Discrimination in AI: A Human Problem

Daughtry just dropped a new album: Shock To The System. As a divorced dad rock fan—a cool genre that only the cool kids listened to—it's been on a repeat on my gym playlist… while lifting weights… or when adjusting the weights of hacked-up AI models… (Let's not stray off the topic, shall we?)

 

As if on a cue, Grammarly drops its AI detection tool! The first tool by a mainstream after ChatGPT pulled its not-so-well-done AI detection tool—for right reasons. Now, I am a curious soul and Grammarly addict—who earned me the title of "grammar nerd" among a few... well, spoiler: it's all a sham, and it's always been Grammarly—gotta try it.

 

The playlist is on, and the song "Artificial," in particular, is blasting:
Apocalyptic lyrics!
Cynicism dripping!
And fear of AI? Oh, it's just palpable.

 

Meanwhile, the Grammarly AI detection tool (presumably an AI—missing the "intelligence" part) is busy... well... busy, aggressively AI-shaming any kind of good grammar. Bravo, way to kill career paths here!

 

"This text appears to be AI-generated" - As Grammarly helpfully cares to inform me. OMG, thanks! Should I mess up my grammar to humanize it?... Or perhaps, pass it through another GPT wrapper to supposedly "humanize" it?

 

Hey, listen—I'd rather take my grammar lesson from an AI (or someone at Reddit!?) than a textbook... for better analogies, you know? Besides, I was a lousy student anyway back in school. Thanks to those AI lessons, I can now unleash the wrath of punctuation—em-dashes, in particular—upon my readers.

 

We need to understand something here... AI is an influential medium in our lives at this point. I mean, we are already approaching 2nd anniversary of the beloved ChatGPT. Isn't that enough?

 

We, humans—the actuals with the intelligence here—have this tendency to unconsciously adapt language patterns from our environment. The more we communicate with a medium, the more there is a cross-pollination of vocabulary and language patterns.

 

...And now, you wonder: Is this text a byproduct of an AI prompt that needs to be cited?

 

As the lyrics of the song echo—"No love, no embrace... the death of who we are is right here. It's never gonna stop..."—I couldn't help but wonder how the dangers of AI are just a mere reflection of our own thought patterns. Dark enough?

 

Our tendency to oversimplify complex behaviors of human nature—with profiling and algorithms—is what will be reflected in AI and its algorithms, too: turning them to bias towards imperfections.

 

AI, for the most part, in its current form, is not intelligent either way. We humans are the one with the triggers to the AI—and our fate through them. As the grassroots architecture, we may wanna be mindful of not transferring our own weights—biases and all—to AI weights.

 

The lyrics echo: "Welcome to your worst nightmare....". We might definitely not be ready to be measured by the skewed scales of these AI models—the very scales we are calibrating.

 

"The real question is not whether machines think but whether men do. The mystery which surrounds a thinking machine already surrounds a thinking man." BF. Skinner, Contingencies of Reinforcement