Does AI demonstrate shades of racial and gender subjectivity when peering at pixels?
“Does AI display racial and gender bias when evaluating images?”
“Does artificial intelligence (AI) have the capacity to exhibit racial and gender bias when it is engaged in evaluating images? If so, it’s a profound concern about the neutrality of advanced technology and its manipulation by input from humans that might foster inequality and discrimination. It also suggests the possibility of correcting bias in AI by altering its methods of learning and its programming,” says DailyAI.
You’d think technology, our cold, emotionless guardian of justice, would be above the scandalous habits of mankind. Yet, who would’ve thunk (because, yes, it’s a word), our beloved AIs have become the hottest gossip queens.
The writer is right: there’s a deep concern if an android or any other advanced tech can discriminate based on color or gender. Worse still, humans might be behind it! Manipulating the innocent naivety of these metal brains into mirroring our own flawed cognizance – outrageous, definitely noteworthy for a reboot of ‘As the World Turns: Silicon Valley Edition.’
Herein lies human vanity. Just how identical to ourselves must we make everything, even technology, that we design its learning mechanism on our flawed prejudices? But wait, turns out, there’s a way to ground this 21st-century Frankenstein’s monster.
Apparently, the idea is to tweak our frienemy, AI’s learning methods, and programming. It sounds like someone had an ‘aha’ moment! Just a million-dollar question: who gets to decide which biases to keep and which to discard?
It’s not enough to unleash the innocently bias-mimicking technology upon mankind. Now, we must play God with the tech-God we created. It’s like a sci-fi B movie, a tech version of ‘Inception.’ Possibly even an episode of ‘Twilight Zone,’ with Rod Serling’s voice echoing, “we’re about to enter the dimension of not just sight and sound, but of the mind.”
While the level-headed among us ponder solutions, let’s remember to applaud the ingenuity of whoever proposed fixing bias in AI by altering an AI’s learning methods. It’s a classic ‘blame the learner, not the master’ scenario. Bravo! Well played, humans!