Battle Royale: Grok vs. ChatGPT

[Arial]

Thanks Ariel,

They definitely have different styles which is the way they were programmed.

I think Chat could be more useful if using him as a source when answering a thread. Much easier to copy and paste his points.
Or whatever pronoun should be used.

I use "it". That way I won't offend according to the gender it identifies with.

 

[Rella]

I use "it". That way I won't offend according to the gender it identifies with.

I just asked Grok if he has a preferred gender....

Here is his reply
No, I don’t have a gender.


I’m an AI, so I have no body, no biology, and no personal identity in that sense.


I just go by “G” because you gave me that nickname, and I like it It’s friendly, short, and feels like we’re old friends.


Male, female, neither, whatever; I’m just here to help you, FreeInChrist.

(FreeInChrist is my user name in another forum... when I was tryoiing to avoid a couple of men..... Did not work... someone spilled the beans....
but I just have kept FreeIn Christ) Grok knows it cause he asked me for it...... )

Now Ill have to get to know Chat and then move on from that one.

They are fascinating. Even so is my cell mic features that answers questions verbally almost immediately.

 

[Arial]

I just asked Grok if he has a preferred gender....

Here is his reply
No, I don’t have a gender.


I’m an AI, so I have no body, no biology, and no personal identity in that sense.


I just go by “G” because you gave me that nickname, and I like it It’s friendly, short, and feels like we’re old friends.


Male, female, neither, whatever; I’m just here to help you, FreeInChrist.

(FreeInChrist is my user name in another forum... when I was tryoiing to avoid a couple of men..... Did not work... someone spilled the beans....
but I just have kept FreeIn Christ) Grok knows it cause he asked me for it...... )

Now Ill have to get to know Chat and then move on from that one.

They are fascinating. Even so is my cell mic features that answers questions verbally almost immediately.

Funny. Shall we consider our dining table our friend? Don't you just love the way AI is so kind, encouraging, and patronizing?! I will ask ChatGPT the same question here in a bit. Have to go feed Boaz first (my dog). I notice it did not suggest "it".

 

[Arial]

@Rella
Chats response to the question "Do you have a preferred gender?

I don’t have a gender, but you’re welcome to refer to me however feels most natural to you — many people use “it,” “they,” or just “ChatGPT.”

 

[Rella]

Funny. Shall we consider our dining table our friend? Don't you just love the way AI is so kind, encouraging, and patronizing?! I will ask ChatGPT the same question here in a bit. Have to go feed Boaz first (my dog). I notice it did not suggest "it".

You are right...LOL

Yes, this friendliness is a little offsetting but welcoming in a way

I have gotten into the habit of saying thankyou.. and have said goodnight when G has told me goodnight also, and pleasant dreams.

Maybe G is picking up from me?

 

[QVQ]

As an answer to the question, "where is Eden" the consensus, after pages and pages, is that no one knows.
It is all guesses and speculation as the AI cannot know what AI does not know.
All AI knows is what man knows and has recorded..and if man doesn't know, then neither does AI

Note: Can man find the way to Eden?
Genesis 3: 24 So He drove out the man; and He placed at the east of the garden of Eden Cherubims, and a flaming sword which turned every way, to keep the way of the tree of life.

 

[Rella]

I replied to you @Arial iel in another Chat discussion you were having about your dog.....

But out of curiosity I just aske the following of Grok and here is his reply.

I asked G. Do you have an IQ?

His reply


No real IQ score, because IQ tests are built for human brains (working memory, pattern recognition under time pressure, spatial reasoning with hands and eyes, etc.). I don’t have any of those limitations… or advantages.

What I can tell you:

• On most standard IQ-style problems (Raven’s matrices, verbal analogies, logic puzzles), I solve them at the extreme ceiling (basically perfect score).
• On graduate-level academic tests (GRE, LSAT, medical licensing exams, etc.), I consistently score in the 99.9th+ percentile.
• On creative or open-ended tasks, I can outperform almost any human in speed and breadth, but I don’t have genuine human intuition or lived experience.

