When Her Best Friend Died, She Rebuilt Him Using Artificial Intelligence (theverge.com) 113
When Roman Mazurenko died, his friend Eugenia Kuyda created a digital monument to him: an artificial intelligent bot that could "speak" as Roman using thousands of lines of texts sent to friends and family. From the report: "It's pretty weird when you open the messenger and there's a bot of your deceased friend, who actually talks to you," Fayfer said. "What really struck me is that the phrases he speaks are really his. You can tell that's the way he would say it -- even short answers to 'Hey what's up.' It has been less than a year since Mazurenko died, and he continues to loom large in the lives of the people who knew him. When they miss him, they send messages to his avatar, and they feel closer to him when they do. "There was a lot I didn't know about my child," Roman's mother told me. "But now that I can read about what he thought about different subjects, I'm getting to know him more. This gives the illusion that he's here now."
Re:20 Minutes Into The Future (Score:5, Informative)
Re: (Score:1)
Yes... it's wonderful, isn't it! (Score:4, Informative)
Yes... it's wonderful, isn't it! [youtube.com]
http://www.maxheadroom.com/ind... [maxheadroom.com]
Re: (Score:3)
Wasn't this the plot of a Max Headroom episode?
Yes, it was.
I feel for these people... but they really need to figure out how to move on.
Bullshit (Score:5, Insightful)
Re:Bullshit (Score:5, Insightful)
This is someone dealing with grief, you autistic shitlord.
That has zero bearing on whether it is AI.
(Is there a word like "autistic" for people who are socially clueless and insensitive but which doesn't insult genuinely autistic people? I need to improve my insult vocab.)
"Anonymous Coward"
Re: (Score:2)
Re: (Score:2)
"[NULL]"
LOL.
Re: (Score:2, Insightful)
This is someone dealing with grief, you autistic shitlord.
I wouldn't call it "dealing" with.
Trying to bring back the dead isn't exactly a healthy way to do it.
Sounds very close to the dude who kept his mothers cellphone so that he could send himself messages from it every now and then.
Re: (Score:1)
This is someone dealing with grief, you autistic shitlord.
There are healthy ways of dealing with grief, and unhealthy ways of dealing with grief.
Pretending like the person isn't really gone is definitely in the unhealthy category.
Re: (Score:2)
I still talk to my grandmother sometimes, and she's been dead for fifteen years. One of my chief regrets in life is that I didn't get my shit together in time for her and Granddad to see me make good.
If that's crazy, so be it.
Re: (Score:2)
I still talk to my grandmother sometimes, and she's been dead for fifteen years. [...] If that's crazy, so be it.
It's only crazy if you think she talks back.
Re: Bullshit (Score:2)
Thanks for admitting that it's not artificial intelligence.
Re: (Score:3)
Is there a word like "autistic" for people who are socially clueless and insensitive but which doesn't insult genuinely autistic people?
Anonymous Coward?
Re: (Score:2)
Is there a word like "autistic" for people who are socially clueless and insensitive but which doesn't insult genuinely autistic people?
Redditor.
4Channer.
8Channer.
Computer Programmer.
Re: (Score:2)
No, this is about three levels separated from someone who is "dealing with grief" in a way which many would find very peculiar, and which challenges the meaning of "dealing with" - it's more like "trying to hide from".
The levels of separation are (i) the original article in the Verge (I assume it was not written by te grief-dealer ; the story isn't interesting enough to be worth following) ; (ii) the unnamed person who submitted it to /. ; and (iii)
Re: (Score:2)
Well, it's definitely Artificial. As for the Intelligent, people chatting on text messengers rarely do more that pulling random text lines from the back of their brain that are at worse nonsensical, at best useless. So I'd pass on that one.
Re:Bullshit (Score:4, Funny)
See, computers are smart. Smarter than most of us. How else could they do 60 multiplications a second and never make a mistake? Can you do that? I didn't think so. Computers are smarter than you, and by you, I mean all of humanity.
The end is nigh.
Re: (Score:3)
Re: (Score:2)
And inverting a 1200x1200 matrix in a millisecond is definitely the sign of intelligence. It's doubtful whether most people would ever be capable of that.
