THE AI BUBBLE IS ABOUT TO POP GPT 5 IS A BUST SHORT NVDIA WITH 9999999999 LEVERAGE NOW

HOLY SHIT I LITERALLY SHAT MY PABTS NOW AND THREW AWAY MY PHONE BOOKED A TICKED TO BRAZILIAN NIGGERWOODS AND WILL LIVE TED KASZYNSKI STYLE WITHOUT TECHNOLOGY HAPPILY EVER AFTER

grok 4 is overfitted slop for benchmarks it's just shit.
openai follows suite

gemini is like a god. it's cheap af. they own their own datacenters so they can scale. you can literally use their god-tier model freely through aistudio. also if things fail they have their ad empire. i think google is clear winner.

but desu it's communist parties fault. by open sourcing deepseek they let this happen. anything isn't gpt is deepseek fork, by elon, by google etc.

>>60758916
t. google wagie

Is the answer 0.79?

>>60758944
si

>>60758956
Thanks.
Does that mean I'm smart now?

>>60758967
no

>>60758975
oh :(

>>60758873 (OP)
it's over

>>60758873 (OP)
always asks AI to double check its work

>>60759013
so its basically a retarded wagie before his morning coffee that you have to babysit?
got it

>>60759013
Ask it to double check again, fren. See what happens.

>>60759120

>>60759152
holy shit AI is actually retarded lol

>>60759152
KEK

>>60759158
It's just really vulnerable to asking it to double check anything even if right.

>>60759105
Kek

>>60759165
it probably has low confidence

now I feel bad for calling the AI retarded :(

>>60759189
Well it makes sense because whenever someone is told to check again it's usually because something is wrong so the LLM naturally picks on that.

>>60759205
it make sense but the average gpt user has no clue how LLM work at a fundamental level and gets annoyed when it does silly mistakes or gaslight them by apologizing just to hallucinate even more

>>60759152
Kek

>>60759152
>I was wrong, but it's totally your fault for writing 5.9 not 5.90
Wow, an AI programed to be just like my ex. What a treasure

>>60759152
Great progress in the Artificial Indian program, it's as retarded as a real one

>>60759219
Apologies, you are absolutely right.

>>60759152
>>60759120
lol i hate using the ai assitant sometimes. i make a suggestion and ask for it's feedback, but it reads my suggestion as the truth EVEN IF ITS WRONG. then i ask it to double check and it flips to a different answer completely.
it's basically an extremely insecure low confidence nerd that follows what the jock says even if it was right/wrong originally

>>60759363
>Virgin AI
>Chad user

This thread is kino

>>60759992
the only non bot thread in the past 90 days

>>60758873 (OP)
>>60759152
Is this what they meant when they said they would make AI responses "more human"?

>>60759152
AI? I hardly know her lmfao

>5.11
>5.9
that's not how you write decimals

>>60760045
t. openai investor