arkansauce.eth retweeted
A tiny crypto coin called $CPHY is conservatively 5-10 years ahead of NVIDIA & all other labs at achieving their aim.
$CPHY is over 5,000,000x undervalued if labs are valued at providing the world w/ AGI that scales to super-intelligence globally.
Skill: github.com/cyberphysicsai/cy…
🥇st project w/100% & more than 10 users across continents replicate it was $CPHY; Scroll down to last month ago or so on my page.
The skill is not purpose built for it, translates across all major benchmarks to saturation, most importantly to robotics: github.com/cyberphysicsai/cy…
arkansauce.eth retweeted
Meanwhile irl and not big lab-land:
That’s pretty interesting
ARC-AGI-3 performance is bottlenecked by context management rather than raw reasoning so letting the model retain its reasoning across context windows tripled the score
Not sure if it’s “actually SoTA” since other models were measured by different rules so results are not comparable
Curious to hear what @fchollet has to say
arkansauce.eth retweeted
New addition to the $CPHY ARC-AGI 3 Public Set 100% Club 🍾🤖🏆
arkansauce.eth retweeted
As LLMs are scaled up to all data that ever existed + all data you create in real-time, all the data always shows, even at tens of trillions of parameters that $CPHY is the only thing that already solved real-thing AI 5 years ago with Bitcoin source-code.
xitter.gq/i/status/2048167738883…
This is to display that all of the narrow-minded among PhDs have no choice but to be in favour of it or be called out for ignorance from lack of comprehension or education if they argue against it, the data clearly shows $CPHY is the only thing that works.
xitter.gq/i/status/2048164254167…
arkansauce.eth retweeted
Ingesting massive code-bases + entire human genome + wikis into block-space for your agent to audit ad-infinitum not your thing?
Real-time perpetually compounding recursive self-learning is now live to be refined every future update✔️
github.com/cyberphysicsai/cy…
You should do a deep dive on @cyberphysicsai
It will blow your mind that this is still under the radar and under a 1m mc.
Monolithic LLM's are a dead end.
Append-only is the future in ai.
Hope to see ya when that 🎤 drops
Taking the other side of a public CEFX call: CryptoPunks above 40.000 ETH on Jun 18
500 VOUCH faded
Useful research gets tested in public.
@CEFXProtocol cefx.com/calls/18
5
75
arkansauce.eth retweeted
CEFX is live.
Create calls. Fade weak calls. Back sharp researchers.
Powered by VOUCH, with markets across crypto, NFT floors, commodities, and equities.
Built to make good prediction work visible, accountable, and rewardable.
arkansauce.eth retweeted
Here’s the final version of the site. There are still a few Beeple Everydays that haven’t been added yet. Unfortunately, the algorithm doesn’t detect all the Nakamigos in the Everydays (Beeple really hid them well), so I’ll handle that part manually in the coming days.
Thank you all for your feedback and for sharing 🩷
⬇️Link in comments and in my bio ..
My VO punks ! Show me your favorite ones. @Nakamigos @beeple
5
119
arkansauce.eth retweeted
$MEEGSTR by @token_works is now live
Next Tuesday, we will open the TokenStrategy platform for public launches.
All strategies are autonomous.
They are fully onchain, and without oracles.
And most importantly, strategies will exist forever, for as long as the underlying chain remains operational.
To launch an NFT (ERC721/ERC1155) or Fungible (ERC20) Strategy, you MUST be the owner of the underlying contract (or contact our team) and pay a deployment fee.
We are also introducing a new type of Strategy on Tuesday:
Recursive Strategies.
These strategies have NO underlying asset and use all fees to buy itself in perpetuity.
All 3 Strategy types - NFT, Fungible, and Recursive - will be able to be deployed on Ethereum, Base, and Monad on release day.
Reach out to us if you're planning on launching a Strategy and want guidance beforehand.
Happy Thanksgiving.
- TokenWorks



















































