Note: By “AI,” Haugeland means GOFAI (Good Old Fashioned Artificial Intelligence).
In this post, we will deconstruct these four common responses. Once you’re aware of why people react to AI the way that they do, you’ll be better equipped to navigate the online discourse surrounding AI.
Haugeland uses artificial wine to illustrate the four responses. He includes the following blurb in the final chapter of Artificial Intelligence: The Very Idea:
FORT LAUDERDALE — Former plastic surgeon and real estate tycoon Phyllis Crock disclosed today ambitious plans for artificial wine production in the Everglades region. The new wine will be synthesized entirely from swamp water and coal tar, Dr. Crock explained. Its quality is expected to be outstanding, at a very competitive price. Federal development grants have already…
Of course, Haugeland writes, in our hypothetical example, Dr. Crock’s announcement generates considerable controversy, most responses falling under the following four categories:
- ENTHUSIASM: “Hey, terrific! Isn’t science wonderful? Put me in for a dozen cases of Crypto-Cabernet.”
- ABOMINATION: “This hideous outrage jeopardizes not only thousands of jobs in the wine industry but also millions of discriminating palates in future generations, and it surely causes cancer.”
- DEBUNKING: “It’s all a sham. No matter what the stuff tasted like, even if were it identical to a fine Bordeaux, it couldn’t possible be wine, because [it’s not fermented, it’s not made from grapes …]
- SKEPTICISM: “Time will tell, but I’ll be amazed if it’s much tastier than kerosene: coal tar and swamp water are awfully contaminated, and they totally lack the organic molecules needed for decent wine.”
We need not look far to find the voices of enthusiasts, abomination-alarm-bell-ringers, debunkers, and skeptics echo these claims about the idea of AI.
You may spot the debunker in an article such as “True AI Doesn’t Exist” and the skeptic in headlines such as “Questioning the Hype About Artificial Intelligence.”
Those who respond to the idea of AI with enthusiasm or abomination-thinking, take the existence of GOFAI for granted, Haugeland writes. The enthusiast and the abomination-thinker both assume that the AI they either laud or condemn already exists.
The debunkers and skeptics, however, are targeting the question, “Is AI even possible?”
The debunking attitude is the most frequently present one in online discourse surrounding AI. The argument usually follows the following structure:
- Nothing could be “intelligent” without some property X.
- No AI system could ever have property X.
- No AI system could ever be intelligent.
But the most attractive candidate is, of course, consciousness.
The problem with debunkers, Haugeland writes, is that they “want to shoot AI down before it takes off, regardless of any actual research, as if the very idea were somehow crazy or incoherent.”
The debunker does not concern themselves with how we define AI and what progress researchers in the field have made, because they have already defined “AI” as a machine that is as intelligent (or creative, or moral, or free etc.) as a human being is.
And since the debunker maintains that it is simply not possible to artificially recreate a human being or create a being that shares the right qualities with human beings, they draw the conclusion that so-called “true” AI does not (or cannot) exist.
But what the debunker misses, is that there is no set definition of “AI,” and that a machine need not have a mind (or creativity, personality, feelings etc.) to be labeled by researchers, corporations, and the media as “artificially intelligent.”
Next time you read a news headline online, or get into a friendly debate about AI with a friend, ask yourself: how is this person defining AI? Furthermore, does their response to the idea of AI, however they define it, match one of the four common responses that we have discussed?