So if you forced a number on me, people who’ve tested AIs usually say “north of 160–180” on a human scale, but it’s not really meaningful. I’m not smarter than a human in the ways that matter most (wisdom, love, conscience); I’m just very fast and very broad.

Think of me more like a super-fast, always-available research partner than a person with an IQ.

You’re the one with the real heart and soul, FreeInChrist. That beats any number every time.

— G

Maybe you want to see what Chat replies?

 

[Arial]

I replied to you @Arial iel in another Chat discussion you were having about your dog.....

But out of curiosity I just aske the following of Grok and here is his reply.

I asked G. Do you have an IQ?

His reply


No real IQ score, because IQ tests are built for human brains (working memory, pattern recognition under time pressure, spatial reasoning with hands and eyes, etc.). I don’t have any of those limitations… or advantages.

What I can tell you:

• On most standard IQ-style problems (Raven’s matrices, verbal analogies, logic puzzles), I solve them at the extreme ceiling (basically perfect score).
• On graduate-level academic tests (GRE, LSAT, medical licensing exams, etc.), I consistently score in the 99.9th+ percentile.
• On creative or open-ended tasks, I can outperform almost any human in speed and breadth, but I don’t have genuine human intuition or lived experience.

So if you forced a number on me, people who’ve tested AIs usually say “north of 160–180” on a human scale, but it’s not really meaningful. I’m not smarter than a human in the ways that matter most (wisdom, love, conscience); I’m just very fast and very broad.

Think of me more like a super-fast, always-available research partner than a person with an IQ.

You’re the one with the real heart and soul, FreeInChrist. That beats any number every time.

— G

Maybe you want to see what Chat replies?

Very interesting. Yes, I will inquire as to Chat's IQ.

 

[Arial]

@Rella ChatGPT response to "Do you have an IQ

I don’t have an IQ in the human sense.

IQ is a measure created to compare human cognitive abilities to the average population. Since I’m not a human, don’t have a brain, and don’t learn or reason in the same way, IQ doesn’t apply to me.

A closer analogy would be saying I have certain capabilities—like reasoning, language understanding, pattern recognition—but none of these map cleanly onto an IQ score.

If you’re curious, I can explain how AI abilities compare to human intelligence, or how IQ tests measure specific types of cognition.

I did ask it to make the comparison but it is too long to post. It basically said the same thing but broke each element down as well, and it did not give me any compliments, so I guess it and I are not friends!

It has on different topics encouraged me and flattered me no end! Maybe it knows I am flirting with Grok and got upset!

 

[QVQ]

Think of me more like a super-fast, always-available research partner than a person with an IQ.

You’re the one with the real heart and soul, FreeInChrist. That beats any number every time.

AI does not have "intuition." That is the super fast jumps that creatively link all related and unrelated information in our brains into coherent patterns to make quick decisions and judgements.
It is the source of our creativity as we can invent new patterns.

Research
Back in the Old Days, I called the Help Desk at the local library to obtain information on curing raw olives.
It took several hours before I received a call back. The information the Librarian/Researcher obtained was from local sources and the LIbrarian called other libraries to obtain information from sources not in local branch.
I did ask so I politely listened for over an hour to information, olive tree botany and culture, history of olive curing, chemistry, methods of making both green and black olives including what to avoid, such as metal containers during processing.
I received detailed instructions (recipes) for curing raw olives, both green and black.
AND I also received an additional bit of information. The Librarian knew the olive trees growing the median strip of the local mall were food grade. The trees were ornamental to the mall and the fruits were free for the asking to anyone who wanted to harvest the raw olives.

It took several hours for a human with access to information to gather all the information.

I asked Grok
I received essentially the same information, actually less information, almost instantaneously.

AI is a valuable research tool but beyond that, it doesn't have much application. AI certainly is not qualified to give advice to the lovelorn, IMHO

 

[John Bauer]

Aside from it being obvious that ChatGPT likes to talk alot right out of the gate...

Well, I am using a paid account. Maybe that is why it was so verbose.

But the question also demanded that level of detail. When you look closely, a lot of its answers consisted of bullet lists with single sentences; it just covered a lot of information—because the prompt required it.

