Sabitlenmiş Tweet
What is Zama and What Problem Are They Solving?
@zama_fhe
Web3 is built on transparency, everything on-chain is public.
That includes:
• Inputs to smart contracts
• Outputs of computations
The logic itself:
This is fine for things like DeFi, but terrible for privacy-sensitive use cases like:
• Private voting
• Encrypted identity
• Medical or financial records
• ML inference on private data
There’s a growing need for confidentiality on-chain, without breaking the trustless model. This is where Zama comes in.
• What Does Zama Build?
@zama_fhe develops practical implementations of Fully Homomorphic Encryption (FHE), a cryptographic technique that allows:
• "Computation on encrypted data, without decrypting it."
This means:
• The user encrypts their data.
• The smart contract processes the encrypted input.
• The output is also encrypted.
Only the user (or intended recipient) can decrypt the result. No intermediate step ever reveals the data,not to the smart contract, not to the validator, not to the public.
• Finally, How Does Zama Apply This to Web3?
Zama is working on bringing FHE into blockchain environments (specifically EVM chains like Ethereum). Their goal is to allow developers to write confidential smart contracts, contracts that operate entirely over encrypted data.
With Zama’s stack, a Web3 developer can:
• Write a smart contract that accepts encrypted inputs.
• Use Zama’s tooling (like Concrete or TFHE-rs) to perform encrypted computations.
• Store and update encrypted state.
• Return encrypted outputs to the user.
This means validators don’t learn anything about the transaction or user data, yet the computation is still correct and verifiable.
Why Is This Better Than Existing Web3 Privacy Tech?
You might ask: why not just use zero-knowledge proofs (ZKPs)?
ZKPs prove that a statement is true without revealing the data, but:
• They’re one-shot: you can’t do dynamic or repeated computation easily.
• They don’t protect the state or outputs just prove the correctness of one statement.
• They’re limited in terms of logic complexity (and expensive to compute).
FHE, by contrast:
Lets you perform arbitrary computation on encrypted data.
• Keeps inputs, state, and outputs encrypted.
• Is more flexible for multi-step applications, like private machine learning on-chain.
• ZK and FHE are complementary. ZK proves something was done correctly. FHE actually does the thing on encrypted data.
@zama_fhe is building privacy.
English
