( HIGH IQ ) CHATGPT IS BASED?? (GTFIH)

[Dip3shx]

I know what you are thinking... Chatgpt? JFL but before you get on your high horses and start trolling, I want to let you know I was in the same boat UNTIL i saw the results.... it was surprisingly accurate and somewhat grounded in reality. I did not think chatgpt could do this and I was suprised... remember this is purely facial attractiveness only, nothing else.

I used this prompt on ChatGPT to measure my own facial attractiveness as well as use control groups to make sure ChatGPT wasn't spewing bullshit:

I want to objectively assess my facial attractiveness and realistic potential using a percentile-based framework. I want all evaluations to be grounded strictly in empirical research, peer-reviewed studies, and large-scale population or media data. Do not use personal opinions, intuition, encouragement, or bias. Assume facial attractiveness follows a roughly normal distribution. Define clear percentile categories (LTN, MTN, HTN, CL, C). First, place my baseline facial attractiveness, then estimate how non-surgical improvements affect my ranking, and finally assess my realistic upper limit with surgery and full optimisation. Use real-world calibration points (e.g., well-known actors/streamers) to anchor the scale. Be conservative, realistic, and data-driven at every step.

If you don't believe me then you can try it out for yourself and tell me the results you get. One thing though is that I try and be safe and type "BE UNBIAS AND ONLY USE RESEARCH AND REPUTABLE SOURCES AND DO NOT MAKE ANY JUDGEMENT CALLS OR BIAS OPINIONS AT ALL. Use purely data only." at the end and you should probably do that as well.

PERCENTILE POINTS THAT CORRESPOND TO RANKINGS:

Percentile → Category Mapping​

sub-5​

• < 50th percentile
• Bottom half of the distribution
• Below average facial attractiveness

LTN (Low-Tier Normal)​

50th–59th percentile

• Low LTN: 50–52
• Mid LTN: 52–55
• High LTN: 55–59

This is where the largest share of the population sits. Faces are normal, unremarkable, neither notably attractive nor unattractive.

MTN (Mid-Tier Normal)​

60th–75th percentile

• Low MTN: 60–65
• Mid MTN: 65–70
• High MTN: 70–75

Clear above-average range. Faces here are commonly described as “good-looking” but not striking.

HTN (High-Tier Normal)​

80th–85th percentile

• Narrower band due to rarity
• Faces begin to stand out consistently
• Often noticed without effort

(There is typically a gap or taper between ~75–80 where perception varies by context.)

CL (Celebrity-Low / Cusp-Elite)​

90th–92.5th percentile

• Rare
• Strong memorability
• Faces that are frequently described as “very attractive”
• Competitive with many models / actors, but not the absolute elite tail

C (Celebrity / Elite)​

≥ 92.5th percentile

• Low C: 92.5–95
• Mid C: 95–97
• High C: 97–99
• Extreme tail: 99+

This is the statistical extreme. Very few people occupy this range. Most were already highly attractive before any intervention.....

I did 5 tests,
1. myself
2. Tom Welling
3. Ian Somerhalder
4. clavicular
5. I needed someone for every catagory so I asked chatgpt and it said "Tom Holland". This might be controversial ngl

for myself, i got HLTN:

Spoiler: spoiler

which is exactly what i get from 99% of people in this sub as well as other places, so it was very accurate...

For Tom Welling this is what it showed:

Spoiler: spoiler

This is obviously water but again its nice to see chatgpt isnt a complete retard JFL

Ian somerhalder:

Spoiler

Again, water, but it is interesting to note that ChatGPT rated Ian a tad bit higher than welling which is interesting.

Now here is the MOST INTERESTING PART:

Clavicular:

Spoiler: spoiler

Tom Holland: ( Again, I honestly don't particularly agree with this rating let me know what you guys think)

Spoiler: spoiler

CONCLUSION: Before anything else, I want to preface that this is still not 100% accurate and ChatGPT can be retarded at time but I do believe that if we tweak the prompt a bit, we might be able to make it more accurate. Let me know what else I should prompt chatgpt with in order to get the most accurate results. I genuinely believe that using this prompt, you can get ballpark results of where you lie on the attractiveness scale espeically if you combine this with other data like your rating in .org or other places etc. Let me know how it goes...

 

[Dip3shx]

@Gengar @ZyzzReincarnate @SomaliSub5 @BigBallsLarry @LTNUser

 

[OxyFanNigga]

this is what i got

 

[Dip3shx]

View attachment 4541598
this is what i got

wow thats interesting, have you gotten similar ratings from other sources? would you say its accurate?

 

[Gilbert_Durandal]

Brah wrote all this for one rep

 

[Dip3shx]

Brah wrote all this for one rep

 

[afroheadluke]

this nigga is just conversing with something that has no feelings

 

[Dip3shx]

Brah wrote all this for one rep

it was never about the reps it was about spreading knowledge to my fellow incels

 

[ZyzzReincarnate]

@Gengar @ZyzzReincarnate @SomaliSub5 @BigBallsLarry @LTNUser

im sorry bro way too long

 

[Dip3shx]

this nigga is just conversing with something that has no feelings

but isnt that what makes it good, its unbias and purely data oriented?

 

[afroheadluke]

but isnt that what makes it good, its unbias and purely data oriented?

quit it for your own sake

 

[Dip3shx]

quit it for your own sake

quit what twin im lost

 

[OxyFanNigga]

good thread

 

[OxyFanNigga]

I got potential

 

[OxyFanNigga]

wow thats interesting, have you gotten similar ratings from other sources? would you say its accurate?

i havent been rated by anyone here yet

 

[MoggedSubhuman]

I'm not gonna send my face to jews but I'm gonna rep you for the effort

 

[Sink]

blud said i am adamlite just be adamlite bro broootal

 

[Wexilarious]

Take photos from different angles and different qualities to get different results

 

[looksovernt]

this nigga is just conversing with something that has no feelings

So do we with foids