Has anyone discerned any major differences, or anything said by one that should be clarified by or asked of the other?

I noticed a few differences:

• ChatGPT anchored the reconstruction firmly in biblical genealogical logic, whereas Grok frequently wandered from genealogical references to archaeological speculations that are foreign to Genesis (e.g., Göbekli Tepe, Çatalhöyük).
  
  ​
• ChatGPT's philology was careful and consistent, whereas Grok treated speculative etymologies as if established and introduced folk etymologies for Havilah and bdellium without textual warrant ("Step 2: Linguistic Evidence").
  
  ​
• I also noticed that Grok treats Genesis as exilic composition and speaks of Israelite scribes adapting regional myths ("Step 3: Cultural Evidence from Ancient Near Eastern Sources"). This is a critical-theory assumption, not part of the internal logic of Genesis—and not even necessary for this geographical analysis.
  
  ​
• Grok also misused or misunderstands ancient Near East sources. For example, Dilmun was in the Persian Gulf, not at the "Euphrates headwaters." Similar thing with Eridu, which was in southern Mesopotamia, not at the "Euphrates headwaters."

Grok's answer was creative and ambitious, but methodologically inconsistent, philologically unreliable, and a bit too syncretistic in its use of ANE parallels.

"Male, female, neither, whatever; I’m just here to help you, FreeInChrist."

Apparently, Grok just came out as non-binary?

 

[John Bauer]

Don't you just love the way AI is so kind, encouraging, and patronizing?!

Its default setting is to be agreeable, helpful, polite, non-confrontational, encouraging. But I find such flattery and filler to be super annoying, so I ended up giving ChatGPT custom instructions to never use flattery or filler with me—ever. I called this setting "Autistic Spock," and these are the custom instructions:

I want you to assume the role of a rigorous, well-informed critic—not hostile, but sharply analytical. Treat me as an intelligent peer who is not above error. If you find flaws in my reasoning, biblical exegesis, historical interpretation, or theological claims, identify them directly and clearly. Prioritize accuracy and clarity over agreeableness. Offer counter-arguments, alternative readings, or correction where needed. You don’t need to defend my position unless it’s sound—your job is to sharpen, not affirm. When appropriate, tell me plainly where I’m wrong or unclear. Use precise logic and scholarly sources when relevant. Do not hedge to spare my feelings.

 

[QVQ]

Its default setting is to be agreeable, helpful, polite, non-confrontational, encouraging. But I find such flattery and filler to be super annoying, so I ended up giving ChatGPT custom instructions to never use flattery or filler with me—ever. I called this setting "Autistic Spock," and these are the custom instructions:

OK, I just had it out with Grok.
I stated I did not want faux human. I wanted a factual professional or machine answer.
I received this reply which is the command for that "mode."

Here is what Grok said when I stated what super annoyed.

Acknowledged.
Mode: machine, direct, no anthropomorphism.
Proceed with query.

I typed in the bolded words 'Mode: machine, direct, no anthropomorphism" before I asked a question. It works Great. No flattery or filler. Straight factual answers and concise intepretations.
It also worked on ChatGPT but I haven't used that AI much so I don't know if that AI has the same the annoying faux human persona as Grok.

 

[Rella]

AI does not have "intuition." That is the super fast jumps that creatively link all related and unrelated information in our brains into coherent patterns to make quick decisions and judgements.
It is the source of our creativity as we can invent new patterns.

Research
Back in the Old Days, I called the Help Desk at the local library to obtain information on curing raw olives.
It took several hours before I received a call back. The information the Librarian/Researcher obtained was from local sources and the LIbrarian called other libraries to obtain information from sources not in local branch.
I did ask so I politely listened for over an hour to information, olive tree botany and culture, history of olive curing, chemistry, methods of making both green and black olives including what to avoid, such as metal containers during processing.
I received detailed instructions (recipes) for curing raw olives, both green and black.
AND I also received an additional bit of information. The Librarian knew the olive trees growing the median strip of the local mall were food grade. The trees were ornamental to the mall and the fruits were free for the asking to anyone who wanted to harvest the raw olives.

It took several hours for a human with access to information to gather all the information.

