Mr. kazz (Ø,G) ❖,❖
2.8K posts


According to the image below, it seems like Zama will go for a token sale before the actual launch of Mainnet {speculative} Rand basically confirmed that the token sale will be completely independent of their previous funding rounds. That alone opens up a lot of possibilities, maybe a fresh valuation, maybe a more open distribution, maybe something tailored specifically for creators and early users rather than investors. Then there’s Adam’s point about pricing creator reward tokens differently. If @zama decides to give creators a meaningful discount or early pricing advantage, that could reshape the incentives around building on FHE. Imagine creators or OG NFT holders getting in at a preferential level instead of being lumped in with late public buyers, that would be a huge signal of long term alignment. Nothing is confirmed yet, but the image here suggests they’re still shaping the mechanics and when a team is open to adjusting token economics based on community feedback, it usually means they’re aiming for a launch that people can rally behind. X: @zama #ZamaCreatorProgram

I want to talk about FHE a little more today and how privacy signing works... Not a lot of people realise this but FHE completely changes how wallets handle privacy. With Zama, your wallet can sign actions without ever exposing what you’re doing, not to the chain, not to the dapp, not to anyone watching the mempool. Traditional wallets sign raw transaction data,that means your amounts, addresses and intent are basically sitting in the open until they land on chain. This is why snipers, MEV bots and chain analysts eat so well. Zama actually changes this because the computation happens on fully encrypted data, your wallet only needs to sign an encrypted intent, then the chain never sees the plaintext, it only verifies the result. To explain private signature with an example, let's say you approve a swap, bid or transfer just like normal, but the payload is encrypted end to end. Your wallet doesn’t leak the what, who, or how much only that you authorised something. On chain, the @zama network computes over that encrypted payload using FHE. Validators confirm your logic, balances and rules without ever decrypting anything. It’s the same trust model as Ethereum just with confidentiality baked in. This means wallets can finally offer true privacy without special networks, mixers or sketchy bridging. You keep your existing chain, you keep your existing wallet. You just gain the ability to act on chain without broadcasting your life to the world. X: @zama #ZamaCreatorProgram


We are kicking off the Mainnet Mega Giveaway, the biggest coordinated community event in Monad so far, with 20 ecosystem partners united under one goal: reward the early believers. 🎁 1,000+ prizes ✅ USDT rewards ✅ NFTs & mainnet whitelist spots ✅ Bonus testnet points ✅ Partners Discord role ✅ Exclusive Memesteroid Discord role with mainnet rewards for those who interact with our app Holding any partner testnet NFT/token? Join our Discord, verify, and secure your exclusive role. Finish the 5 day tasks on the campaign page and secure your seat before Monad mainnet lights up. 🔗 Link in comment




People often assume that adding privacy to blockchain means building a whole new network, new tokens, new wallets and isolated bridges. That’s how projects like Monero and Zcash were built...strong privacy, but disconnected from the rest of Web3, but Zama is doing the opposite. Instead of starting from scratch, zama is using Fully Homomorphic Encryption (FHE) to bring privacy directly into the chains we already use. It lets contracts compute over encrypted data, so privacy becomes part of Ethereum, Solana and others, not a separate ecosystem. With what Zama is building, here is what that unlocks; >>Smart contracts can stay private without leaving their home chains >>Encrypted data stays compatible with public protocols >>Apps can mix open and private logic in the same environment Zama isn’t replacing zk or MPC, it connects them all into a single, usable framework where privacy is just another tool in the dev stack. This turns privacy from a standalone blockchain idea into a core layer of crypto infrastructure practical, composable and chain agnostic. With @zama_fhe, privacy isn’t a destination, it’s built into the journey. #ZamaCreatorProgram








𝐒𝐄𝐍𝐓𝐈𝐄𝐍𝐓 𝐉𝐎𝐔𝐑𝐍𝐀𝐋 𝐃𝐀𝐘 𝟐𝟖 Sentient recent tweet made mention of SPIN bench and here is a TL;DR of what it is all about... Most AI benchmarks test short tasks, like answering questions or simple reasoning. But real intelligence isn’t about quick answers, it’s about long term reasoning, coordination and adapting to other's goals. That’s exactly what SPIN Bench was designed to measure. Developed in collaboration with Princeton and UTAustin and presented at COLM conf 2025, SPIN Bench (Strategic Planning, Interaction and Negotiation) is a unified benchmark for testing how AI systems plan, cooperate and negotiate. It’s a new way to measure strategic, multi agent intelligence, the kind we see in real social and economic systems. Unlike most single agent tests, SPIN Bench adds multi agent environments and long horizon tasks where agents must; >>Plan over many steps. >>Predict others’ intentions. >>Adapt strategies dynamically. It’s built to reveal where reasoning fails and how models can improve through better coordination. SPIN Bench spans four domains that capture every side of intelligence which includes; >> PDDL Planning: long term goal decomposition. >> Competitive Games: strategy under pressure >> Co op Card Games: coordination with partial info. >> Negotiation (Diplomacy-style): alliance building and bluffing It’s a lab for testing how agents reason, communicate and adapt in complex settings. Each agent follows the same loop which is... Observe → communicate → act across controlled environments with shared metrics. This makes SPIN Bench a powerful lens into social reasoning, communication and multi agent coordination, exactly the frontiers @SentientAGI is pushing forward. X: @SentientAGI || @sentient_found

𝐒𝐄𝐍𝐓𝐈𝐄𝐍𝐓 𝐉𝐎𝐔𝐑𝐍𝐀𝐋 𝐃𝐀𝐘 𝟐𝟓 Firstly I want to say congratulations to Sentient for winning the "AI Startup or the Year" award at the Minsky 2025 Awards... I also read through the OML workshop track post made by Sentient few hours ago and here's what I learnt from it.... Today’s AI distribution model is broken. You either get closed models which is gated behind APIs, where transparency and local execution are impossible or Open models that are freely downloadable, but impossible to monetize or control, but Sentient is introducing a third path. That is why they introduced OML {Open access, Monetizable and Loyal model serving} OML is a new AI primitive that makes it possible to: >>Distribute models openly for local use. >>Enforce usage rights cryptographically. >>Preserve both transparency and sustainability. Think of it as open models with built in loyalty. Anyone can run them, but unauthorized use or monetization attempts can be cryptographically prevented, no gatekeepers, no centralized APIs just provable enforcement. To make this work, Sentient formalized two new security goals for open AI models: >> Model extraction resistance, this protects against copying or cloning. >> Permission forgery resistance, this ensures only authorized users can access premium model capabilities. All these are tailored to the white box challenge, once a model is public, how do you stop misuse without closing it? {I'll talk about this in another thread} Sentient’s research also proves fundamental limits, what’s possible and impossible when balancing openness, monetization and security. They explore everything from obfuscation based approaches to cryptographic model serving. And its not theory, they actually build it, take a look at OML 1.0, it introduces a practical system combining AI native model fingerprinting and crypto economic enforcement mechanisms. Together, these ensure models report back and remain economically loyal to their creators. This means builders can share models openly while keeping monetization rights and usage integrity intact, a sustainable path for open AI ecosystems. No need to choose between open source and survival anymore. OML sits at the intersection of cryptography, ML and mechanism design, opening a new research direction for AI distribution and governance. It could redefine how open models are shared, verified and rewarded in the coming AGI era. Reading through everything above will make you understand that @SentientAGI is building toward an open, decentralized AGI ecosystem and OML is a key piece of that foundation. X. @SentientAGI || @sentient_found

