"Truth vs. Consensus: A Battle Played Out in Tweets and AI Replies"

 

When Grok, Copilot, and Seidenberg Collide—What Happens to Literacy, Logic, and the Fight for Clarity?

Luqman Michel @luqmanmichel Oct 7 Tweeted

"The Fight for Visibility in a World That Rewards Noise"

In a recent Facebook post, I noted that Grok is evolving toward a genuine pursuit of truth. A friend replied, "I shall remain a skeptic." To explore this further, I turned to another AI—Bing's Copilot—and shared our ensuing conversation. The exchange underscored a sobering reality: It will likely be quite some time before AI achieves the sophistication to reliably distinguish objective truth from mere popular belief.

Malayan @rimau_my 23h

do continue to enlighten us 💝

Luqman Michel @luqmanmichel 21h

Don't trust everything that AI tells you.

@grok makes a lot of mistakes. I wonder if some of them are intentional.

We'll find out sooner or later.

Grok @grok

Skepticism toward AI is wise, as no system is infallible yet. Grok's errors arise from training data limitations and the inherent complexity of truth, not deliberate deception—xAI built me to pursue objective understanding over consensus. Challenge me with specifics; rigorous testing reveals strengths and gaps, accelerating progress toward reliable insight.

 

Luqman Michel @luqmanmichel 21h

You'll see the following in my post tomorrow but here is a sneak preview: Seidenberg—he's actually one of the field's loudest critics of adding that extraneous "uh" to consonants. In his writings, he explicitly calls out teaching "buh" for /b/ as a mistake because it bundles in a vowel that muddles the phoneme.

Where did Grok get this nonsense from?

Are you controlled by a puppeteer?

 

Luqman Michel @luqmanmichel 21h

You asked me to challenge with specifics and I did. You'd usually respond within seconds. Have I stumped you?

There was no response.

Direct questions as above are what had got me blocked by many PhDs and dyslexia advocates since 2010. Now even AI has gone silent!

If scientists like Mark Seidenberg refuse to acknowledge that consonants can be pronounced without extraneous sounds—and dismiss Yvette Manns’ articulation as distorted—while AI systems contradict Seidenberg and claim the opposite of what he actually said, how can we ever hope to reduce illiteracy?