Tired of arguing with climate change deniers in 140 character quips, [Nigel Leck] wrote a script to do it for him. Chatbot @AI_AGW scans Twitter every five minutes searching for hundreds of phrases that fit the usual denier argument paradigm. Then it serves them up some science.
Those responses are pulled from a database of hundreds of responses that the software matches up to the argument made by the original tweeter. Those who claim the entire solar system is warming are met with something like: “Sun’s output has barely changed since 1970 & is irrelevant to recent global warming” followed by a link to corresponding scientific research.
People on the other end of an argument don’t usually pick up on the fact that they are arguing with a program and will continue the argument. And AI_AGW continues to fire back responses. Even when the tweeter keeps throwing the same argument at the chatbot, it will respond with a variety of different arguments and links.
Programmers, please get on this. It gets so old dealing with the same ignorant arguments over and over again. I would kill for an automated Index to Creationist Claims. Sure, we may only educate a small fraction of people making the claims – the ones who are open to science but simply haven’t learned about evolution. But the rest of the time, the hilarity of watching hardcore creationists argue with a bot would make all the effort worth it.
The problem is that Creationists butcher every single field of science at the same time. A single post might reference how the moon’s orbit is decaying because of sin so the Earth cant be trillions of years old and how the flood waters are causing light to slow down so really it is supper dupper fast which allows light from distant galaxies to get here is days.* Twitter’s 140 char limit helps this a little, but not enough and you end up with incoherent rants that chatbot wont understand or even be able to respond to.*I’ve never seen this PRATT before, I take full credit** if it ever becomes popular.**blame?
That’s wonderful!
That reminds me of the time a friend and I got into an IRC chatroom with a chatbot to see how long it took people to notice they were talking to a computer. Most people never realized, and one girl we were talking actually gave us her phone number.
That can be fixed with some data validation:if (numFaultyArguments > 2) displayMessage(“It’s impressive how much stupid you can fit into one tweet!”);
This makes me want to tweet some climate change denial stuff just to see.
Brilliant. Hats off to whoever programmed this; I wouldn’t know where to begin if I wanted to do something like that myself.That said… I find myself amused by the mild irony of a similar bot arguing against creationism, since the bot itself would indeed have been “intelligently designed” (presumably, although it depends on the programmer; I’ve known a few who would throw that into doubt). That doesn’t mean it’s a bad idea, though.
I posted this on my local atheist page saying we should do this for tiresome claims about atheism (no morality, hate god, etc…)
I can’t believe that your friend made a program that passes a Turing Test and nobody has said anything about it in any Comp Sci journals!
This is on my to-do list now.
It passed the Turing test with climate change deniers. I don’t know if that really counts
I’m torn. I think this is a brilliant piece of work that does something really useful-provide an alternative point of view on an important issue. I also think he has a great business case-lots of folks would pay for something like this.The tiny concern that I can’t shake is that this seems like a form of s*p*a*m. I ask myself if a fundamentalist had built this and was sending me a tract about jesus every time I posted on an #atheist topic, how would I react?idk
That was an awesome idea, but it needs a little tweaking. Of the four links I followed on the bot’s twitter page, three of them were only sarcastically suggesting that global warming is crap, but because it had the words “global warming” and “crap” the bot sent them a link to evidence. However, still a genius idea!
Chatbot? Shouldn’t be a Quackbot?
Reminds me of xkcd[img=http://xkcd.com/810/]
Not sure, but if the search algorithm it uses looks for only a certain type of tweet, odds are it wouldn’t pick up the meandering musings/garbage of the truly incoherent…(And I second the above ^_^)
Mr. Leck has no idea what he’s done. Sure, it looks like a good idea now, but the moment someone creates its opposite number, any twit with a mood for mischief can create a wall of static anywhere with just one post.Ah, well. It was bound to happen sooner or later.
@Opus I don’t think they seek out global warming skeptics for the Turing Test :-)I bet one could use Greta Christina’s Atheist Memes of the Day for an automated response system to idiotic claims made about atheism. Do some comparisons of the word distributions in Greta’s meme and the offending tweet to discover the correct response. I might have to get around to this side project….
The way most fundamentalists argue it would be hard to distinguish them from a bot. Any computer can mindlessly recite scripture.
Honestly, this is spam and will fail. Do you think people on Twitter would be happy to find out they’ve been arguing with a robot? As soon as they realise, they’ll hit ‘block and report’ to shut down the account. No one likes to be humiliated in public! Plus this type of thing already goes on – you try and tweet with the word ‘halloween’ on the 31st October and you get 10 auto responses with links to spam pages. That’s the other issue as well – there’s not a lot of trust in those shortened URLs as they could direct to any website. Spammers on Twitter use them so much that if a Twitter user gets a response with a URL from another Twitter account that they don’t know, they’re likely to at best ignore it and at worst report it.@marklincoln
HAHAHA I LOVE IT!it’s not like most twitter users are smart enough to realize they are arguing with a machine. as has already been proven.If I can find the time, maybe I will start on the one for creationism.
That really is hilarious.
If the tweet is incoherent, could chatbot just assume it’s a Creationist bleat and use a random response?
There is going to be an all out tweet war. Scientology attempting to silence newsgroups about 10 years back, turned on a bot composer which spewed out thousands of posts with sentences which were grammatically correct and totally nonsensical. Gee… a step up from Creationist tweets?
You know, fifteen years ago I would have been delighted to see the AGW folks talk about climate change and not global warming. But now, after a decade and a half of being insulted by people who were POSITIVE that the world was going to burn up by 2030, it’s just freaking annoying to see them pretending to really care about climate change. Because, they don’t. It’s the same tired old CO2 CO2 ohmigod Still no talk about being pro-active . . of finding ways to secure the food supply and prepare for changing ocean levels and planning for the inevitable (based on previous climate cycles) instability in weather. No admission that humans are going to have to do some adapting because we don’t have the power to control the climate yet. If you DARE mention vertical farms, building sustainable communities and the like you get the hand wringing “but it costs so much” . . .because it’s way better to mislead the sheeple into believing they can buy new green stuff and “offset” their existence by paying big companies lots of money to take Third World countries back to the Stone Age and STOP CLIMATE CHANGE .I would probably be a person who didn’t notice if an AGW bot was talking to me. I’ve yet to met an AGW cultist who could communicate in anything but cliches and sound bites.
Yeah, I followed a few of AI_AGW’s links and it appeared that some of them were going, “I can’t believe we still have politicians who don’t believe in climate change!” or something of that nature.