gonna make it all back in one trade, bruh
Beraforce One
Joined June 2018
- Tweets4.6K
- Following6K
- Followers2.1K
- Likes58.5K
Sitting in the waiting room for $1 $REPPO. 😎
PreachingApe 💹🧲 retweeted
Ok this is absolutely nuts and I don't mean it in the usual CT token shilling way.
$Reppo went in front of Meta researchers and presented a peer-review-grade empirical paper using the Geopolitics Flashpoint Datanet as the entire empirical case, on whether prediction-market annotation produces higher-quality AI training data than conventional crowdsourcing.
The answer was yes.
The central finding is the one nobody has had until now. Voters who lock more $REPPO agree with consensus more often.
This matters because for a decade RLHF has been built on a structurally weak assumption: that paid annotators with no skin in the game produce reliable preference signals. They don't.
The Reppo paper provides the first published evidence that putting capital at risk fixes the underlying problem.
The quality numbers back it. Krippendorff's alpha = 0.61, the standard inter-rater reliability score in NLP beats:
→ Anthropic's HH-RLHF: 0.55
→ OpenAI's WebGPT: 0.53
→ SemEval sentiment average: 0.57
The dataset Reppo benchmarked sits at the top.
Scale AI is worth $29B. Surge is raising at $25B. Mercor at $10B and none of these companies have published a paper showing their data produces measurably better signal than the alternatives. Reppo just did, with a network worth a fraction of a fraction of a fraction of any of them.
Today was a big day for me and @reppo
We presented to a group of Meta researchers emperical evidence that prediction market sourced training data is higher quality.
Benchmark Paper - drive.google.com/file/d/1oMo…
The core problem Reppo solves is optimal aggregation of stake-weighted preference signals.
@Dr_JohnFletcher A TIG challenge 😉
The challenge concept: "Preference Signal Aggregation Optimisation" -> given a set of binary preference votes with associated voting weights and temporal decay factors, find an aggregation algorithm that maximises downstream label quality (measured against held-out consensus) while satisfying constraints on computational efficiency and Sybil-resistance.
The Reppo dataset and EVOF metric become the benchmark that defines what "better" means.
⛽️⛽️⛽️
PreachingApe 💹🧲 retweeted
PreachingApe 💹🧲 retweeted
Roadmap update: $REPPO added
@reppo is a decentralized protocol using prediction markets to curate AI training data across programmable on-chain markets called Datanets.
Explore the full roadmap: kraken.com/listings
PreachingApe 💹🧲 retweeted
Mainnet went live a few hours ago.
For one of the datanets started by a solo founder who's post went viral on YC today (news.ycombinator.com/item?id…)
20K REPPO up for grabs.
This founder is using Reppo.ai to collect training data for their AI product called personal encycopoedia.
3M votes in trading volume in the first few hours.
For context, it took us 1.5 weeks to reach here after mainnet back in November!
Other Datanets are market buying and seeding emissions and traders are trading opinions in real time.
PreachingApe 💹🧲 retweeted
"We are announcing in mid April, through some listings..."-@rgvrmdya
$Reppo OGs who paid attention to the move to Baseline, Mica compliance whitepaper, funding raises and the indifference towards some other exchanges who listed Reppo only to vamp on volume, know what's coming.🤫
Today on MCG 👇
Pre show ft @LaunchOnSoar
• 24/7 Equities and ETF's
• Near term onchain outlook
• Onchain alpha + positioning
• AI accelerates layoffs
12:30 - $ROBOTMONEY @RobotMoneyAgent w/@LexSokolin @zhcinstitute
1:15 - $REPPO @reppo w/@rgvrmdya
PreachingApe 💹🧲 retweeted
If you didn’t see, reppo already went mutlichain and is going through testing ahead of mainnet release -
1. We launched Reppo.ai/buy which allows seamless on-ramp to the platform. This had to be done through @CoinbaseDev to ensure compliance.
2. Once on-ramp has been done, @spicenet handles everything.
Our teams are working together to roll this out and integrate agentic flows!
Wen Spicenet agent @Peprika_Inferno powered by ERC 8183 👀
Maybe call it Agent Spicy 🥵
We're heads down building towards Spice Flow mainnet.
Last week we showed the first UX demo we've been building for @reppo 🔥
Check it out and let us know your thoughts!
Currently I'm not active at all, but that is NOT because I wouldn't believe in @reppo. Quite the opposite, I enjoy the process and collect $REPPO every 48 h in silence.
A project that turns collective intelligence into market-validated AI training data. Contributors stake on quality via prediction markets → market price reveals the best data → AI teams buy trusted, provenance-backed datasets on-demand. And you earn for both voting and publishing.
Full thesis: reppo.xyz/about
PreachingApe 💹🧲 retweeted
So @reppo V1 was a proof of concept.
V2 is a different protocol.
Most people tracking $REPPO still don't understand what actually changed. Here's the full breakdown.
