Blur price

Let me start with an observation. In 2023, the VC circle suddenly began to talk frantically about FHE (Fully Homomorphic Encryption). At that time, I was quite confused: this technology was proposed as early as 2009, so why is it only becoming popular now? Later, I understood: it wasn't that the technology suddenly matured, but that the narrative needed it. ZK (Zero-Knowledge Proof) focuses on scalability, while FHE is packaged as "a supplement for privacy." But what really made FHE popular wasn't its relationship with ZK, but the market's shift from "Crypto Native" to "external narratives"—AI, RWA, and payment scenarios are where FHE truly shines. The three stages of FHE: First stage: Supporting role of ZK (2023) In 2023, FHE gained traction in the VC circle mainly because it was positioned as a complementary technology to ZK. The logic at that time was as follows: ZK solves scalability issues (Layer2, Rollup) FHE solves privacy issues (complex information encryption off-chain and on-chain) Sounds reasonable, right? But the problem is: the demand for FHE in Crypto Native scenarios is actually not high. In on-chain transactions, DeFi, and NFT scenarios, ZK is already sufficient. FHE's "encrypted computation" capability finds little application in purely on-chain scenarios. So during that stage, FHE was more about "conceptual heat, little implementation." Second stage: Expanding scenarios (2024) In 2024, the market began to shift from "native narratives" (Layer123, on-chain assets) to "external narratives" (AI, RWA, payments). At this point, the demand for FHE was truly activated. Why? Because these external scenarios share a common characteristic: they need to handle complex, sensitive, off-chain information. AI: Needs to train models and process data in an encrypted state. RWA: Needs to encrypt and protect user identity and asset information. Payments: Need to encrypt and protect transaction details and user privacy. These scenarios cannot be solved by ZK; only FHE can do it. Thus, FHE transitioned from being "ZK's supporting role" to becoming "the main character of external narratives." Third stage: Accelerated implementation (2025) Now, FHE is starting to be implemented in reality. The leading player in the field is Zama, with a valuation of $1 billion, having raised funds from a mix of traditional and Web3 VCs. Zama provides a complete technical foundation, similar to "AWS for FHE." But there's an interesting detail: Zama is unlikely to issue a token and is pursuing a public listing path. What does this indicate? It indicates that Zama's target customers are not retail investors, but enterprises, institutions, and governments—these customers need stable technical services, not token speculation. The project that has gone the furthest in implementation is Mind Network, which recently launched on-chain message encryption, supporting wallet-signed key generation for end-to-end encryption protection of user semantic information. This feature sounds very technical, but the actual application scenarios are quite clear: RWA and payments. For example, if you are doing RWA transactions on-chain, you need to submit identity information and asset proof, but you don't want this information to be public—FHE can solve this problem. The explosion of technical narratives requires corresponding scenarios. I have always believed in one logic: the explosion of technical narratives must be supported by corresponding scenarios. Technology without scenarios is just a PPT. Here are a few examples: Meme has revitalized NFT trading tools: Tools like Blur and Tensor were originally designed for NFTs, but the explosion of Meme has allowed them to find new applications. The influx of Sol users has made data service companies profitable: Node service providers like Helius and QuickNode have profited immensely due to the surge in Sol users. If RWA explodes, FHE will be one of the biggest beneficiaries: because RWA requires a large amount of data encryption and verification. Therefore, FHE's opportunity lies not in the technology itself, but in the scenarios it can serve. My advice: If you want to pay attention to the FHE field, my suggestion is: Don't just focus on Zama: Zama is not issuing tokens, retail investors cannot participate. Pay attention to implementation projects: Mind Network, Fhenix, and other projects that are genuinely working on applications may have airdrop opportunities. Look at scenarios, not just technology: FHE's opportunities lie in RWA, AI, and payments, not purely on-chain scenarios. Take it slow and steady. FHE is not an overnight wealth opportunity, but a long-term field that requires patience to wait for the explosion of scenarios. @zama_fhe #ZamaCreatorProgram

All the open weights models in mlx-lm. Gives you a pretty good landscape of open weights LLMs.