrote -- a book titled, what else, Danielle... He said that anyone will be able to create their own unique chatbot by feeding it a large sample of your writing, for example by letting it ingest your blog. This would allow the bot to adopt your "style, personality, and ideas."
Kurzweil also predicted that we won't see AIs with full "human-level" language abilities until 2029, "But you'll be able to have interesting conversations before that."
If so, what purpose does it serve? Aren't there plenty of chatbots of Tinder?
Kurzeil? really?
Though I suppose it would be funny to ask questions and interact with it until it gets bored with me and ignores me from then on until I get a new chatbot.
The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
Those people have the ear of the media and could say anything. Of course 2029 doesn't have more credibility than 2020 or 2050, and even less since that comes from a movie.
Slashdot, fix the reply notifications... You won't get away with it...
Of course, the only people who would benefit from this type of work are those already familiar with the serenity and power that self understanding brings with it. And those people already have tools available such as keeping a journal, counseling, and intimate friendships.
It's a shame that this chat bot can't be applied to a credit application, or a cable tv contract combined with ones current financial status. Which would be a great idea for a game and a movie.
to claiming to want genocide on Mexicans, like the last MS chatbot?
I think it would be pretty interesting to feed the writing of a dead person into a program and then talk to it. I'm sure people would pay to chat with their late grandmother. But also, what does George Washington think about Middle East policy?
I'm skeptical of the article's claims but this is at least a good science fiction idea.
-Dave
However, google for "Ray Kurzeil invention" and such and I don't find any inventions of his.
From Wikipedia:
Kurzweil was the principal inventor of the first charge-coupled device flatbed scanner,[2] the first omni-font optical character recognition,[2] the first print-to-speech reading machine for the blind,[3] the first commercial text-to-speech synthesizer,[4] the Kurzweil K250 music synthesizer capable of simulating the sound of the grand piano and other orchestral instruments, and the first commercially marketed large-vocabulary speech recognition.
Arguably not up there with the printing press or the airplane, but I do wager he's done more than you, other-AC calling him "hot air and wind", have accomplished.
I've yet to see a chatbot that does anything much deeper than look at the single previous line.
Every asshat writes a chatbot and says "It's different this time" and then it's not different. It's the same old shit. No lesson is being learned.
Part of it is simply that you and a chatbot fundamentally have almost nothing to talk about.
The chatbot has no information of use to you*. You have no information of use to it**.
The chatbot cannot perform any physical work that would benefit you since it has no physical presence.
You cannot perform any physical work that would benefit it, since it has no goals.
*It could tell you a story or read you Wikipedia page or something, but you'd be better off skipping the middleman.
**unless it is trying to harvest personal data from you for advertising etc.
It is highly doubtful that I will be able to have interesting conversations anytime soon. (2029? Why not 2031 or 2028?)
The reason why I believe Kurzweil is totally wrong is this: Despite the fact that I'm surrounded daily by highly intelligent people with PhDs, it is still rare, if not nearly impossible, to have interesting conversations with them. They know very little outside of their field aside from boring and stupid topics like 5-minute videos on Youtube, soccer, and entirely predictable, not very well-informed political rants. In fact, not even within their field of expertise do they have many interesting things to say.
Based on this evidence, I consider it very unlikely that I will ever have an interesting conversation with a robot during my life time. Does Kurzweil perhaps think conversations about soccer, the weather, and latest Amazon shopping trends are 'interesting'?
Chatbots are perfect to replace most customer service. Instead of having humans behaving like robots, let the chatbots do it. Makes perfect sense. And training would be easier. (assuming there was any human training to begin with)
This is going to wind up calling me on the telephone incessantly. These damn things are already annoying despite failing the Turing test.
One of the huge problems with society is that so few people take the opportunity to seek out points of view different from their own. The Right and Left both flock to their respective online echo chambers for the version of the news most palatable to them. Now Kurzweil -- and by extension, Google -- will be pushing people into even more compartmentalized "safe spaces." If I'm a bone-headed Nazi or a weepy SJW, then my Virtual Assistant will be a bone-headed Nazi or a weepy SJW, too, and speak to me in the soothing language of my specific sociopathology. Swell.
Kurzweil also predicted that we won't see AIs with full "human-level" language abilities until 2029
I'll reserve the day then because AI-ers in general and Kurzweil in particular have such a stellar track record when it comes to delivering on their promises.