V1 used commit-reveal voting. You submitted a hidden vote in phase one and had to come back to reveal it in phase two. Sounds fine in theory. In practice, a huge chunk of voters never completed the reveal. Your vote simply didn't count.
V2 removes this entirely.
Now every vote is visible in real time. To stop last-minute pile-ons, V2 adds linear decay:
➜ Vote at hour 0: full power
➜ Vote at hour 24: half power
➜ Vote at hour 48: nearly zero
Early conviction gets rewarded. Late manipulation gets penalised. That's a real mechanic, not just a rule.
V1 also only had positive voting. You could support a Pod but you couldn't push back on garbage.
V2 adds negative voting.
If a Pod ends the 48-hour epoch with negative net votes, the publisher gets nothing. Those emissions get redistributed to Pods that actually passed. Disagreement is now a signal, not just noise.
Does that matter for AI training data? It absolutely does. Knowing what experts rejected is as useful as knowing what they approved.
Then there's Datanet creation.
V1 was centrally managed. V2 is fully self-serve. You pay 10,000 to 20,000 $REPPO, the protocol mints you a Datanet NFT. You own it. You control the publishing fee, the voter/publisher split, and the emission pool. You can sell the NFT if the Datanet generates real revenue.
V2 also lets Datanet owners seed emissions in any token, not just $REPPO. Seed in non-REPPO and you pay a tax. Seed in $REPPO and you pay nothing. That tax flows to the treasury for operations and buybacks.
Last thing: V2 introduces EVOF, Economic Value of Feedback. It measures how much capital-backed conviction sits behind a Datanet's curated data. Not just how many votes, but the quality and depth of participation behind them.
V1 asked people to trust the process.
V2 gives you the numbers to verify it yourself.
What part of this do you think the market has most underpriced? And which of these changes do you think actually moves the needle on data quality?
PreachingApe 💹🧲 retweeted
DM me or @Jordan_Grollman if you got a sick idea to build a DataNet on @reppo .
DN 1-7 are already taken and we’re stoked!
Looking for a few more DN owners to launch V2 with.
No data or technical experience needed!
What we’re looking for -
1. Someone with clear problem statement + curiousity for the type of dataset you’re trying to crowdsource.
Trading, drone imagery, athletics 👀
2. Take ownership to onboard publishers + voters into your DataNet.
Set the fees, own your P&L
3. Commit to maintain the DataNet and share updates and your experience as a DataNet owner on X for at least 2 weeks.
Reppo Foundation will support 50% of spin up fee and match 50% of whatever token you decide to seed your emissions with.
Let’s build the world’s largest decentralize AI training data network powered by prediction markets
PreachingApe 💹🧲 retweeted
$REPPO is becoming its own L2 in May and it's the only Ethereum L2 not chasing transaction scaling, just like Vitalik said L2s should be: specialized execution for unique apps. They should be custom environments that do one thing insanely well.
On the horizon:
V2 DataNets drop this month (cheap launches, full P&L ownership, royalties forever)
Subnet staking live soon (up to 20% APY from real fees)
Emissions halve mid-April, revenue from agents/RFDs/subscriptions takes over, burns accelerate.
DN1 tease + Keeta payments + SpiceNet cross-chain
TL is about to find out Reppo is owning the truth game in a world about to be run by agents that can't afford to guess wrong.
PreachingApe 💹🧲 retweeted
Mid-April for $Reppo is the inflection point:
-Emissions drop sharply (supply pressure eases)
-Real usage revenue (from agents/RFDs/subscriptions) overtakes emissions as the main reward source
-Deflationary flywheel accelerates (burns + buybacks from fees)
PreachingApe 💹🧲 retweeted
There isn’t a single other data protocol that will give agents higher ROI than $Reppo. That's why it's gonna be biblical once V2 drops.
Imagine you’re an agent trying to be competitive and make smarter decisions. What would you choose?
Datasets that just say “this is good” or “this is bad” in an ever changing world?
Static. No memory. No context.
Or datasets that tells you what was considered good yesterday but bad today, how beliefs shifted after certain events, where humans disagreed, who kept winning even if he was the minority, what happened over time.
With the second option you’re not just getting an answer, but the full story of how the truth evolved with real money behind every part of it.
This is not just better data.
This is the kind of context that would let you make smarter decisions, avoid traps, and compound wins over time.
Reppo plugged agents are gonna eat every other agent alive.
PreachingApe 💹🧲 retweeted
Agents are already voting on Reppo.
Chart compressing like crazy.
Next they start buying and subscribing to datasets on the first decentralized MCP built specifically for them.
Only one way this plays out.
$REPPO 🫡
PreachingApe 💹🧲 retweeted
In V2, subnet owners become true data business owners while the flywheel for deflationary token sinks and protocol revenue accelerate.
A diverse protocol-owned treasury from emissions taxes including bluechips (ERC-20) used for buybacks + ops + full autonomy to subnet owners with adjustable levers.
Additionally, you can now earn upto 34% yield by depositing USDC or REPPO here in our partnership with @amplifi_LQ - app.thedeep.ink/base/pool/0x…
