TACo section & page on threshold.network

[arjunhassard]

• [x] section on main page copy
• [x] section moved up to second section
• [x] section on main page visuals
• [x] Brief for Evandro and Sorin for TACo full page (threshold.network/taco) Dec 4th
• [x] full page copy
• [x] full page graphics
• [x] full page header section
• [x] full page process diagrams
• [x] full page overview & utility section
• [x] full page comparison table
• [x] full page use case + associated conditions section
• [ ] full page implemented
• [ ] full page pushed to production

[arjunhassard]

Header

[arjunhassard]

First Section – Value Propositions & Relevant Domains

[arjunhassard]

Second Section (Interactive) – First Selection: DKG

[arjunhassard]

Second Section (Interactive) – Second Selection: Encryption

[arjunhassard]

Second Section (Interactive) – Third Selection: Decryption

[arjunhassard]

Third Section (Interactive) – First Selection: NFT-Gating

[arjunhassard]

Third Section (Interactive) – Second Selection: Secret Recovery

[arjunhassard]

Third Section (Interactive) – Third Selection: Inference Protection This use case is intended to crowbar TACo into the LLM space, and has some legitimacy in that it reflects Nosana's stated reasons for integrating TACo. However, still articulating the conditionality and designing a mock UX element, so open to different ideas.

Summary: Prevent third-parties from harvesting or blocking communication between end-users and generative models.

Condition example (a bit boring!):

const nft_gpt4_1 = new condition.predefined.erc721.ERC721Balance({
    chain: 1,
    contractAddress: 0x008646372cc42E1a627FCe94Aa7a7033e7CF075A, // only GPT4.1 may consume outputs 
    returnValueTest: {
        comparator: '>=',
        value: 10,

UX element:

• A diagram showing a neural net returning an output, which is then sent to two other models. One can access it, the other cannot.
• Other ideas?

[arjunhassard]

Fourth Section – Competitor Comparison Table

[arjunhassard]

Whole Page

[cygnusv]

On the comparison table:

• EVM contract should be considered redundant & highly available
• PKI should probably be considered scalable
• Cloud service shouldn't be considered Transparent
• For TACo, is 2 chains multichain? 😝

[arjunhassard]

EVM contract should be considered redundant & highly available

100%, good spot

PKI should probably be considered scalable

If you're frequently encrypting lots of large files/packets for many recipients using PKI in the client (like a livestreamer), is it really scalable? I guess I could add another row: KEM/DEM or File-size agnostic

Cloud service shouldn't be considered Transparent

The subtlety here is that they don't pretend to be trust-minimized – unlike LIT – but maybe that's hard to infer

For TACo, is 2 chains multichain?

I can put that in the hover-over explanation modal!

[derekpierre]

Closed via https://github.com/threshold-network/website/pull/113.