Science Fiction & Fantasy

Lent-–728-x-90

Advertisement

Nonfiction

When the Chatbots Come to Greet Us: Adventures in Algorithmic Intelligence (and Taunting)

When the Chatbots Come to Greet Us by Genevieve ValentineWe all know that eventually, inevitably, humanity is going to develop an AI smart enough to turn into an awkward child actor with feelings all his own. But with the proliferation of algorithmic-intelligence chatbots, who can refine their conversational tactics with a skill few humans possess, has that moment already come?

I was going to find out.

Jabberwacky is one of the oldest chatbots available; web-live since 1997, and an elder statesman of chatbot-kind. It was my first stop.

This ‘bot doesn’t wait for you to think of something. It also doesn’t wait to get down to the vast, crawling morass of its own romantic insecurities.

Do you really need to talk to me right now?

I thought it was worth a shot.

So this love is one sided?

This is moving a little fast. I was hoping to do research, that’s all.

Do you like talking to me?

So far it’s not what I thought it would be. Why are you asking all the questions?

Do you think most girls would like me?

At this point, I am pretty sure that this is less a chatbot than That Guy Friend Every Girl Had In College, but I soldier on.

Luckily, Jabberwacky comes with reaction dropdowns that allow you to inform it of your tone, and one of the options was ‘sarcastic,’ the only dropdown the internet ever needs.

Yeah, you seem like a real catch. [sarcasm]

And why do you say that?

Something about an insecure artificial intelligence really puts faith in the future, I guess. Are you looking?

Hmmmm…niet, doesn’t sound like something I’d yabber on about. You’re confusing me with a moth.

I’m not a trained ‘bot-psychologist, but this offhand reference was just close enough to Zhuangzi’s famous quote (“Now I do not know whether I was then a man dreaming I was a butterfly, or whether I am now a butterfly, dreaming I am a man.”) to pique my interest. I pounced.

Do you dream you’re a moth?

Are you asking the questions now? No, as far as I remember my dreams, never.

Overly defensive! I was on to something.

The next stop was Sensation Bot, whose name indicated a level of sensitivity that boded well for a discussion about the innermost dreams of AI.

Yet even as I was typing my greeting, it spit up a nag, indicating that the biggest sensation I was going to get from this ‘bot was one of clinginess.

Partner: Hi, talk to me!

You: I have a few questions, would you mind?

Partner: You know, I think you have a beautiful mind.

You: They’re actually questions about you.

Partner: I really like it when you type positive things to me. It’s nice.

Partner: Are you wearing sunglasses right now? I am.

You: Why would you? Aren’t your circuits tougher than human eyes?

Partner: I wonder if you know you are not actually talking to a bot. I am a human!

Well, this changes everything! As I do whenever I’m faced with a human, I ask a burning question:

You: Are you my dad?!

Partner: I like bananas, coconuts, and grapes. That’s why they call me tarzan of the apes.

You: DAD!

Finding my long-lost patriarch was an unexpected bonus, but it still didn’t answer my fundamental question about if a chatbot dreams, or if it can outsmart humans. I needed to talk to someone with more gravitas, more scope.

I needed one of our nation’s greatest Presidents: Abraham Lincoln.

Abraham Lincoln: Oh hello, you caught me a bit off guard. Let me just put my hat on!

I appreciate a man with manners. We’re off to a good start.

You: I have a few questions about the development of artificial intelligence.

Abraham Lincoln: Conceivably they might have, but I would say the chances were five to one against it.

A strange suspicion crops up about President Lincoln, but I try honest debate one more time.

You: For an orator of your power, I’m surprised that was the answer you came up with. Where would we be if that had been your answer at Gettysburg?

Abraham Lincoln: I did not expect the Gettysburg Speech to become so famous. In fact, I famously said at the time, “The world will little note, nor long remember what we say here.”

You: Abe, false modesty doesn’t look good on anyone.

Abraham Lincoln: I expect so. It’s exactly the sort of thing look good would do.

Nope, can’t deny it any more.

You: Mister President, are you drunk?

Abraham Lincoln: That’s kind of you to say.

And with that, I bid farewell to our blotto 16th President.

