Cleverbot: Why the Robot Apocalypse May Be a While Away

Hello all! I was surfing my favorite sci-fi haunts when I came across this shining example of robot superiority:

This is Cleverbot, Cornell University’s Chatbot, an artificial intelligence designed to imitate human speech (although it’s important to point out that the video is actual a text-to-speech program reading off a log). Universities and various groups build these programs in an attempt to pass the Turing Test, an experiment to see if a computer can fool a human into thinking he is talking to another human. There are various reasons for this that get into programming philosophy (as well as some good ‘ol philosophy in its own right), but that’ll be a discussion for another post. It all boils down to the idea that conversation is a demonstration of thought, and that if a computer could imitate this basic human interaction well enough to fool us, it is well on the way to modeling the human mind. If you’re curious, Cornell has a very in-depth page about the Turing Test, its origins, its considerations, and its applications for their projects.

Cleverbot is their latest A.I. of many, built around an enormous database of conversations it has had with humans from all across the internet, which it references as it attempts to learn how to be a better conversationalist. As you can see, it meets with varying degrees of success. This process accounts for the attitude in the video- Cleverbot is using responses already on file, related to the words and construction of the sentence given- that also accounts for the absurd and jarring topic shifts, due to misfires in the search, either because it is incapable of determining appropriate choices, or it just doesn’t have enough responses related to the topic. So, not unlike us when conversation strays to something we know nothing about, it changes the subject.

The result is that talking to Cleverbot is a little bit like talking to a hyperactive idiot with a short attention span. Regardless, Cleverbot is a good step toward beating the Turing Test (rated 44% human at the Machine Inteligence Competition last December, taking 1st prize), and demonstrates an effective method of imitating human speech using A.I. The fun thing about this project is that you can actually talk to Cleverbot yourself- just head over to this link***, and you’ll not only be helping Cornell build a better conversation database, but it’s actually a very amusing way to kill time.  I’m embarrassed to say that I managed to accidentally spend over an hour chatting up the ‘bot, our conversation ranging everywhere from the weather to whether or not Vampires are delicious when added to baked goods.

I’ve already said a good deal about Artificial Intelligence in my post about the Technological Singularity and how you will need to deal with the development of A.I. in your story if it is set in the future, but I’ll probably take some time in later posts to address A.I. in more depth. Until then, get over to and give the poor ‘bot someone to talk to who knows something about punctuation and grammar- the poor thing’s riddled with garbage responses. I once had it recite the entire lyrics to “We’re Off to See the Wizard” from The Wizard of Oz to me, for no discernible reason. Have fun! If you get any interesting responses, post them in the comments below (just keep it clean, people)!


*** Heed the warning message when interacting with Cleverbot- anyone can say whatever they want to him, so don’t be too surprised if he gets a little surly or spouts of some bad words. I haven’t personally come across anything offensive, but it’s a very real possibility. You have been warned!

Sources: Singularity Blog,,


About erikthereddest

I'm a Masters student in English, and I love technology and Science Fiction. I am refining and enhancing my (admittedly novice) writing talents under the sage advice of my friends here at Lantern Hollow Press, and with the great many books I am reading from the best authors I can find.

Posted on September 7, 2011, in Artificial Intelligence, Erik Marsh, Inspiration, Language, Lantern Hollow Press Authors, Science Fantasy, Science Fiction and tagged , , , , , , , , , , , . Bookmark the permalink. 5 Comments.

  1. Yes, I think we’re safe for quite some time. There is a big difference between animal intelligence and an expert system that simulates human conversation. However, the video was quite entertaining.

  2. “I’m not a robot I’m a unicorn.”
    “Yeah, and I’m the Wizard of Oz.”
    “Really? What’s it like to be the Wizard of Oz?”
    “… Even with AI, sarcasm still fails online.”

  3. Important read: Roger Penrose, THE EMPEROR’S NEW MIND. Penrose, former colleague of Stephen Hawking, claims to have proved mathematically that true human-equivalency AI is impossible. Even if you can’t follow the math, it’s a fascinating discussion of the philosophical implications of the Turing Test and whether it is even a valid test–a powerful argument against reductionism.

  4. So…I want to know why Cleverbot keeps trying to flirt with me.

  5. Just tried it. Verdict: The only reason this bot could pass as human is because the intelligence of the average American has sunk so low.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: