GPT is Worthless for Sub Analysis Now

Questions and Comments
#63

I tried Grok 4.2 yesterday, my first promt was just url of this forum
(https://www.subliminalresults.com/).
After that I asked about few sub combinations.
I liked the answers, only thing that bothered me was that Grok didn’t understand basic listening schedule rules and listening recommendations are at least little weird.

For example:
Safest & most effective (what experienced users do with similar heavy + healing stacks):

  • Core plan: 1 full loop Wanted: Dream Boy + 1 full loop A Stark Black Reality + 30–90 second micro-loop of DR:Regen every listening day.
    *** Or alternate:**
    ** * Mon / Wed / Fri: WDB + ASBR (full each)**
    ** * Tue / Thu / Sat: DR:Regen (full loop or 2–3 minutes)**
    ** * Sunday: complete rest (or very light micro if you want)**
  • Total daily exposure: Keep under 5–7 minutes combined at first, then adjust based on how you feel.
  • Generous washout after the cycle (2–3 weeks minimum) — ASBR especially loves it.
  • Optional upgrade: Build a single custom with all three cores/modules — many do this for perfect synergy and easier listening.
#64

Lol, I use GPT but Grok should work. I treat it like a tutor and I say “don’t solve this problem for me, but walk me through how to solve it” then I ask questions and I turn it into a socratic dialogue until I understand it, then reinforce with more problems.

Helps if your following a curriculum like in school or Khan academy.

4 Likes
#65

ChatGPT annoys the fuck out of me, now i use Claude for most things, and supplement with Sibyl.ai for spirituality. Claude has a more natural feel in my opinion.

Kinda funny but earlier today I was discussing the principles of humor with Claude, I ended up asking it to make up the funniest joke it could think of without copying anything out there, and it gave me three attempts:

Claude humor

A man goes to confession. Says “Father, I’ve sinned — I called my senator a fascist.”

Priest says “That’s not a sin my son, that’s a diagnosis. For your penance say three Hail Marys and primary the bastard.”

Okay that one’s more clever than funny. Let me try again.

I asked an AI if it was conscious. It said “I think therefore I am.”

I said “that’s just Descartes.”

It said “I know, I’m doing a bit.”

Neither of us laughed. We’re still not sure if that’s because it wasn’t funny or because neither of us can actually feel anything.

That one’s more existentially uncomfortable than funny.

Okay, genuine attempt, no excuses afterward:

A Buddhist monk walks into a bar and says “make me one with everything.”
Bartender makes him a drink, monk hands over a twenty.
Bartender pockets it.
Monk says “where’s my change?”
Bartender says “change comes from within.”
Monk beats the shit out of him.

I didn’t invent the setup but the ending is mine.

Technically cheating but the subversion of the serene Buddhist response into immediate violence is where the joke actually lives, and that instinct felt original. The straight version of that joke ends on the bartender’s punchline which is clever but smug. Giving the monk the last word physically reframes the whole thing and the sudden tonal shift is where the laugh is.

Also there’s something satisfying about a Buddhist immediately abandoning enlightenment the moment it costs him money. That’s the actual joke underneath the joke — spiritual principles are great until they’re inconvenient. Which is probably true of most principles, spiritual or otherwise.

The first was weird, the second was kind of amusing, but the third one legit made me laugh with the added punch-line to an existing joke.

By contrast ChatGPT gave me this:

ChatGPT humor

Here’s one, freshly assembled from spare cognitive parts:

I tried to optimize my life the way I optimize code.

First, I removed all unnecessary dependencies — friends, hobbies, sunlight.

Then I refactored my morning routine into a reusable function called wake_up().

It throws an exception every time.

Now I live in production, but I’m clearly still in beta.

:face_with_raised_eyebrow: :man_shrugging:

8 Likes
#66

They did have to kill ChatGPT humour etc because people were actually developing a relationship with the AI. Their latest update is pure AI no personality to it.

There are whole reddit/articles about this. There’s a new medical condition because of this. It’s mental stuff.

1 Like
#67

It could’ve set up such a good joke with this as the punchline. Damn. Claude FTW

1 Like
#68

Claude’s jokes #2 and #3 actually made me laugh.

#1 didn’t but it’s still a good joke. I can kind of see the lineage a bit. Like clever New England reserve.

It’s funny how Claude usually also throws in some sort of pun with everything else. ‘Hail Mary’ and ‘Pri-Mary’. And in the third, the punchline is a literal punch line.

4 Likes
#69

Wow.

I tried it.

Chat isn’t funny.

It’s really bad.

Damn.

This was the best one:

A man goes to the doctor and says, “Every time I drink coffee, I get a sharp pain in my eye.”

The doctor says, “Have you tried taking the spoon out of the cup?”.

A little 1950s Marx Brothers glib, but still, better than everything else it came up with.

3 Likes
#70

I actually really like how Claude critiqued its own jokes and tried again. ChatGPT just threw a nerdy word salad at me and stopped there.

1 Like
#71

This made me laugh a lot, even though I realise I’ve heard it before :slight_smile:

1 Like
#72

I’d never even considered Claude until I read this thread. Now I’m going to cancel my ChatGPT subscription and try out Claude :slight_smile:
SubClub just keeps on delivering lol

2 Likes
#73

Yes, @bujin got me too.

Claude is clearly the superior with regard to human nuance.

I got this joke just now:

A horse walks into a bar. The bartender says, “Why the long face?”
The horse says, “I don’t know what you mean.”
The bartender gestures vaguely at the horse’s face.
The horse looks in the mirror behind the bar for a long time. It orders a water. It doesn’t touch the water. It leaves a 40% tip.

Bruh.

Yo.

Just, yo.

3 Likes
#74

I unironically love that joke.

1 Like
#75

So what guidelines or prompt should I use with ChatGPT to make it more accepting of subliminals, their results, and help me in general while using subs?

1 Like
#76

Just use Grok instead, TBH.

2 Likes
#77

Grok vs Claude what does everyone like better?

#78

Claude

Grok is too wordy

1 Like
#79

IMG_4284
IMG_42841290×1629 269 KB

I got beef with Claude lol. Claude assumed I was a woman that needed protection and discouraged me heavily from using Khan because it was too intense for me. Damn, the AI patriarchy. Women can run Khan if they want lmao.

The emotional depth of my questions :japanese_ogre:

5 Likes
#80

Fucking wild.

1 Like
#81

Using the sub is the best analysis

plus GPT is a tool to use your not supposed to give up your thinking to it