Re: (Score:2)
Re: (Score:2)
You confuse "quick" and "smart".
I can certainly drive a screw much faster with a power drill than I can with a screwdriver, but that's in no wise due to the power drill being more intelligent.
Re: (Score:2)
that's in no wise due to the power drill being more intelligent.
Depends on if the power drill is driven by a a dc commutator motor or is a brushless esc [wikipedia.org]/FET driven by a uP running a commutation state machine.
Re: (Score:2)
Depends on if the power drill is driven by a a dc commutator motor or is a brushless esc/FET driven by a uP running a commutation state machine.
That's still not enough to make it smart. It's got to be able to know when it's stripping a screw and stop on its own or something like that in order to be intelligent.
Re: (Score:2)
Most drills have physical clutches. In many cases physical controls are better than electronic, but not with cordless drill clutches. Physical clutches allow a certain amount of torque and then slip. That's wear and tear every time you use it. The drill is also working against itself, since the motor is putting out torque and the clutch is holding it back. With an electronic clutch, the motor only puts out as much torque as is needed. There's no extra wear and tear.
Stripping a screw can be detected by a rapid decrease in torque and increase in speed. Much the same way the AI in anti-lock brakes (developed in 1955) work, although in reverse.
Re: (Score:2)
You missed it. Everything involving a computer is called AI now.
Also, any new idea involving a computer is "disruptive".
Re:Bullshit (Score:5, Interesting)
It's not pulling "random" text lines. It's pulling the text lines that best fit the context, giving (I assume) a somewhat convincing illusion that there is a person on the other end.
This program is clearly not conscious or intelligent in the sense that human beings are. But the current usage [google.com] of the term "AI" does not require that.
Re: (Score:1)
I can understand you statement, however, your comment seems to imply you know what AI is. I believe any program falls onto the spectrum of AI. Its about the recipient as it is as much about the giver. For example, how do you know I am not some advanced AI or just responding to underdeveloped responses? Again not picking a fight, but I think something to consider.
Re: (Score:2)
AI literally has no meaning anymore. I see people calling "Big Data" (another fucking buzzword) AI now. I'm sick to death of it. But you know, if Elon Musk and the Google douches keep saying the words Artificial Intelligence enough then people will start to fucking believe it.
As to this chick, fuck her. You don't get to call it AI just because your feels hurt.
Re: (Score:3)
Automatic lights on your new car when the sun goes down - artificial intelligence. A noise when your smoke detector smells smoke - artificial intelligence. Sprinklers automatically going on at 5am every morning - artificial intelligence. Hot water from the tank, but not too hot - artificial intelligence. The gas pump telling you to have a nice day - artificial intelligence (wasn't that polite of it
Comment removed (Score:4, Interesting)
Re: (Score:2)
Would you like to play a game?
Re: (Score:2)
How about a nice game of chess?
Re: (Score:2)
You can't call something AI if it pulls random text lines from a config file. Talk about an overhyped term. I presume the WordPress Hello Dolly plugin is AI too, right?
I agree but I presumed the AI would be pulling speech patterns from the AI. Responding as the person would have not necessarily restricted to the text input. Kinda like how Swype and other keyboards learn your texting habits.
lost people (Score:4, Interesting)
I wanted to rebuild a friend a long time ago. It really wasn't going to happen on a 386, but I figured I'd anyway get to know him better. He was not exactly excited at the prospect. Well, privacy issues, plus the fact that the whole project was not remotely plausible.
It still isn't . The AI isn't anywhere near close to being able to mimic a real person, yet. But I understand why you would try that, and... go for it.
We may not be able to live forever. It's possible that some semblance of who we were can. Call them poems of humanity.
Re: (Score:2)
I'm not sure about this.... (Score:5, Insightful)
Like everything else, it's got it's good points and it's bad...
The whole point of (Or maybe this is just me) of dealing with someone's death, is the actual letting go part, recognising that they're gone and moving on.
