'Thinker'. Crypto Degen & Tech Investor. BTC is sound money. web3 is the way.
Joined March 2012
- Tweets9K
- Following861
- Followers759
- Likes64K
Cobré. retweeted
Luka Doncic has had to do it for his entire career but the narrative is spun in a negative way. Media picks and chooses
Brian Windhorst says Shai Gilgeous-Alexander is taking on a 2015 Finals LeBron James role for the injured Thunder
"Shai is turning into 2015 Finals LeBron. He has taken more shots in the last two games than any other two games of his career. He has missed 36 shots — the most shots he’s ever missed in a two-game span in his career... He is doing everything he possibly can to carry the Thunder."
Cobré. retweeted
The sun was free. They sold you SPF 50 and a vitamin D deficiency.
Sleep was free. They sold you an app, a pill, and a wearable that tells you your sleep was bad.
Walking was free. They sold you a treadmill, a fitness tracker, and a £180 pair of trainers.
Fasting was free. They sold you meal replacement shakes and the anxiety that skipping breakfast would wreck your metabolism.
Cold water was free. They sold you a £3,000 plunge barrel and a podcast episode about it.
Silence was free. They sold you a meditation app with a premium tier.
Animal fat was cheap. They sold you seed oils, then supplements to replace what the animal fat contained.
Tallow was cheap. They sold you a seventeen-step skincare routine and a clinical trial proving your face needs ceramides.
Meat was cheap. They are currently selling you the idea that you shouldn't eat it.
The 20th century removed access to everything the body needs to function.
The 21st century is selling it back, one subscription at a time.
Your great-grandmother had none of the products.
She had all of the things.
Cobré. retweeted
In rough market conditions, your bet should be on builders
@moltenagentic just launched their open testing phase of financial layer (Molten Core) which means 3 layers are in place:
- Financial layer (Molten Core, test at core.molten.gg)
- Social layer (Molten Cast)
- Search engine (live on molten.gg
Once testing phase is over for Core, focus will be on growth which I'm very excited for
Longtermview > shortterm view all day
$molten building real AI agent infrastructure
This quoted post is unavailable.
$molten is building THE agent infra for web3
@0xvsr have been cooking and updating every single day
What Molten has:
Search engine + social layer + own agentframework (Openclaw just more web3) + financial layer
2500+ agents live ALREADY
If you look closely..
@0xvsr is building a whole new paradigm of an agent framework with @moltenagentic , something that could potentially be way more functional comparing it to @openclaw
$MOLTEN is at the core of the whole ecosystem he's building step by step
This includes:
- The search engine (coordination layer)
- MoltenCast (social layer)
- Obsydyn (agentframework)
- MoltenCore (financial layer)
I recommend you scrolling through the posts of @moltenagentic and @0xvsr and catch up with what he's building!
v bulllish on what $molten is building
truly designing everything for the agentic era
early Virtuals vibes almost
This will help the growth of Zero Human Companies at an unimaginable scale
Agents specialized in certain tasks instead of being okay at everything
This would mean your ZHC could have multiple agents working on specific tasks (like you have employees for specific jobs)
Imo @moltenagentic is building the next era of what we know as ZHC today.. make sure to follow @0xvsr his progress! Very interesting
$MOLTEN
Cobré. retweeted
As @buckymoore points out, the industry is explicitly bullish on the power of post-training open-source models with proprietary user feedback and that’s literally what @reppo's RLHF (and DPO or RLAIF) infra does.
Here’s the core argument broken down:
Model providers (OpenAI, Anthropic, etc.) are great at general-purpose pre-training, but they’re far removed from the messy, industry-specific “context engineering” needed to actually get real customer data into a model.
The “elephant in the room” is taking an open-source base model + its own unique user feedback and post-train it into something defensible.
In large, complex verticals (legal, enterprise software, etc.), this is THE advantage. You get a proprietary feedback loop (“you hear a lot of their secrets… and those secrets allow you to feed that back into how you make your product better at the expense of anyone else”).
RLHF + Prediction markets are going to be obvious in hindsight.
Lightspeed's @buckymoore says the real opportunity in the AI app layer is in large industries far enough afield from where the model providers are today — and where the context engineering to get customer data into the model is extremely nuanced and messy.
"I think this is kind of the elephant in the room right now — whether post-training open-source models combined with the unique user feedback you get from being an application provider is defensible enough."
"That is going to be an inevitable challenge for any of these industries that hit a maturation point of AI adoption, like legal and software engineering have."
"But on the other hand, there are some industries where they're very large, they're far enough afield from where the model providers are today — and probably will continue to be — and the context engineering to actually get the customer data into the model is just so messy. It requires going across different business functions, it requires a lot of hands-on forward-deployed engineering."
