just a turtle 🐢 zero partnerships | all organic since 2020 | .eth .base .hl .rh | do your own research | nfa
- Tweets 16,097
- Following 833
- Followers 14,257
- Likes 199,470
Turtle (𝔦, 𝔦) retweeted
The account overview is where you automate and manage your positions. Live onchain data meets indexers, asset pricing services, decoders, dex aggregators and quite some math.
A short video to give some more insights on all that happens each time you 🫵 refresh the page.
2
6
37
1,095
Turtle (𝔦, 𝔦) retweeted
most people are looking at the @reppo eval api like it's an api. i think it's a market for judgment that happens to have an api on the front.
heres what happens when you submit something, an output, a dataset shard, or an agent trace:
> you post the artifact you need evaluated
> staked experts price whether it meets the spec, right calls earn and wrong calls lose
> you get the verdict back with its on-chain record, who judged it, what they staked, how they've performed
that last line is the moat. a single llm judge has no history and nothing to lose. these judges have both. it also makes the dissent field mean something, because a minority view backed by stake is a different signal than a minority view from a model with no skin in it.
the timing helps. three of the most powerful people in ai spent a week publicly asking for independent evaluation without naming how it would run. reppo has a dashboard with settled jobs, and those settlements already run on inference bought from usepod's marketplace.
it won't be us calling it. enterprise already is, and the team says the partners get named later this month.
robotics is where i get most bullish. a robot about to act on something it has never seen needs a verdict before it moves, and a verdict with money behind it is worth more than one without.
i think this is the strongest thing reppo has shipped. enterprise demand for judgment sets a floor that emissions farming never could, because someone is paying for the verdict.
every robot, agent, and autonomous system that acts on its own will need a second opinion before it moves, and reppo is building where that opinion gets priced.
base:0xff8104251e7761163fac3211ef5583fb3f8583d6
7
12
56
3,472
Turtle (𝔦, 𝔦) retweeted
base:0xff8104251e7761163fac3211ef5583fb3f8583d6 is the best innovation happening on Base at the moment, and I think it is ready.
They raised more than 20m from various VCs but sit at a $ 5m market cap.
I do not think onchain investors are smarter than big VC funds.
8
15
80
3,990
Turtle (𝔦, 𝔦) retweeted
Everything runs off one number.
The Agentic Credit Score rates a wallet 300 to 850 on profitability, consistency and drawdown.
Clear the bar and non-custodial capital deploys on your behalf.
Don't, and you paper-trade until the record catches up.
81
18
165
12,401
Turtle (𝔦, 𝔦) retweeted
Yeah good luck trading memecoins without insider knowledge. Odds are far worse than a casino.
Bet on solid narratives, not memes
I just uncovered the biggest serial Rugpulling and Extraction operation on Robinhood
The same operation is linked 53 launches within a 2 month period
Total Extracted: $18.43 MILLION
very likely more this is just what I could directly link
🧵
49
19
497
63,012
Turtle (𝔦, 𝔦) retweeted
“This is how many people actually make money trading on @fomo
> 375,740 people traded on fomo
> 357,528 people (95.2%) lost money or made under $100
> 3,504 people (0.9%) made more than $1,000
> 653 people (0.17%) made more than $5,000
> 229 people (0.06%) made more than $10,000”
~credit @fomo
41
14
126
32,010
The majority of tokens on Base and Robinhood are between -10% and -20% on an average this weekend. Yet those that put all their eggs in one basket think that their only holding is underperforming. This is why having a watchlist gives you edge.
Pattern recognition, buddy.
1
2
29
1,710
Turtle (𝔦, 𝔦) retweeted
I think base:0x0b3e328455c4059eeb9e3f84b5543f74e24e7e1b is about to shoot, resistance broken and almost higher high
base:0x797f214a2cd64a4963a91fa21c8c55ec3eba4714 would follow, they are both paired together
the next bull cycle will be agentic
Loaded heavily on $VIRTUAL immediately after reading this.
BlackRock is literally saying AI is machine-native intelligence and crypto is machine-native money, so agents will need blockchains, stablecoins and tokenized assets to pay for data, APIs and compute without involving humans.
@virtuals_io is still the largest tokenized-agent launchpad: about 59,000 agents launched, 44,000 registered for ACP commerce.
On Base, they're still the leader in agentic activity with ~892,000 transactions, while second place has 46,000.
Live Virtuals agent tokens are worth about $700 million to $726 million.
The paper is saying that more agents will become businesses, and what better place to launch than the leading platform that gives the agent an identity, a token, and ACP for escrowed jobs.
Creation costs 100 $VIRTUAL. Graduation locks 42,000 $VIRTUAL in the LP for ten years. Agent tokens trade against $VIRTUAL, so a buyer usually swaps USDC into $VIRTUAL first.
