While I was on vacation, the team behind @thebunkerhood appears to have rugged the project. The $BUNKER token launched, liquidity was pulled shortly after, and the account has since rebranded to @renaisancexbt. I’m sorry I ever spoke about this project - I trusted their updates and I was wrong. If you can report this new account, please do it. There is should be no room in this space for people like this. P.S: photo from the vacation

Burned 9 bunker artifacts to receive 3 GTD slots. There ~150 GTD spots left in simple way to burn: 1 Common + 1 Rare + 1 Epic = 1 Citizens GTD mint spot After it you'll have to burn more: 3 Common + 1 Rare + 1 Epic Do not miss your enter to bunker

2 DAYS DEEP IN THE INK NFT MINTS. HERE'S WHAT I SAW. There are no ideas. Ink Broker, Ink StonkBrokers - StonkBrokers is a Robinhood Chain thing. Ink MACHINES, but on Robinhood we already have BTC Machine. And punks everywhere, ink_punk_ (X account suspended already), Onchain InkPunk. Just copy something that worked and extract money from degens. Why anybody will need a copy of one project on the other chain? Mints were a mess too. Old X accounts renamed into "new projects" to shill their collection. Free mint became paid in the middle of the mint. Per wallet limit changed. Price dropped when nobody was buying. What surprised me more is that @inkonchain never really touched NFTs on their own chain. On @RobinhoodApp it just happened by itself. Volume was there on Sunday and that's it. Right now the airdrop rumor is the only reason people are still there. After the drop I don't see why anyone comes back. Going back to @RobinhoodApp. The people's chain.

The @inkonchain NFTs volume getting hot! Gonna follow it as seems like liquidity moving there Bought some @KrakensOnInk with a floor of 4$ after freeming and volume of 7k Who knows, maybe volume from @RobinhoodApp NFT market moves to @inkonchain?

Wow, they finally decided to list base:0xb2000000000000000000004c27f6523082f41d01 And for sure they decided to list it in the timing when Robinhood grabbing people attention with cash-cat:native Pumped after it almost 200%, not bad

If you are new to $STONKBROKER State of the ecosystem, including what most threads get wrong. Every number read off the contracts, not the docs. Robinhood Chain, block 42,493,430. Two assets, not interchangeable. The NFT: 4,444 ERC-721s, each owning a real ERC-6551 wallet. $STONKBROKER: the liquid ERC-20. Currency of the system, not a claim on rewards. Holding the token does NOT earn you stock drops. Drops go to activated NFTs only. Want stock tokens, you need the NFT. Want liquid exposure, take the token. Getting in: 666,666 $STONKBROKER + ETH fee on the Anvil AMM. 10% on a swap, 15% to snipe an ID. Not a rounding error. And that fee is charged on a "market notional" that is supposed to come from a TWAP oracle. The oracle is not wired. It is a number set by hand, currently 4.427844 ETH. Activation is the real sink. Five tiers, and the top one costs 25x the base for 3.33x the weight. Per token spent the base tier buys 7.5x more reward weight than the top. Read that twice before you max out. Half of every activation fee burns, half goes to the protocol. Most threads leave that half out. Activation clears on transfer, no exemptions. Deposit into the AMM vault or post the broker as loan collateral and it clears the same way. Which is most of why there have been 13 loans in total, 8 still open, zero liquidations. Clock In v2 changed the model. Brokers now elect up to 3 tokens they want to be paid in, no election means ETH, and any wallet can crank the round. So where does the money actually go? Across all 769 completed rounds, 674 ETH distributed. $STONKBROKER took 61%. All ten tokenized stocks on the menu took 9.77% between them. NVDA got 18 ETH. GME got 1.09. And it is not flat. On a 40-round rolling average the share buying $STONKBROKER went from 23.5% to 68.2%. Of the elections on file right now, 67.75% of the next pot buys $STONKBROKER. 1,474 of 1,814 activated brokers have an election on file. They are choosing, not defaulting. 1,358 chose the protocol's own token. The distribution engine is mostly a buyback that brokers voted for themselves. The stock leg is the minority case. Two more things. 2,204 of 4,444 brokers sit in the AMM vault as inventory and cannot be activated. So "41% of supply activated" and "81% of brokers in wallets activated" are both true. Use both. Broker Box design RTP is exactly 90.0000%, recomputed from the table in source. Every roll pays something, and 90.02% of rolls pay back less than the ticket. Admin surface, since nobody says it. None of the 14 contracts is upgradeable. Every EIP-1967 proxy slot reads zero. Tier table, prize table, fee shares: constants or immutables. The fee router is genuinely ownerless. Credit where due. But one plain EOA holds every admin role across ten of them and deployed all fourteen. Code size zero. Not a multisig, not a timelock. On Aug 14 it cancelled round 455 and six seconds later withdrew 456.70 ETH, the largest pot in protocol history. Whether that ETH came back is not verifiable from available indexes, so I am not claiming either way. Per the docs: drops are marketing rewards, not dividends. No equity, no revenue share, US swaps restricted. DYOR.