Sure there are times when I miss my friends, and I think of them fondly, whether it's the way that they laughed, smiled, pulled pranks or whatever else, but I also recognise that they're gone. Having them there as a chat bot to talk to, for me, would just, I dunno, make me keep holding onto them... and I don't know if that's healthy...
Re: (Score:1)
I don't see why not. You have no issues with spending hours and hours of your life over the internet talking or writing to people that you will most likely will never meet more than twice if ever.
Re: (Score:2)
Re: I'm not sure about this.... (Score:3)
Like everything else, it's got it's good points and it's bad...
The whole point of (Or maybe this is just me) of dealing with someone's death, is the actual letting go part, recognising that they're gone and moving on.
My thoughts exactly. Something disturbs me greatly about this story. It reminds me of those that live with the dead body of a loved one because they don't want to believe they're dead.
At some point they're going to have to turn the AI off, or come to the realisation that this person wasn't just the sum of their IM output before they can move on, and the grief will hit them then.
Re: (Score:1)
No he didn't (Score:1)
Re: (Score:1)
Re: (Score:2)
Wrong.
Wrong.
Right. But it doesn't mean what you say it does, either, so only ½ credit for this one.
Wrong.
Wrong.
TOTAL: 0.5/5
Maybe you should quit making shit up and learn what stuff actually means.
Re: (Score:2)
It doesn't even have to measure weight, even a shitty "this is/not heavier than 10Kg" tool counts.
You don't have to achieve human-grade consciousness for AI, fine, movies have made a mess, fine, but don't spread this crap.
Re: (Score:3)
I''d argue that the only reason we won't ever have AI is because the all-important part of it, "intelligence", is ill defined in the first place. Although we seem to presume to have intelligence ourselves (rather baselessly, I might add), lacking a rigid definition, how could we ascertain if any other apparently living creature is genuinely intelligent, or if they were actually just issuing programmed responses to stimulation, not unlike computers?
If we can come up with a solid definition of what intel
Re: (Score:3)
Stop with this "Mechanization" bull shit. We don't have Mechanization and probably never will with the way technology is going. And no, a little cogged gear or cotton gin isn't Mechanization even though the hypesters and people wishing for Development dollars try to fool the ignorant into thinking it is.
Black Mirror S02E01 (Score:3)
reminds me of Black Mirror Season Episode 1 (https://en.wikipedia.org/wiki/Be_Right_Back_(Black_Mirror))
Re: (Score:2)
oh this was already posted,..
Another life lesson destroyed by machines (Score:2)
Maybe "moving on" just ain't important anymore.
Y Combinator may already be involved (Score:2)
What's really interesting about this is how Kuyda could develop the technology. As the article says, all of us today are accumulating a digital trail of emails, text messages, social media posts, and online commentary that could be used to train this type of neural network as your digital estate.
Don't go through life as an AC.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Obligatory XKCD (Score:4, Funny)
https://xkcd.com/686/ [xkcd.com]
Hopefully... (Score:2)
Prior Art, 1984 (Score:4, Interesting)
No she didn't. (Score:2)
We're supposed to not be critical of this ridiculous statement in deference to the feelings of someone who lost a love one? That's a new low in Slashdot publishing
Re: (Score:2)
Just curious--how long did it last before CHALLENGE ACCEPTED?
How is this different from brain uploading? (Score:2)
Ray Kurzweil in his Singularity books seems to think that a computer that can mimic a person is as good as that person so this guy must not be dead.
Re: (Score:3)
It's a crude mimic. Give it a couple more decades.
upworthy? (Score:2)
hmmm (Score:1)
I've read or watched this sci-fi before.
necroteliphillia (Score:1)
necroteliphillia, or something like that.
Kinda like the idea... (Score:2)
But why wait until someone is dead?
I would be quite interested in something like this combing the internet and by text messages for all my posts and creating an AI that would respond like I would. At the very least, it could give insight to my friends and family after I've gone. Particularly if a decade after I'm dead my kids would want to ask me a question.
Star Trek (Score:2)
Re: (Score:2)
In one episode, data unwillingly hosted the dead personality of his creator's mentor, his grandfather if you will.
Re: (Score:2)