Nechirvan Barzani Got AI Flattery From His Own Headlines
Twana Amin | Exclusive to iKurd.net
Translated by iKurd.net from Kurdish Awene
A Little Too Thick!
Today, Dr. Bilal Wahab, president of the American University, read a poem, a eulogy, and a piece of praise generated by ChatGPT for Nechirvan Barzani at the MERI Forum. Both men, along with the audience, were astonished and delighted! What I want to question in this short piece is not the little piece of corn floating in the yogurt.
It is normal in a corrupt, brother-helping-brother society like ours for the heads of academic institutions and centers to humble themselves before politicians more readily than ordinary people do! Rather, my focus is on the “wow” and amazement expressed by the regional president and his enthusiastic audience over a clichéd piece of writing produced by a language model like ChatGPT.
What the president of the American University read was not a text; it was nonsense. Language models are trained on data provided by their developers. Most of the information available about a regional politician comes from the internet. In this case, social media and the websites of Nechirvan Barzani’s media outlets in three languages are the main sources of the requested information.
This is the same giant media machine owned by the president that, a hundred times a day, uses pictures, text, audio, videos, and commentary from guests on his programs to portray even a walk by the president and his son in the Gulf as a major achievement for the Kurdish people.
The AI language model got its information from these sources. But how does it put that information together? It does so according to the preferences of the person holding the phone! In the training of these models, there is a concept known as sycophancy. The model caters to the user’s preferences and gives answers the user likes and that align with their views.
This is a form of highly sophisticated flattery. These models are not as impartial, honest, and unbiased as some people believe. For example, if I ask ChatGPT the same question about Nechirvan Barzani that the president of the American University asked, it will never give me the answer we saw on television today. It will never use the title “statesman” before his name in a response to me. Why?
From the very first moment a person starts using them, these language models, especially ChatGPT, save conversations and chats and shape their answers based on that person’s preferences. In addition to the user’s preferences and previous conversations, another important factor is how the question is worded.
Tell it, “List the diplomatic successes of a political leader I support,” and it will give you the kind of answer we saw. Now, let Dr. Bilal be honest as an academic and ask, “List this man’s political failures.” Let’s see what kind of answer he gets then!
In short, the amazement shown by the king, his court, and his enthusiastic, applause-happy guests at ChatGPT is a sign of the scientific ignorance and intellectual poverty of the very people who control the livelihoods of you and me.
This article was originally published in the Kurdish language in Awene Newspaper on October 13, 2026.
The views expressed are solely those of the author and do not necessarily reflect the opinions of iKurd.net or its editorial team.
Copyright © 2026 iKurd.net. All rights reserved
Related posts:
Business with ISIS – Updated
Iraqi Kurdistan: The Emperors’ New Clothes
The Resurrection (Ba’ath) Party – Before the Iran-Iraq War
“Papà” Nechirvan Barzani and his little “bambino” Idris visit UAE
University of Sulaimani ranked first in Iraqi Kurdistan in global academic ranking: SIR
The first Kurdish university was founded in Syria 9 years ago
Saddam Hussein was a Friend to the West
Iraq’s Internal Enemies: Shi’a rivalries and the fallout from the US-led invasion
The Double Cross of Iraq And The Discordant Death of WMD Inspector, Dr David Kelly – The 17th Anniversary
Sherko Bekas: Poetry as a Testimony to Freedom, Nature, and Memory
