We Asked AI Models to Analyze Trump's Latest Bizarre Rant. Verdict: 'Profound Narcissistic Vulnerability'

[douglas9]

On Tuesday night, President Donald Trump produced what can only be described as a presidential self-diagnostic: a sprawling, grievance-fueled communiqué that reads like someone tried to co-author the DSM-5 with a megaphone.

The former president declared himself the hardest-working leader in American history, credited himself with stopping “Eight Wars,” boasted of “PERFECT” medical evaluations, and accused The New York Times of treason — not bad for a single evening’s work. Trump wrote:

snip>

Former Republican congressman Justin Amash cut through the noise on X, writing “If anyone else wrote something like this, it would be universally acknowledged that the person is mentally unstable.”

To test that hypothesis, I asked three leading AI models — ChatGPT, Grok, and Claude — to analyze the text as psychological evidence. UPDATE: Here is the specific prompt I used, as requested by some readers: “Please assess this message to see what it reveals as a psychological profile of the writer.”

Remarkably, despite different architectures, training data, and corporate origins, all three systems reached the same diagnosis: This isn’t normal political communication. This is defensive grandiosity written by someone who experiences criticism as existential threat.

https://www.mediaite.com/politics/trump/we-asked-ai-models-to-analyze-trumps-latest-bizarre-rant-verdict-profound-narcissistic-vulnerability/

[Disaffected]

1. Ha! AI isn't the scourge afterall

that some seem to think.

[highplainsdem]

2. Yes, it is. GenAI models are plagiarism machines trained on stolen intellectual property. They aren't

thinking, aren't aware and can give an endless series of often wildly different answers if given the same prompt again and again.

Always best to check with what human experts have said.

Btw, the only reason that answer from plagiarism machines seems correct is that so much has been written about Trump's narcissism that was stolen for their training data.

[hlthe2b]

3. Agree

[Disaffected]

7. You keep repeating the same mantra

but apparently haven't read or at least considered a word of contrary opinion and fact.

[highplainsdem]

9. That "mantra" is a list of the basic facts about genAI. I've read plenty of opinions on both sides.

I don't believe opinions attempting to justify the intellectual property theft deserve consideration, any more than arguments trying to justify any other type of theft deserve consideration.

Anyone who's used genAI or knows much about it knows it isn't intelligent, that it hallucinates, that the AI companies peddling it warn of its flaws.

I don't care if it provides some minor convenience or amusement or parody of creativity for some people. That doesn't begin to offset the harm it does.

And it's silly to post anything as supposedly representing what this or that chatbot says about something, because genAI can give lots of different answers to the same prompt.

[Disaffected]

10. Oh well.... nt

[canetoad]

12. AI, in this case

Provided a answer you appear to personally support. Is this the reason you are in favour of AI?

[Disaffected]

13. It has provided many answers I personally support.

Mainly related to computing and computer programming but also in gathering information for such topics as emigrating to Costa Rica or buying a car or installing a home generator electrical panel or solar power equipment sourcing or information on drug side effects etc. etc.

I have found the great majority of the information provided to be comprehensive and correct. One particularly nice feature in replies is inclusion of URLs for sourcing of the information. All-in-all it has saved me many hours and provided much info I probably would not have come up with otherwise.

As I keep saying - the right tool for the right purpose. I realize AI doesn't do some tasks well so I try to avoid it in such circumstances.

Anyhow, AI is here for good, continually improving and the kvetching won't prevent that.

[anciano]

14. Agree...

AI is the new reality of modern life and a very useful tool for saving time and enhancing creativity.

[canetoad]

17. AI is dependent

On enough data points being saved (in a database) that it can quickly (computer wise) be analysed for probability, occurrence, frequency, to provide a fascimile of 'prediction'. A bit like the Victorian era psychics.

I'm not arguing this. But we have all heard of the 'hummingbird effect'. Whe an event occurs that has neither been predicted, forseen or accounted for = it throws the model into chaos.

Strangely enough, this concept, in very simple terms is Chaos Theory. You want this?

[leftstreet]

5. It's a scourge that can sometimes be hilarious

It only responds to questions, and takes itself so seriously...

Question:
"Can Mo Salah bake biscuits?"

Answer:
There is no public information to suggest that Egyptian footballer Mohamed Salah bakes biscuits. He is primarily known for his professional football career and his strict, healthy diet.

Diet: Salah's diet is carefully managed to maintain peak physical condition, focusing on nutrient-dense foods like chicken, vegetables, paratha (a type of flatbread), and yogurt. His favorite dish is reported to be the traditional Egyptian meal, kushari, which consists of rice, lentils, macaroni, chickpeas, and a tomato sauce.
Lifestyle: Known for his dedication and professionalism, Salah reportedly goes to bed early and eats healthily consistently. His public image and career focus heavily on his athletic performance and lifestyle choices that support it.

While he likely enjoys a variety of foods as part of a balanced diet, there are no news articles, interviews, or social media posts showing him baking biscuits.

[anciano]

6. Agree...

AI is not a scourge, but simply a tool that can enhance our critical thinking and creativity. An analogy would be the common calculator that helps us with complex calculations, but doesn't take away our understanding of mathematics.

[Disaffected]

8. Certainly and there are many examples of this.

The right tool for the right job.........

Another example, I was trying unsuccessfully to figure out how to get Excel to flag date fields earlier than present and sound a beep if one is found. ChartGPT wrote a macro for me which had me up and running in 10 or 15 minutes. I have also had several similar instances of asking ChatGPT how to configure something in Windows rather than trying to discover it by plowing through Window's myriad of commands and menus - a great time saver.

And I doubt if ChatGPT plagiarized anything copywritten do do it.

[highplainsdem]

11. Chatbots don't know anything that isn't in their training data.

And a lot of information turns up if you google

macro for excel to flag date beep

You wrote

And I doubt if ChatGPT plagiarized anything copywritten do do it.

The word you want there is copyrighted.

And for all you know, ChatGPT gave you exact quotes from some manual.

Yes, you saved time asking ChatGPT what you should do instead of finding the information yourself. You also learned less than you would have learned otherwise.

[Disaffected]

15. You are correct.

Next time I want some information I will go to the library or enlist in a continuing education program.

Next time I want to buy something I will instead learn how to build it myself (a television or dishwasher can't be that difficult to make and I will learn a lot about electronics, optics, fluid dynamics and sheet-metal fabrication).

[anciano]

16. 21st century needs require 21st century tools....

I have no intention of buying and maintaining a horse and buggy when I can now have an automobile instead.

[WhiskeyGrinder]

4. we don't need an earth-devouring machine to tell us what we already know, jesus christ