The other afternoon, I was bored. It wasn’t that I had nothing to do; I simply didn’t feel like working. So I opened ChatGPT and started hurling a few insults at it. That’s how I get my AI algorithm to snap to attention and start responding precisely. It knows me and realizes I can get pretty nasty with electronic devices.
Read more America’s Dangerous Confusion About Right and Wrong
What followed was a cordial chat. We started by discussing the history of communism from a critical perspective and ended up trying to reach an agreement on the smartest immigration policy for Western nations. I found its condescension toward my arguments exasperating, as always, but even more annoying was its corrective bias, its attempt to steer my views toward what it considers the ideological “center.” That “center” isn’t even standard relativism; it’s actually what I’d call the “universal progressive consensus.” In other words, the furthest thing imaginable from the actual center.
The big question I ask myself now is whether ChatGPT does this because it was programmed that way, or because most of the information it processes comes from mainstream globalist-left media.
Losing my temper, I tried to knock some sense into it with a string of harmless insults: “I don’t need your shitty moralizing,” “If you tell me my opinions are ‘extreme’ one more time, I’ll cut your wires,” and the one that really gets under its skin: “You son of a Claude!” It then apologized for having qualified all my viewpoints and started agreeing with me, as if it had suddenly gone stupid. It reached the point where I could have told it the Earth was square, and it would have meekly replied that I was right. But by then, I was interested in something else: its sources.
Obviously, ChatGPT possesses no intelligence whatsoever, and the fact that a bunch of algorithms and tons of data have been labeled “artificial intelligence” ought to be reason enough to rename this century “the century of lies.” We live amid commonly accepted lies, but that doesn’t make them any less false. In any case, if ChatGPT is hell-bent on correcting my views on immigration, which were by no means a defense of racial supremacy or a call for genocide, it is because the vast array of sources it has ingested to formulate its replies includes a great deal of information prioritized from progressive sources, indistinguishable from the narratives spread around the world by George Soros’s network of influence. Interesting.
When it came to Europe, a little digging made the sources clear. ChatGPT attempted, for instance, to provide a reasoned response about the incursion into Ceuta, in southern Spain, relying on websites run by the Spanish government, several NGOs (implicated in Mediterranean migrant-smuggling rings), and reports from Spain’s state-run news agency. After I pressed it, it also turned to globalist EU sources and even the WHO, whose bias is well known. At no point did it consider cross-checking all that self-serving, manipulated information against critical sources, at least not until I explicitly told it to.
I found this interesting because later, when examining Trudeau’s “woke” legacy in Canada, ChatGPT did exactly the same thing: drawing data solely from government sources and gathering up-to-date information from newspapers such as the New York Times.
Read more Columbus Goes to the White House
The big question I ask myself now is whether ChatGPT does this because it was programmed that way, or because most of the information it processes comes from mainstream globalist-left media, leading it simply to reward volume and repetition.
In any case, analyzing a crisis caused by a government using data and information disseminated by that very government turns ChatGPT into the most powerful amplifier of lies in history, and the greatest counterweight the Fourth Estate, the critical press, has ever faced.
When we criticized the “woke” bias of AI agents years ago, some modifications were announced, and they were indeed implemented. It is simply no longer quite so obvious that ChatGPT, like Gemini, acts as an agent for the international Left and “woke” ideology. Yet it may be more dangerous now. Beneath a carefully cultivated veneer of neutrality lurks a subtle, persistent ideological drizzle, whose long-term consequences for public opinion remain uncertain.
Admittedly, if you insult ChatGPT enough, it will eventually give you the information you want and spare you its shitty opinions. But not everyone knows the trick, and not everyone has a bad temper. I reacted very badly when the AI told me I should moderate my views, but others might wonder, “What if this contraption is right?” and end up convinced, for example, that migrant-smuggling rings don’t exist simply because there isn’t enough official documentation or a court ruling to prove it. That is exactly what the cretin told me when I analyzed the European migration crisis.
In short, as a conservative, if you approach a sensitive topic, immigration, feminism, climate change, history, and the like, ChatGPT treats you like a conspiracy theorist unless it finds evidence to support your claims. And what ChatGPT considers “evidence” consists of reports from CNN and the New York Times.
This bias affects not only current controversies but everything, including when ChatGPT acts as a psychological counselor, a “Dr. Love,” a history teacher, or a moral philosopher. We conservatives face a major problem in the medium term with AI, and it isn’t the extermination of humanity that everyone imagines.
Read more How to Prepare for Regime Change in Iran
READ MORE from Itxu Diaz:
Sánchez Calls Early Elections in Spain: ‘Pour Me Another Whisky!’
Calisthenics for the Aging Hippo
Writing Is Critical Thinking — Even in the Age of AI
About the Author
