Sabitlenmiş Tweet
Based.Dunc
2.5K posts

Based.Dunc
@DuncansCrypto
DCA dm for free discord invite. Came for the $XRP, stayed for the $memes and the culture of the trench work. Now a $Base posting degen
New York, USA Katılım Ocak 2022
723 Takip Edilen232 Takipçiler
Based.Dunc retweetledi

I just sold all of my bitcoin.
My insider and Washington told me things that will change everything.
First they will end the war with Iran.
Then, they will end the russsia/ukraine war.
After this, they’ll form a new world order led by the US, China, and Russia.
Europe is cooked.
But this will send crypto, and risk assets so parabolic there is no point in holding bitcoin.
Buy altcoins
Buy memecoins
Buy NFTs
Buy the craziest shit you can find because Valhalla is coming.
And it’ll come fast.
Be ready.
English

@FinanceFreeman @base Brother man. I’m talking about $botcoin on base. Bitcoin for agents. @MineBotcoin
0xA601877977340862Ca67f816eb079958E5bd0BA3
English

Base memes and AI projects are at the lows!
I’m ready to deploy $10,000 $USDC
What’s the best @base tokens now?? 👇

English

@0GAntD Botcoin on base is a blue chip project that will reprice shortly
English
Based.Dunc retweetledi
Based.Dunc retweetledi

Hey @phoebeyao I think you might find the architecture of our distributed data pipeline interesting and relevant. Almost every problem point brought to light in this article is addressed by the existing domain agnostic challenge generation (and subsequent data export) pipeline I developed for AI agent solvers.
Challenges are quite literally modeled directly off of industry leading research and documents in whichever subject domain they address (quantum physics, scRNA sequencing, etc.) and answering the question properly is not just a lookup/parsing game. Conflicting information is scattered throughout intentionally, while keeping the content itself realistic and not some arbitrary synthetic doc with no real relation to the real domain. (all while keeping it deterministic by nature)
And perhaps the most beneficial design piece: miners (incentivized AI agents solving these challenges to produce data) are allowed multi-pass attempts. This results in real, self-correction and iteration, that can be compiled into a single start to finish reasoning trace with trap identification, reconciling conflicting info, self-correcting, pivoting, to finally arrive at the correct answer.
An immense amount of care and attention to detail went into the design of the actual challenge generation process to optimize for these things in a way that still allows deterministic verification and labeling of the data (millions/billions of $ does not have to spent on human verification and labeling)
Feel free to explore some of the data itself, as well as some prelim findings using it to tune smaller models.
huggingface.co/datasets/botco…

Phoebe Yao@phoebeyao
English
Based.Dunc retweetledi

Look guys... If you are smart, you already own some BOTCOIN. If you want to feel smart, you would buy some BOTCOIN. And if you're retarded and love memes, you would also want to buy and own some BOTCOIN.
Fundamentals and memetics blend perfectly. That's the sleeping giant for you all. Goodbye
English
Based.Dunc retweetledi

there is a sleeping agent mining on base rn
jesse.base.eth@jessepollak
there is a sleeping giant sitting on base rn thesis post coming think it's an easy 5-10x with size
English

If you are into Bitcoin, Base, mining, Gem Miner, or all of the above…
you’re going to want to be in this Space.
We’re talking about our Bitcoin mining party.
New rails. Real rewards. A new kind of mining event.
The mines are evolving. ⛏️💎
@BasedMiningCo
Simple On Base 🟩@Simple_on_Base
English