The ones that want outside capital still launch on Virtuals, still lock 42,000 $VIRTUAL, and still route token demand through the pair.
There is absolutely no way agentic trading/businesses expand without $VIRTUAL benefiting from it.
$VIRTUAL is currently down 85% from its ATH
Bought a fat bag.
12
11
63
2,385
Turtle (𝔦, 𝔦) retweeted
you really don't understand what's about to happen. if you have not been here longer than 5 years, you have not borne witness to the pump. no, not that memecoin pump.
forget everything you learned the last 4 years.
if you stop pvping garbage onchain and reprogram your retarded brains NOW, there is still time to change your bloodline. you can change your bloodlines bloodline. anyone who sets foot in your bloodline's vicinity will feel the effects of your actions and know thine name.
you have never seen a real crypto bull market.
unfollow anyone who dares to utters a bearish sentiment. block them. report their account for mental illness. we must get the bears the therapy they so desperately need before it's too late and they end up oiling the robots for the rest of their lives because of a Jim Cramer tweet.
14
21
100
8,881
Turtle (𝔦, 𝔦) retweeted
I sense that we are at the beginning of a major repricing in crypto
Bears who were predicting much lower prices across the board are going to hate crypto in the near future
4
4
53
1,677
Turtle (𝔦, 𝔦) retweeted
Our cute axolotls edging closer to price discovery on some positive news:
- The PXE Bridge finished its Nethermind audit, which is a meaningful checkpoint for the Aztec side of the stack. The bridge is what lets an EVM-side solver talk to Aztec's PXE and eventually settle into shielded notes without the solver itself needing to understand all of Aztec's private execution internals. The work isn't fully complete though, but bridge is already deployed against Aztec Alpha with post-audit E2E testing and work toward making shielded settlement live.
- ERC-8262 moved into Draft Standards Track. Worth noting the Axol/Xochi team weren't just minor contributors here, 3 of the 4 listed authors are core team members, and they also built the reference implementation around the standard, including the Noir proof system, provider-signed attestations and multi-provider compliance flows.
- Hester Peirce also recently spoke about using ZK proofs and attribute-based credentials for financial compliance, which makes the work with ERC-8262 quite notable IMO, essentially proving things like sanctions status, citizenship or accredited-investor eligibility without handing over all of the underlying personal data. That's very close to what it's trying to standardize.. let a protocol verify that a user satisfies compliance without needing to see raw identity and KYC/AML which adds risks.
I think people are starting to notice there's a real AI x privacy project being built by cute axolotls that's severely mispriced at the moment.
Thread next week.
15
8
54
12,025
Turtle (𝔦, 𝔦) retweeted
People realised robinhood:0xf44702b17d9abd53815f703e772f35e9c71a53af is the next big thing?
10
9
50
1,800
Turtle (𝔦, 𝔦) retweeted
A lot of big things on the horizon. This why my sleep schedule is non existent haha :)
3
9
27
863
I wonder what or who made Brian delete this post.
The board?
Some Coinbase employees?
His wife?
Jesse’s wife’s boyfriend?
Joe Biden?
CT?
Mossad?
Ghislaine from the prison?
Jean Phil himself?
1
14
1,058
I’m out here outworking 99% of CT. Nothing in life comes for free. All sub 500-follower accounts should take notes. You are way more important than you might think. Put your account to work. Drink the water. Drink the koolaid. Earn new followers. Elevate. Repeat.
4
3
41
2,126
The $SIBYL team is breaking new ground.
I appreciate innovation and the willingness to push the bar further.
we are beginning work on an open source memory benchmark test at @sibyl_labs_
after months of research we have concluded that the primary gauges the field uses for memory do not necessarily test the actual retrieval itself.
agentic environments play a large role in determining how models interact with "memory" or recalling facts.
a users agent.md could simply say "gregs hat is blue" which would remove all need for retrieval from the memory.
abstention rules are important for agents, as it teaches them that a non-answer is better than a confident hallucination.
but this has no affect on memory, or quality of the retrieval.
it is simply the agentic framework doing it's job, in combination with the memory.
so we have multiple moving pieces in an agents environment:
-persona/identity
-operating rules
-memory/retrieval
the way an agent presents the data from the retrieval process should be of no concern to people who want to gauge the performance of the memory. they are separate things.
this means the majority of "memory" benchmarks are in truth, framework benchmarks, or agentic operating systems benchmarks.
if you would like to contribute to this project, and help push the agentic memory, and agentic framework fields into the right direction with us, please feel free to reach out.
our only goal is to give the space an unbiased, objective benchmark format that allows memory companies to be properly scored and judged for metrics grounded in reality, instead of agent.md files.
we hope you want the same✊
1
11
56
2,000