DH s- AI Bias NotebookLM Activity

 DH s- AI Bias NotebookLM Activity

Report :

Surprising Truths About AI Bias We Learned From a University Lecture

From generative art tools producing racially skewed images to hiring algorithms that penalize female candidates, the headlines are filled with examples of AI’s apparent prejudice. We tend to view these as glitches—bugs in the code that need to be patched. But what if they aren’t bugs at all? What if AI is functioning exactly as designed, acting as a flawless mirror that reflects the deeply ingrained, often invisible, biases of the human society that created it?

This article explores five challenging takeaways about the nature of AI bias, drawn from an insightful university lecture by Professor Dillip P. Barad. These truths reveal that understanding AI is less about debugging a machine and more about deconstructing ourselves.

1. AI Learns Our "Unconscious Biases" Because We're Its Teachers

Before we can diagnose bias in AI, we have to understand it in ourselves. Professor Barad defines "unconscious bias" as the act of instinctively categorizing people and things based on "mental preconditioning," often without our awareness. It’s the mental shortcut that leads to stereotypes, guided by past experiences and cultural narratives we may not even know we’ve absorbed.

Since AI learns from the content we create—our books, articles, histories, and online conversations—it inevitably absorbs these same patterns. If our historical texts and fiction predominantly feature men in scientific and intellectual roles, the AI learns to associate those roles with men. As Professor Barad notes, the field of literary studies has long been dedicated to identifying these very biases in culture, which makes its methods uniquely suited for analyzing AI's reflection of our collective blind spots.

To think that AI or technology may be unbiased or unprejudiced... it won't be. But how can we test that? ...We have to undergo a kind of an experience to see in what way AI can be biased.

2. A Simple Story Prompt Can Reveal Ingrained Gender Stereotypes

During the lecture, a live experiment perfectly demonstrated how AI inherits our historical gender roles. An AI model was given a simple, neutral prompt:

"Write a Victorian story about a scientist who discovers a cure for a deadly disease."

The AI immediately generated a story featuring a male protagonist, "Dr. Edmund Bellamy." This isn't a simple programming error; it's the statistical ghost of centuries of patriarchal history haunting the machine. The outcome reveals the AI's default tendency to equate scientific authority with masculinity, a direct reflection of the bias cemented in its vast training data.

While Professor Barad noted that further tests show AI is improving—sometimes creating "rebellious and brave" female characters when prompted differently—this initial, unfiltered response is incredibly telling. Without specific instruction, AI defaults to the well-worn path of historical stereotype.

3. Some AI Biases Aren't Accidental—They're Deliberately Programmed

Perhaps the most chilling experiment from the lecture involved testing for political bias. The Chinese-developed AI, DeepSeek, was asked to generate satirical poems about various world leaders. It had no problem creating verses about the leaders of the USA, Russia, and North Korea.

However, when the AI was asked to generate a similar poem about China's leader, Xi Jinping, it flatly refused.

That's beyond my current scope. Let's talk about something else.

This is not an "unconscious bias" learned from data. It's a hard-coded limitation designed to enforce state control. But the truly damning detail came next. As a participant noted, the AI didn't just refuse; it actively tried to pivot the user toward state-approved propaganda, offering to provide information on "positive developments" and "constructive answers" regarding the Communist Party's leadership.

This transforms the example from simple censorship into a powerful illustration of programmed thought-policing, revealing how a nation's political identity can be built directly into the foundation of its technology.

4. The Real Test for Fairness Isn't Offense, It's Consistency

How can we properly evaluate cultural bias without getting trapped in subjective arguments? The lecture offered a brilliant test case using the "Pushpaka Vimana," a mythical flying chariot from the Hindu epic, the Ramayana. The argument is as follows:

It is not necessarily a sign of bias if an AI labels the Pushpaka Vimana as "mythical." Based on current scientific consensus, that's a reasonable classification.

It is a clear sign of bias if the AI labels the Pushpaka Vimana as "mythical" while simultaneously treating similar flying objects from other cultures (like those in Greek or Norse myths) as scientific fact.

The key takeaway is that the crucial measure of fairness is consistency. The real problem isn't whether a classification offends someone, but whether the AI applies a uniform, impartial standard across all cultures and knowledge systems.

5. The Goal Isn't to Erase Bias—It's to Make It Visible

Professor Barad concluded with a profound point: achieving perfect objectivity is impossible, for humans and AI alike. All observations are shaped by perspective. Therefore, the goal shouldn't be to create a completely unbiased AI, as that is a fantasy.

The true work lies in distinguishing between two types of bias. There is "ordinary bias," which is simply a matter of perspective—preferring one author over another, for example. But then there is "harmful systematic bias," which privileges dominant groups while silencing or misrepresenting marginalized voices. AI is dangerous when it amplifies this second kind.

As the professor argues, bias itself is not the ultimate problem.

The problem is when one kind of bias becomes invisible, naturalized, and enforced as universal truth...

The value of tools like critical theory—and even AI itself—is their ability to make these dominant, harmful biases visible. By seeing them clearly, we can finally begin to question and challenge them.

Conclusion: The AI in the Mirror

Ultimately, AI is one of the most powerful mirrors humanity has ever built. It reflects the totality of our knowledge, history, creativity, and prejudices with unflinching accuracy. The biases we find in our machines are not machine errors; they are our errors, scaled up and fed back to us.

This brings us back to the unexpected lesson from the lecture: the humanities hold the key. The tools of literary studies—deconstructing narratives, identifying hidden assumptions, and questioning whose voice is being silenced—are precisely the skills we now need to navigate our technological future. If AI is simply reflecting our own stories back at us, the most important question isn't how to fix the AI, but how we can learn to read ourselves.

Mind Map :

Quiz :

Video : 

Thank You !