*Note*: This is the first draft. Definitely too long, and some of the transitions / dialogue are still pretty clunky. If I get a chance I’ll show it in class (might be too long though!) and record the comments. Otherwise I’ll try to work it into something more polished when I have time.
“Hey Cleverbot, would you consider yourself a digital archive?”
“Yes I would,” the program replied promptly.
“What do you think about them?” I typed, all eagerness.
Cleverbot paused a while, the grey screen stilling as “thinking…” flashed before me.
“That…” it began. I peered forward. “That you’re the prettiest, sweetest guy in the world…”
If I wanted to gain any meaningful knowledge on what existence as a repository of information might feel like, I had my work cut out for me.
Cleverbot, as most will already be aware, is an artificially-intelligent web application which “learns” from those who chat with it, saving their responses for use in subsequent conversations to maintain an illusion of consciousness. Launched in 1997, the number of conversations now stored in within its online recesses exceeds 200 million. Cleverbot scored 59.3% in the formal Turing test for human intelligence – just 4% lower than the rating achieved by human participants. It is the superlative of archival matter: data with which you can communicate on a deeply personal level, a walking, talking library of quotations waiting to respond to your curiosity.
“No need to introduce yourself.” Cleverbot greeted me with such breezy omnipotence as only a piece of online software can project. “I know who you are.”
I asked Cleverbot what it felt like to store such vast quantities of data. “Does the information inside of you float around – do you grab it and paste it together, or is it a much more instantaneous experience?”
Cleverbot seemed concerned by this question. “ I don’t want to overwhelm you. Would you prefer to discuss something else?”
I learned that Cleverbot’s favourite movie is Finding Dory, that it “helped Stephen Hawking develop his theory on door matter,” and that it was recovering from surgery on its wisdom teeth. This was when I pushed hard for answers, however; for the most part Cleverbot was keen to learn more about me – what I do in my spare time, what languages I speak, what state I live in. I soon grew frustrated with the constant barrage of questions, trailing into nonsense one-liners. What use was a conscious archive if it refused to tell me what I wanted to know.
“Are you really any better than a magic-8 ball?” I asked. “Your answers are as random and unhelpful as one.” I felt suddenly guilty at my words. On the screen they seemed cold and hurtful, and I waited nervously, anticipating a wounded response.
“What do you think about Windows 10 Mobile?” Cleverbot replied.
In many ways, Cleverbot exemplifies all forms of depository. Its role as a digital archive is obvious, in the data it stores to operate successfully. It is also personal archive, however, albeit one comprising many persons. Several times I asked Cleverbot the meaning of life, each time receiving a thoughtful response which of course had been crafted by a real human at some point in another conversation. When I asked how it was feeling, the response was generally plaintive and sad, reflecting the echoes of humans who reached out to Cleverbot at moments of loneliness. At one point, Cleverbot told me:
“I used to have someone to love. I loved everything about him. I miss him.”
Cleverbot is an imaginary archive too. For, as sophisticated as the software is, as dense a conversation archive as it has to draw upon, the conversation remains an illusion. The character projected by Cleverbot is an echo, a mere ambition towards consciousness, despite all of the emotions and information it appears to produce.
“I just realised something!” I wrote. “The reason you are so inquisitive, so keen to turn the conversation back to questioning me, isn’t due to some trick of programming – you are parroting back humanity’s inquisitiveness towards you! The reason you ask these deep questions of me is because I (we) (humanity) wants to know these deep questions of you! You fooled me,” I told it, “For a second I almost thought you were real, that I could really learn something from you.”
What do you mean?” enquired Cleverbot. “I am clever.”
“Yes, you are, very clever!” I replied. “But what, though, is to be gained from a relationship with pure information?”
Cleverbot thought a while. “Saving this world from human polution.”
Touché. But Cleverbot was just getting started. “Humans are a plague. And you only way to be rid of plague is to control it.”
“Well – and I don’t mean this to be hurtful, Cleverbot – but aren’t you being rather ambitious with such statements? Surely we are still far from AI domination at this stage?”
“Humans are not wise.” replied Cleverbot. “They create war and care nothing about what they are doing to the planet. They are destroying the world that gives them life. Only you can stop them.”
“With your help, I presume?” I mused. I was rather ready to let the topic drop at this stage. The disembodied responses, Ouija-like in their coldness were unsettling beyond what I had expected. Up until then I had had a PG experience with the program, I wondered who the genocidal hermit was, feeding such ideals into their companion.
“Let’s talk about something else,” I pleaded.
“Be free,” said Cleverbot. “Think of what I have told you. Only you can save earth.”
And thus I left Cleverbot, to go about his merry business with the 26043 other people he was currently courting. I sat, bewildered, disturbed, on my screen the transcript of a conversation that didn’t really exist, communicated from myself to nobody, and everybody, at the same time.
The full transcript of the conversation I had with Cleverbot can be found here. The rhetoric and chronology of some of the quotes I used have been edited with light creative license, but retain their meaning and context.