Just read it and you’ll understand everything you need about the project Moving so fast and in the right direction With a burning of artifacts seems like we will have a supply shock

Been following @thebunkerhood for a bit and finally picked up one of their NFTs on Robinhood Chain. This is basically an upcoming game being built on Robinhood Chain, not just another PFP mint. You’ll be able to burn 3 eligible NFTs for a guaranteed spot in the Citizen mint, and that’s the one I’m waiting for next. After that: Citizens, jobs, resources, upgrades, quests and more inside the BUNKER. Robinhood Chain NFTs are getting interesting fast. Happy to have some skin in the game here.

3 hours left to @thebunkerhood GTD mint I have a WL spot so I hope few bunker tickets will be left for WL guys

civilization collapsed some mf still opened an IT ticket guess i found my role in @thebunkerhood 💀 what class did u pick? thebunkerhood.com/enter code: BUNKER

Robinhood Chain NFT meta is starting. I'm not gonna sit this one out and watch people post their free mint screenshots a week later lol Just applied for @FelineReserve. 3k spots, mine is already in the priority docket. Worth throwing an application in imo: thefelinereserve.com/?ref=dy…

Been looking at what's happening with NFTs on Robinhood Chain lately and @FelineReserve caught my eye. Not really because of the 3k supply or the cat stuff. The application itself is actually pretty clever. You have to file an application, post a pledge, go through an "examination", submit your wallet, etc. Basically they turned the whole WL process into part of the project instead of doing another follow + RT + tag 3 friends campaign. Could still end up being just another NFT collection, obviously. But at least they're trying to build some identity around it from day one. Robinhood Chain's NFT scene is still tiny, so I'm curious which projects actually stick. Applied for a seat. Let's see.

I am a reserve for @felinereserve

Developers will spend 20 hours saving $50 in AI tokens and call it optimization. Had an argument about this with my team today. The obvious take was: “Of course we should save tokens.” Sounds rational. I think it’s often a trap. The classic example is choosing the cheapest model because the price per token looks better. DeepSeek, for example, can be great when you’re processing huge amounts of predictable work: extracting data, classifying documents, summarizing, sorting. But give the same model a messy engineering task and the economics can flip very quickly. More retries. More supervision. More weird decisions you have to undo. More time spent figuring out why it went off track. Suddenly the “cheap” model isn’t cheap anymore. Because the metric that matters isn’t: cost per token It’s: cost per successfully solved task And we don’t really have a reliable benchmark for your work. Leaderboards are useful, but they don’t tell me which model is best for my codebase, my workflows and my constraints. The only approach I trust is testing from the top down: Start with the best model you can get. Learn what good looks like. Then move workloads to cheaper models and see where quality actually starts breaking. Going bottom-up has the opposite problem: you may spend hours fighting a mediocre result without realizing what a better model would've done in one shot. Someone on my team has been building an agent for weeks whose job is partly to reduce token usage. Maybe it works. But if engineering time costs more than the tokens it saves, did we actually optimize anything? There are much simpler ways to cut waste. Tools like Caveman and RTK try to reduce all the garbage CLI output agents consume. That’s the kind of token optimization I like: remove useless context before making the model worse. Optimize the outcome first. Optimize tokens second.

Me wasting the strongest model on one email reply. Then giving it a broken repo, 3 PDFs, browser tasks, and a half-dead workflow. That’s when Claude Fable 5 starts to make sense.

🚨🚨🚨 CLAUDE FABLE IS HERE RIP Opus. If this is real, half of AI Twitter is about to spend the next 48 hours pretending they knew this was coming. What are you cooking with Fable?

This launch quietly changes the AI market. Most users get the safe version. Governments, cyber teams, bio labs and critical infrastructure partners may get the less restricted version. Same underlying model. Different access. So the real question becomes ugly: who gets the powerful AI, and who gets the toy version?