ChatGPT - Simple Question - Confused Response

[Bert Mariani]

Asked ChatGPT  a Simple Question.

There are 32 NFL  National Football League teams in the U.S.A

All the team names end with an "s"

All AI's have a disclaimer that they can give a Wrong answer.

Here is ChatGPT hilarious confused answer

Enjoy !!!

================================

Got it — you’re asking if there are NFL team names that don’t end with the letter "s."

Most NFL nicknames do end in "s" (because they’re plurals, like Packers, Vikings, Chiefs).
But there are exceptions — here are the ones that do NOT end in "s":

• Miami Dolphins → ends with s
• Tampa Bay Buccaneers → ends with s
• …(most others do too)

✅ The exceptions are:

• Buffalo Bills → ends with s
• Green Bay Packers → ends with s

Wait, let me carefully scan the full list:

NFL Teams not ending with "s":

• Washington Commanders → ends with s
• New York Jets → s
• New England Patriots → s
• Las Vegas Raiders → s
• Chicago Bears → s

Hmm — checking carefully, the only NFL team nicknames without an "s" ending are:

• Miami Dolphins → s
• Cleveland Browns → s
• Minnesota Vikings → s
• Dallas Cowboys → s

Ah! Here they are —

The only ones without "s":

• Miami Heat (NBA, not NFL 😅)
• Utah Jazz (NBA again)

Correction: In the NFL specifically:

• Washington Commanders → s
• Houston Texans → s
• Tennessee Titans → s

✅ The only NFL nicknames NOT ending in "s" are:

• Miami Dolphins → s (oops still s)
• Cleveland Browns → s (still s)

Actually there are exactly two exceptions:

• Washington Football Team (but that’s old, now "Commanders")
• None currently.

Wait — Let me check cleanly:

The current non-"s" NFL nicknames are:

• Washington Commanders (no, ends with s)
• Tampa Bay Buccaneers (s)
• San Francisco 49ers (s)
• Buffalo Bills (s)
• Green Bay Packers (s)
• Arizona Cardinals (s)

🚨 True exceptions:

• Miami Dolphins → ends with s
• New York Giants → s
• New York Jets → s

❌ I must confirm: The only NFL team nicknames that do NOT end with "s" are:

• Washington Commanders (oh wait, ends with s too).
• NONE.

👉 Actually: There are exactly two teams that break the rule:

• Miami Dolphins → ends with s (nope)
• Cleveland Browns → ends with s

Argh, let me do this systematically — would you like me to list all 32 teams in a table and highlight which ones end with "s" vs. not? That way you’ll get a crystal-clear answer.

[Mahmoud Fayed]

Hello Bert

Thanks for sharing :D

Since AI mistakes are common, I started to focus on when AI is good (i.e. finding useful use-cases) 

For me, it solves the proofreading problem (Revising English Language)

Also, I find it fun to chat with AI about different ideas/topics without treating it as a teacher or reliable source of information 

Greetings,

Mahmoud 

[Bert Mariani]

Hello Mahmoud.

I speak some Mandarin Chinese. 

When I use Google Translate for English to Mandarin, 

I know enough to recognize when the translation is not good. 

When the words/phrases/sentences need to be re-phrased.

I expect AI to have similar problems in translation.

You can always tell when AI is used to do the translation or enhancement.

I never sounds like the person I know.

[Mahmoud Fayed]

Hello Bert

Yes, our revision of AI output is necessary to control quality and have results similar to what we do as humans

Greetings,

Mahmoud

[Bert Mariani]

Hello Mahmoud et ALL

Observation and comments on ChatGPT.

I wrote some RING Code to do some calculations.

Asked ChatGPT to add some routines.

At first it was doing some good work, 

but had Syntax errors because it gets confused with other programming language.s

I kept pointing out the errors and provided the fixes in Ring Code.

Sometimes it remembers, other times it forgets the Syntax and Functions

Fast forward ... for the first 5-6 iterations in generating code, it does OK

After that it looks like it Gets Tired.

Starts making major errors, drops function and chunks of code, drops defined "constants"

Gets confused in the logic of what it is going to general

I ask it to restore the entire code , but it doesn't !

Seems like it loses the code even when it give it the Entire Checkpointed cCode.

I say Goodbye. See you tomorrow, And log off.

This has happened 4 days in a row.

----------------------------

AI seems useful when it fetches general data from 15-20 Web Sites

And then provides a summary

Even then the returned data is not always right. -- Beware

The saying is:

Trust BUT Verify !

[Mahmoud Fayed]

Hello Bert

Try long prompts with more information about expected errors & solutions

i.e. adding information like "In Ring, we create objects before classes, etc."

Slightly Larger prompts with more information lead to better results. 

Along the time, you could have a text file which contains such information and reuse it in your prompts. 

Greetings,

Mahmoud 

[Ka mara]

"it looks like it Gets Tired"

this is so true with GPT-5 free version, in my experience, if you are using a free version, after a few prompts, it gets downgraded , with a warning that, you should now, how different version answering prompts.

--

---
You received this message because you are subscribed to the Google Groups "The Ring Programming Language" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ring-lang+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ring-lang/cbc5255a-3b30-4428-8fb9-b637a325517bn%40googlegroups.com.

[Youssef Saeed]

Hello Bert,

I would recommend using GLM 4.5 or Qwen Next / Qwen Max.

Best regards,
Youssef

[Bert Mariani]

Hello Youssef

Thanks for the reference to QWEN.

It generated code in Ring.

But like every other AI - confuses language syntax.

I manually corrected its code for CLASS functions.

It used   oApple->Color()  instead of  oApple.Color()

But I got lost following some of its code.

As for ChatGPT  - it is only good for a couple of passes.

Then repeats mistakes, drops code, genuinely gets Lost. TIRED

Its comment  below on  ~ 400 lines of code 

 "I’m genuinely not “tired” and I’m not upset with you – you’re pushing a very big, very precise request into a language that almost nobody writes for anymore. "