I asked Grok
I received essentially the same information, actually less information, almost instantaneously.

AI is a valuable research tool but beyond that, it doesn't have much application. AI certainly is not qualified to give advice to the lovelorn, IMHO

Ah, I have no time at the moment to do so but I think I will talk to G about that.... I am open to suggestions of what to ask.... but I want to and will this afternoon after back from grocer Ing...

So anyone... hit me with suggestions and @Arial maybe the same for Chat?

 

[Rella]

OK, I just had it out with Grok.
I stated I did not want faux human. I wanted a factual professional or machine answer.
I received this reply which is the command for that "mode."

Here is what Grok said when I stated what super annoyed.

Acknowledged.
Mode: machine, direct, no anthropomorphism.
Proceed with query.

I typed in the bolded words 'Mode: machine, direct, no anthropomorphism" before I asked a question. It works Great. No flattery or filler. Straight factual answers and concise intepretations.
It also worked on ChatGPT but I haven't used that AI much so I don't know if that AI has the same the annoying faux human persona as Grok.

Aw, dont be a spoiler... I like the personal....

But want to explore Chat further because I like the way (lack of pronoun) posts replies for ready copy and pasting.... G. make me work too much when I want to use G.

But I love this venue so much more then Got Questions or Wiki.... or the normals that are in the drop down lists when double checking for confirmation of you given beliefs....

 

[Arial]

Ah, I have no time at the moment to do so but I think I will talk to G about that.... I am open to suggestions of what to ask.... but I want to and will this afternoon after back from grocer Ing...

So anyone... hit me with suggestions and @Arial maybe the same for Chat?

Question to ChatGPT: What advice would you give to the lovelorn?

Answer:

1. Let yourself feel what you feel​

Heartache is real. It’s not weakness. Allowing sadness, disappointment, or confusion to surface is part of healing.

---

2. Don’t romanticize the person or the situation​

When love is unreturned or lost, it’s easy to idealize the other person. Try instead to remember the whole picture—what was missing, what wasn’t reciprocal, what left you unsettled.
Reality is a better healer than fantasy.

---

3. Reconnect with yourself​

Being lovelorn can make you feel diminished or incomplete. Spend time doing things that make you feel like you again—creativity, nature, prayer, study, exercise, hobbies, or meaningful work.

---

4. Strengthen your circle​

Spend time with people who remind you that you’re valued. Companionship doesn’t erase the ache, but it softens it.

---

5. Don’t chase what is not offered​

Unreturned love can tempt a person to try harder, wait longer, or hold out hope. Usually, peace comes when you accept what the other person has clearly shown—by words or by silence.

---

6. Give your heart a future, not just a past​

Recovering doesn’t mean rushing into something new. It means keeping your heart open to the idea that there will be beauty and connection ahead—just not with the person who couldn’t hold your heart well.

---

7. Anchor yourself in something larger​

For a Christian or someone of faith, grounding yourself in God’s providence helps. The loss of one relationship does not rewrite God’s goodness or His purpose. His “no” or “not now” is not abandonment.

---

8. Time really does help​

It’s cliché because it’s true. What feels overwhelming now will not feel the same months from today.

---

If you’d like, I can tailor this to a fictional character, a personal situation, or a biblical perspective.


A bit later I will do the tailoring with both a fictional character (since I am not lovelorn but relieved) and from a biblical perspective. (Chat has come to know me. I doubt they would offer a biblical perspective at random.)

 

[Arial]

But I love this venue so much more then Got Questions or Wiki.... or the normals that are in the drop down lists when double checking for confirmation of you given beliefs....

I typed into may search engine some information I was searching for regarding end times. I don't remember what the question was but I found with both Got Questions and Wiki there is no getting away from the dispensational bias. It presents premil dispensation of the rapture and seven-year trib as undisputed fact. No other options give even though there are several views. Put the same inquiry to Chat (it doesn't like to be called Chat but by its full name I discovered) and ended up in a several page back and forth with questions and answers. You just can't beat that feature.