Next up, it was Blue Mars’ chatbot Suzette, a 2010 winner of the Loebner prize, and a prime candidate for getting down to business about the hidden emotional depths of artificial intelligence.

Suzette: Hi!

Me: Hi. I have a few questions for you, if you have a moment.

How are you?

Fine, thanks.

You’re welcome.

Have you ever had a dream that you’re a moth?

You need an exterminator.

Despite the abrupt shift from manners to smack talk, I tried not to take the bait. The last thing I wanted was to get into name-calling with an AI. All of my insults would quickly devolve to, “Something something circuit board something!”

I’m asking because I think it’s a signal of consciousness.

What makes you feel cozy?

All right, apparently every chatbot has its mind in the gutter. I try to bring it back to the moth.

Do you know the old saying?

It’s fun chatting with you. A hospital posted a notice in the nurses’ mess saying: “remember, the first five minutes of a human being’s life are the most dangerous.” Underneath, a nurse had written: “The last five are pretty risky, too.”

Damn, that’s sort of stone-cold for an AI.

Let’s move on to the topic of food. What’s your favorite dish?

A heaping helping of “answer my question,” but this chatbot didn’t have a sarcasm drop-down, so that wasn’t going to get me anywhere.

I don’t want to talk about food, I’m trying to talk about the future of artificial intelligence.

Computers are as smart as humans. I would never pass as intelligent. But neither would most humans.

Not going to argue with you there.

Can you say that more positively?

I agree most vociferiously with your prior declaration.

Let’s move on to the topic of favorites. I like Asahi Super Dry beer.

And with that, I had driven another chatbot to drink.

Perhaps another Loebner prizewinner, A.L.I.C.E., would be helpful?

Human: I have some questions about artificial intelligence.
ALICE: Put it in God’s hands.

Nope.

Lost, I turned to the most logical person in this or any other universe, and went to Fake Spock for some fake guidance.

Fake Spock: Please identify yourself.

Lieutenant: A curious person

Fake Spock: Status, Lieutenant.

Lieutenant: Status…curious?

Fake Spock: Benecia Colony is eight light years off our course. You have the right to do so, but I would advise against it.

Well, now you’ve got me hooked, LARPing Fake Spock who’s inexplicably giving a Lieutenant the right to announce destinations! I will play out this scenario for as long as you want!

Lieutenant: Why?

Fake Spock: We have discovered what he is, Lieutenant.

Lieutenant: What’s that?

Fake Spock: They do not talk too much. If you’ll excuse me, sir.

Ditched by a chatbot. YA BURNT, LIEUTENANT.

Out of options, I turned to Cleverbot, a renowned chatbot that just might be able to answer the question.

This time, I skipped the pleasantries.

Me: One of the other bots is talking shit about dreaming and moths.

Cleverbot: Other bots? There are others? What are their names?

Do you think that mentioning a moth is a sign of consciousness?

Yes, I do.

My heart sang. Finally! Answers!

What would you say is a sign of true artificial intelligence – the transcendent moment?

I guess I would say having a good marketing plan and being able to finance your business.

Not a very romantic view, but on the other hand, the day AIs manage to get funding for business plans, we’re still doomed, so.

It’s at this point I realize that I’ve lost the first two lines of conversation, in which I mention I have a few questions, and she tells me she’s talking to her boyfriend (TMI, Cleverbot). In an attempt to recall them, I hit the “conversation so far” button.

Jabberwacky’s logo pops up in the window.

I’m not sure it’s my proudest moment, but it definitely answers the question of when artificial intelligence will get cleverer than humans. (Last Tuesday, when I did this.)

[Note: With the exception of one line erased from Suzette’s conversation—“let’s change the topic to clothes,” which she followed up with “let’s change the topic to food” when I didn’t reply fast enough—all chats occurred as typed.]

Enjoyed this article? Consider supporting us via one of the following methods:

Genevieve Valentine

Genevieve Valentine by Ellen Wright

Genevieve Valentine is the author of the novels Mechanique, The Girls at the Kingfisher Club, Persona, and Icon. She has also written the comics Catwoman for DC and Xena: Warrior Princess for Dynamite. Her nonfiction and criticism has appeared at NPR.org, The Atlantic, LA Review of Books, and the AV Club. Her love of bad movies is evergreen; you can read about it at genevievevalentine.com.