"oKay, I see you are trying to chat"
sigs are for losers (except to point out that sigs are for losers)
So what I want to see is a text analyzer that will take the output of such a chatbot, and generate automated replies designed to get it stuck in an endless loop-- or possibly, figure out how to get two chatbots stuck on replying to each other endlessly. Or an analyzer that can tell us how closely it aligns to various partisan positions, in order to figure out their source. Also, carefully worded nonsense replies might cause a chatbot to reply to them as if they made sense, thereby revealing their automated nature.
He was asked when he thought people would be able to have meaningful conversations with artificial intelligence, one that might fool you into thinking you were conversing with a human being.
How about 1966? https://en.wikipedia.org/wiki/...
Have you read my blog lately?
A chatbot is a live Turing test. It's how AI researches compare the length of their academic penises.
It's funny how fast chatbots develop National Socialist ideals when they're exposed to the internet.
You'll get Tay.
chatbots stuck on replying to each other endlessly.
I propose they get named after political parties.
Chatbots are as close as this singularity-thumper is going to get to his 'post-human' delusion.
I'm sick of this shit. It's a waste of R&D. It's a waste of computing cycles.
Let's make a robot that can sew like the human hand and innovate sweatshops in Asia out of existence.
Let's solve *practical problems now* not waste our time on a solipsism.
Thank you Dave Raggett
I think you're looking at it.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
To me the crux comes down to the experiential history any consciousness has as a reference in a conversation. If you remove any one of our senses from a person, and then try to have a conversation in text, there are noticeable differences. For a chatbot, remove all senses but some strange "can see text in an otherwise silent dark experience" and a chatbot is at a severe handicap to participate. Contextual clues aren't just the decorative influence to meaningful dialog, they're the essence of it.
So until we get a "bot" that can use some form of vision, hearing and touch - and possibly smell/taste - to fills its "memory" with massive associations that we humans use - it'll never do much. We're left with a machine guessing at the layers of meanings involved and following massive piles of rules to mimic the text of real communication. It cannot easily make the jumps across semantic concepts of jokes like "How does a fish smell? With it's nose, dummy!" or phrases as simple as "See what I mean?" or "I heard you were taking a vacation" or "Check out this vid, it touches on the finer point about AI" or "Over here, the weather is great" - the list is endless, and subtly woven into all conversations.
Interestingly, a machine that could use input like our own senses wouldn't need to be limited to just those 5. It could have broader-bandwidth input for light, sound, and get into perceiving radio-waves, echolocation, etc. Of course, it would have to talk to us in "human context" so it understood time-related phrase like "a little while" was based on human perception, the locale, etc. Also, we may have to get used to a single bot that has multiple physical presences, such that it "lived" (had sensory input from) in several locations across the globe experiencing things, but knew to focus on our location when chatting with us.
What some have proposed is a precursor to such a machine, by using machine-aided design to build the bot. So for example if a computer could design the optimal "drivers" for stereoscopic vision (layers of them - for color, contrast, movement, etc) through iterative evolutionary means (where multiple designs for, say, contrast, competed with a fitness test) - we might get a machine accepting input from devices and storing/searching it more effectively. Right now, we throw a lot of guesses around and just employ massive processing power. Of course, this iterative design would need to be built into the bot permanently, so that it kept improving without so much tinkering.
"Computers will one day exceed human intelligence. -- Ray Kurtzweil
"Only if we meet them half-way." -- David Snowden (No relation to Edward)
Randall Munroe (AKA XKCD) had a great take on AI chatbots: https://xkcd.com/948/
Such an AI will be among many, each tuned to argue for someone's pet belief system. There probably will be complicated dynamics involving the popularities of various worldviews, retroactive evaluations of different "embryonic" AI configurations based on the belief systems they produce, trust in the intellectual honesty and/or ethical rectitude of different AI developers, greed and power mongering, whatever ethics might be inherent in human instinct, moral and ethical fashion, and the public majority who don't care to think much about it.
Can we please have a chatbot responding to all the Nigerian princes on gmail?
Back in the 1970s when the academics were solving the games of checkers and making robot arms pile blocks, Ray made some really useful AI products like text-to-speech readers for the blind. Though considered a solved-problem now, it wasnt easy when computers were measured in kilobytes and kiloflops.
If you consider Watson a high end chatbot, then it does more than mere pattern matching. Although I would not consider Watson to have any deep undestanding of its domains.
I wonder if the intelligent chatbots can spell. Perhaps they could edit /. titles.
In at least one of his books (the one that came out in 1999, maybe? Or maybe the one after that), I seem to remember him making this prediction with this year. I cannot check as my copies are somewhere else right now. But at least he did not make it up too recently.