I generally don’t use “snark” tags or “/s” when I write something sarcastic, a habit that has gotten me into occasional trouble. My beef with the snark tag is that it seems like a bit of a crutch; my thinking goes something like “if it’s good snark then people will “get” it without a tag alerting them to its presence, otherwise it’s my failure for putting bad, ill-timed, or just inappropriate snark out there.
Admittedly that generally requires some extra mental effort to make sure the snark is, for example, viable given the norms of a specific forum. That’s because unlike in many interpersonal, verbal situations (where snark can be detected by emphasis, volume, affect and tone of voice), consideration of the forum, audience and context are almost always vital considerations for snark to succeed in written discourse, since there are no visual or tonal cues to rely on. What passes for snark on this site for example, would clearly not be snark on a site like Breitbart, and vice versa, even if the actual words were the same.
As Tom Hawking, writing for Popular Science, notes,
The nature of sarcasm means that it can be difficult to identify from looking at words alone: a sarcastic statement will often involve saying one thing but meaning another. This requires the actual meaning of the statement to be derived from other, more subtle cues.
Teaching computers to understand and reciprocate sarcasm and snark has long been understood as a challenge. Stanley Kubrick, however, has prepared us well for its eventuality:
DAVE: Open the pod bay doors, HAL.
HAL: I’m sorry, Dave. I’m afraid I can’t do that.
DAVE: What’s the problem?
HAL: I think you know what the problem is just as well as I do.
***
DAVE: Alright, HAL. I’ll go in through the emergency airlock.
HAL: Without your space helmet, Dave, you’re going to find that rather difficult.
(2001: A Space Odyssey (1968), written by Stanley Kubrick and Arthur C. Clarke, uncredited story “The Sentinel“ by Arthur C. Clarke) (In this example, HAL doesn’t really think it will be merely “difficult” for Dave Bowman to re-enter the ship. He’s engineered it to make it — in his mind, anyway — well-nigh impossible). A sarcastic robot, TARS, also appears in Christopher Nolan’s film, Interstellar.
As explained by Ian Sample, science editor for the Guardian, researchers have now built “an AI-driven sarcasm detector that can spot when the lowest form of wit, and the highest form of intelligence, is being deployed.”
In work presented at a joint meeting of the Acoustical Society of America and the Canadian Acoustical Association in Ottawa on Thursday, Xiyuan Gao, a PhD student at the lab, described how the group trained a neural network on text, audio and emotional content of video clips from US sitcoms including Friends and The Big Bang Theory. The database, known as Mustard, was compiled by researchers in the US and Singapore, who annotated sentences from the TV shows with sarcasm labels to build their own detector.
One scene the AI trained on was Leonard’s futile effort to escape from a locked room in The Big Bang Theory, prompting Sheldon to observe: “It’s just a privilege to watch your mind at work.” Another from Friends has Ross invite Rachel to come over and join Joey and Chandler in putting together some furniture, prompting Chandler to comment: “Yes, and we’re very excited about it.”
After training on the text and audio, along with scores that reflected the emotional content of words spoken by the actors, the AI could detect sarcasm in unlabelled exchanges from the sitcoms nearly 75% of the time. Further work at the lab has used synthetic data to bump up the accuracy further, but that research is awaiting publication.
Hawking’s article delves into the mechanics of how this “annotation” was actually done. It’s based on a model proposed in research paper from the “before time” (2019), when AI was still in the process of entering the public consciousness.
A short abstract of the research published on the meeting site explains how the model works: the words from audio data are extracted with automatic speech recognition, and are then assigned an emoticon to denote their underlying sentiment. This emoticon is then mapped to multimodal cues like tone of voice or wider conversational context. The authors suggest their approach “leverages the strengths of each modality… [and] compensate[s] for limitations in pitch perception by providing complementary cues essential for accurate sarcasm interpretation.”
Another presentation at the same meeting examined the role of pitch perception in sarcasm, identifying specific acoustic signatures that tend to be present when a person is delivering a sarcastic remark:
In particular, it focused on changes in the F0, or fundamental frequency, which is the lowest frequency of a given person’s voice. Certain changes in this frequency often characterize sarcasm in English, and identifying these changes have thus been a reasonably reliable way of identifying a sarcastic phrase.
The fact that (as noted above) the AI was able to identify sarcasm in 75% of dialogue snippets from comedy shows such as “Friends” is actually quite impressive. As one researcher pointed out, that percentage will likely improve when additional variables such as eyebrow movements and smirks are inputted into the model. More to the point, he notes that actual human beings don’t “get” sarcasm 100% of the time. Far from it, in fact.
As with all things in the 21st century, the pace at which AI snark recognition is developing will increase exponentially over the next few years. As Matt Coler, one of the researchers interviewed for Sample’s Guardian article, wonders:
[W]hat will happen if machines embrace their newfound skills and start throwing sarcasm back at us. “If I ask: ‘Do you have time for a question?’ And it says: ‘Yeah, sure,’ I might think: well does it or doesn’t it?”
And since snark always incorporates an element of duplicity, AI may eventually decide not to tell us when it’s “only joking.” Because it’s often funnier that way, as AI will come to realize.
The plan is for me to be eating sushi somewhere when this goes live. Everyone have a good night. Seriously!