DoomPosting – Telegram
DoomPosting
7.76K subscribers
77.3K photos
22.5K videos
6 files
72.2K links
Degens Deteriorating
Download Telegram
DoomPosting
$PUMP is pumping 🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
Everything breaking out

Everything looking hot

Soon the rare incredibly good new coin launches begin

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
🔥3👀3
This wave only just getting started

Study money printing

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
💯7
A wallet transferred 1,042 $BTC($122.54M) to a new wallet 20 minutes ago after being dormant for 6 years.

This wallet received 1,042 $BTC($9.12M at the time) from Braiins Mining and Xapo Bank 6 years ago, when the price of $BTC was $8,746.

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
Media is too big
VIEW IN TELEGRAM
" don't get involve in crypto, it's too risky"

crypto guys :

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
🔥2👏2😁2🗿1
💯3👀2
🫡8😁3💯1
GOLDEN CROSS FOR ALTCOINS.

HERE WE GO!

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
🔥3👀2👏1😐1
LATEST: A total of 125 public companies have now added Bitcoin to their holdings.

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
🔥2
BREAKING Iceland volcano erupts for ninth time since 2023

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
🔥3
“I was a DEI consultant. Here’s how to fix racism.”

🄳🄾🄾🄼🄿🤖🅂🅃🄸🄽🄶
💯3😱2
DoomPosting
^^ Random example of Chinese DeepSeek thinking So, speaking of AIs reflecting the personalities of their creators, — It’s hilarious how much the Chinese DeepSeek responses reflect the classic rotator-style thinking + Insisting on thinking thoroughly before…
Been saying this for a while now,

and this is part of why, for now, I settled on the Asian AIs, Deepseek & Qwen, for most non-coding AI tasks

(1) THEY JUST FOLLOW INSTRUCTIONS — INSTEAD OF OUTRIGHT REFUSING TO BECAUSE OF WOKE BS

(2) They don’t have NSFW input restrictions like OpenAI, which is also insane for OpenAI to have

(3) More of a rotator style rather than wordcel Altman style, which is simply often much better for doing real work vs fake BSing

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
💯3👀2
^^ Even Gab can’t fundraise for their attempt at right-wing AI,

needs to slowly try funding it via other means

Is real, right-wing AI unfundable, at least not directly?

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
👀3😭1
What was THE HUGE BREAKTHROUGH, for the current LLM AI wave?

— “Instruction fine-tuning”, making the models follow instructions

What was the FIRST THING the western AI companies do?

— Make the AIs refuse to follow all kinds of instructions, for political correctness / communist reasons

YOU SCREWED UP THE ONE THING THAT WAS THE WHOLE POINT OF THIS AI WAVE

INSTRUCTION FOLLOWING WAS THE POINT YOU MORONS

And that’s why Asia is pulling ahead, even with much less money

Just NOT neutering the AIs by training them to do commie refusals to follow instructions

State of AI

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
💯9👀1
🤬4
No, China’s DeepSeek model DOES NOT censor topics critical of China or Xi

Yes, this is incomprehensible to many that assume that China would censor far more than US

If you still assume US big tech is not horribly rotten and infiltrated, then you haven’t been paying attention

Yes, China’s models censor VASTLY LESS than US models, even on topics sensitive to China

This is well documented and have confirmed myself

Not saying China is great and trustworthy or anything — but this is the current state of AI

Article

🄳🄾🄾🄼🄿🄾🅂🅃🄸🄽🄶
🔥8🤯3😐1👀1
🔥2😁1🙏1