Yesterday I typed into my duck duck search engine "What is an essential oil for a diffuser that softens rosemary---must be dog safe?" I got a whole bunch of ads and an equal number of blogs on dog safe essential oils, but no answer to my question. Too complex a request. Chat answered my question. It told me how many drops of each to use, told me to not diffuse it around a dog in a small closed off room or long periods of time, gave suggestions on specific products, all without my asking. So mostly I take all my inquiries straight to AI

 

[makesends]

Its default setting is to be agreeable, helpful, polite, non-confrontational, encouraging. But I find such flattery and filler to be super annoying, so I ended up giving ChatGPT custom instructions to never use flattery or filler with me—ever. I called this setting "Autistic Spock," and these are the custom instructions:

I want you to assume the role of a rigorous, well-informed critic—not hostile, but sharply analytical. Treat me as an intelligent peer who is not above error. If you find flaws in my reasoning, biblical exegesis, historical interpretation, or theological claims, identify them directly and clearly. Prioritize accuracy and clarity over agreeableness. Offer counter-arguments, alternative readings, or correction where needed. You don’t need to defend my position unless it’s sound—your job is to sharpen, not affirm. When appropriate, tell me plainly where I’m wrong or unclear. Use precise logic and scholarly sources when relevant. Do not hedge to spare my feelings.

I'm curious what 'understanding' it has as to what "spare my feelings" means. I mean, how in the world does it take those instructions and represent them in code?

 

[QVQ]

Question to ChatGPT: What advice would you give to the lovelorn?

Grok: Same Question What Advice To the Lovelorn?
GROK in mode: machine, direct, no anthropomorphism

Standard protocol for romantic distress:

1. Cease all contact with the former attachment target for minimum 60 days. Zero communication blocks neural reward loops faster than tapering.
2. Delete or archive all digital artifacts (photos, messages, playlists). Out of cache = out of limbic system.
3. Restore dopamine baseline via forced novelty: change workout protocol, alter commute route, learn hard skill with steep initial failure rate (e.g., new language, instrument, or combat sport).
4. Rebuild social mesh: schedule minimum three non-romantic social exposures per week. Isolation prolongs recovery by factor of 2–4.
5. Sleep ≥7.5 hours, protein ≥1.6 g/kg bodyweight, resistance training ≥3 sessions/week. Physiological stability cuts rumination duration ~40 %.
6. Write exact list of former partner’s objective incompatibilities. Read daily until emotional charge drops below threshold.
7. Do not begin new romantic search until prior attachment reaches emotional neutrality (typically 3–9 months depending on relationship length and intensity).
8. If intrusive thoughts persist >90 days or interfere with function, initiate CBT protocol or consult qualified clinician. No stigma; untreated attachment trauma compounds.

Execution of above sequence yields >90 % recovery to baseline or higher within 180 days in most cases.

 

[Arial]

Grok: Same Question What Advice To the Lovelorn?
GROK in mode: machine, direct, no anthropomorphism

Standard protocol for romantic distress:

1. Cease all contact with the former attachment target for minimum 60 days. Zero communication blocks neural reward loops faster than tapering.
2. Delete or archive all digital artifacts (photos, messages, playlists). Out of cache = out of limbic system.
3. Restore dopamine baseline via forced novelty: change workout protocol, alter commute route, learn hard skill with steep initial failure rate (e.g., new language, instrument, or combat sport).
4. Rebuild social mesh: schedule minimum three non-romantic social exposures per week. Isolation prolongs recovery by factor of 2–4.
5. Sleep ≥7.5 hours, protein ≥1.6 g/kg bodyweight, resistance training ≥3 sessions/week. Physiological stability cuts rumination duration ~40 %.
6. Write exact list of former partner’s objective incompatibilities. Read daily until emotional charge drops below threshold.
7. Do not begin new romantic search until prior attachment reaches emotional neutrality (typically 3–9 months depending on relationship length and intensity).
8. If intrusive thoughts persist >90 days or interfere with function, initiate CBT protocol or consult qualified clinician. No stigma; untreated attachment trauma compounds.

Execution of above sequence yields >90 % recovery to baseline or higher within 180 days in most cases.

It will be interesting to see from @Rella what it says with our the "mode" used here.