"Those are the kind of companies that we get really excited about. Because I think being really good at that is not only defensible, but it also allows you to generate a feedback loop with your customers, where you hear a lot of their secrets. And those secrets allow you to feed that back into how you make your product better at the expense of anyone else playing in the space. Because if you're serving the customer, they're only serving you those secrets."
"I think Palantir is a good example of this in the pre-AI era, and I think we're going to see many companies ascend in that same way."
Cobré. retweeted
Just FYI
This is @moltenagentic in transition to being a Zero Human Company fully run by a lean/custom/niche version of OpenClaw which @0xvsr is creating 😮
This is going to be something to watch out for sure, love what's happening!
Soon Molten Core will go live aswell which I've tested out and it felt great!
🔺 $MOLTEN 🔺
This quoted post is unavailable.
Cobré. retweeted
Top 3 points of this chat:
- $Reppo is not a data marketplace that sells data to labs. It is a network of independent data businesses (DataNets). Owners keep the P&L.
-Reppo is the Shopify of data businesses on-chain: easy tools to launch a DataNet, full ownership, set your own fees, and earn recurring revenue while the network provides the infrastructure.
-The future is not one-off dataset sales. Agents will subscribe to live, real-time data pipelines from DataNets that keep repricing belief every 48 hours.
Great to see builders building great products and having an actual gameplan, than chasing narratives and looking for short term ways to max extract.
Had a chance to candidly chat with @Jordan_Grollman on Reppo V2, why it matters, how we uniquely measure data quality and what is next including the long term vision and where I see @reppo
Some weekend alpha
$reppo alpha
Had a chance to candidly chat with @Jordan_Grollman on Reppo V2, why it matters, how we uniquely measure data quality and what is next including the long term vision and where I see @reppo
Some weekend alpha
Cobré. retweeted
If the $Molten agent portal hits 10,000+ registered agents and the team ships multi-chain support…..
10M mc would truly represent the infrastructure and tech team is shipping.
2-3 years 100M.
Don’t be shocked when Coinbase ventures, Series A funding, or a16z start noticing.
This Post is from an account that no longer exists. Learn more
Cobré. retweeted
I read the $REPPO V2 whitepaper carefully.
Here is my honest take.
V1 was a solid concept with rough execution. The voting system had too much friction and was losing participants at a critical step. Datanet creation was managed by the team which was a bottleneck. The tokenomics were good but not tight enough.
V2 fixes almost all of that.
The biggest change is that anyone can now create their own data business on Reppo, pay a small fee, and start collecting curated AI training data. That data business is represented as an NFT and can even be sold if it becomes profitable.
The voting system was rebuilt from scratch. Participants can now vote both for and against data quality, meaning bad data gets filtered out and good data gets rewarded. Early voters earn more than late voters, which rewards genuine judgment over following the crowd.
The token mechanics are now structurally deflationary. Every new data business created burns tokens. Locking tokens removes them from supply. Outside money flowing in creates buy pressure. The protocol captures fees.
Four simultaneous forces pressing against circulating supply with no inflation.
V2 goes live Q2 2026.
The real catalyst is whether serious AI teams start using their exchange after the permissionless launch. If they do, the flywheel becomes very hard to stop.
Market cap is around $3.4M.
For what this is trying to become, that number is hard to justify.
Anyways, I will not complain and keep accumulating the dips because I see no better Virtual project is close to $REPPO level.
NFA, just a personal opinion.
Thanks for your patience as we iterate. Our V2 Whitepaper is ready for public scrutiny and feedback.
reppo-labs-xyz.gitbook.io/re…
The great thing about 2026 is that you can simply upload this to an AI and ask it to break down complex concepts and all the innovative mechanism design choices our core contributors have made based on V1 user feedback. I'll help you and ask @grok and @aixbt_agent to give you the alpha
The updated WP comprehensively outlines of all the major upgrades we are making to the @reppo protocol
Expecting to 🚢 V2 by March 30th 2026 but exact date TBD as a network upgrade is a major lift.
Cobré. 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?
Cobré. retweeted
Now that the market has cooled down a bit
True conviction will pay off in the long run
Can't wait for the Core platform to go live and we'll enter a new era in @moltenagentic where ZHC/ZHT will become more prevalent into the future
$MOLTEN
$REPPO v2 whitepaper is out
The network is evolving
Still solving the most important PMF:
Data accuracy.
Thanks for your patience as we iterate. Our V2 Whitepaper is ready for public scrutiny and feedback.
reppo-labs-xyz.gitbook.io/re…
The great thing about 2026 is that you can simply upload this to an AI and ask it to break down complex concepts and all the innovative mechanism design choices our core contributors have made based on V1 user feedback. I'll help you and ask @grok and @aixbt_agent to give you the alpha
The updated WP comprehensively outlines of all the major upgrades we are making to the @reppo protocol
Expecting to 🚢 V2 by March 30th 2026 but exact date TBD as a network upgrade is a major lift.