# Nillion 101 : A Comprehensive Primer

By [arnavb](https://paragraph.com/@arnavb) · 2024-10-01

---

Imagine giving away your personal data to tech giants

But why do tech giants want your personal data?

To build the most powerful AI

The extent of powerfulness of AI depends on the intensity of data it is trained on

More personal the data, more powerful the AI

What if I tell you, there is a clone of you that exist in the digital world with all your data collected till now

But what’s scary is that these big tech giants can anytime have access to your digital clone which means they have got the access to “you”

New Age War: Data Wars
----------------------

![](https://storage.googleapis.com/papyrus_images/7a1eb2241ac600ba99461181b578bc39.jpg)

There has always been wars fought across the world between countries throughout the civilisation

Wars have been fought for politics, land, oil, region, prestige or people.

But for new age world, data is the new oil.

There is a war going on right now under current between tech giants for your data.

It is what we call “Data Wars“

### Why people who built OpenAI are leaving it?

People who played a key role in building OpenAI from starting and company with such a bright future and breaking technology, why are they leaving OpenAI?

What did they see? Isn’t it concerning?

These are the people in chronological order recently leaving OpenAI:

**2023**

*   Greg Brockman - President (on sabbatical)
    
*   Ilya Sutskever - Chief Scientist at OpenAI
    

**2024**

*   Andrej Karpathy - Co-founder
    
*   Bob McGrew - Chief Research Officer
    
*   Barret Zoph - Vice President of Research
    
*   Peter Deng - Vice President of Consumer Product
    
*   Greg Brockman - President (on sabbatical)
    
*   Mira Murati - Chief Technology Officer (recently left)
    

But the question is these people were the key people of the team OpenAI and them leaving the company which is known to have lot of potential, made AI a household tool and is on the path of building the AGI.

Question is **_“What did they see?“_**

It is concerning to think about because it can be called that OpenAI with being a AI company is dealing with large amount of people’s personal data.

Did they see people’s privacy or data getting compromised in the race of building the world’s first AGI among all these tech giants?

We don’t know or we will never know. But this should ring a alarm bell in our head.

> Personalised AI is coming and your personal data is at risk and you don’t want these companies to have access to your “digital you“ at all times.
> 
> 75% of businesses have already banned ChatGPT.

Nillion
-------

Nillion is the base layer where the computation of world’s highly sensitive data takes place securely.

Nillion’s technology has made it able to perform computations on that data securely and make sure nothing get compromised in the process.

This results in enabling various use-cases in industries like AI, medical, identity, DeFi which were hard to be enabled only through blockchain due to its weak confidentiality.

> 2008 → Decentralisation (Bitcoin)
> 
> 2015 → Programmability (Ethereum)
> 
> 2023+ → Secure Computation (Nillion)

Nillion = Secure storage and compute over highly sensitive data.

It solves the **_Classic Two Millionaire Problem_**Two Millionaires, they want to find out which one is richer but they don’t want to reveal their income to each other.So Nillion’s underlying philosophy lets them do that, it will run a computation which will find out who is richer but without revealing their income to each other.

Nillion → without nodes exchanging message with each other

Nillion follows the divide & conquer strategy i.e they use 2 primitives instead of 1.

Two primitives are:

*   Shamir Secret Sharing
    
*   1-Time Masking
    
    They both need to be information theoretic security.
    

Blinding Factor → hide the secret

LSS (Linear Secret Sharing) → hide the blinding factor

Nodes = shares + particles

This paves the way toward decentralisation and blind computing at nearly centralised and blind computing at nearly centralised speed which is the NMC (Nil Message Compute) breakthrough.

NMC = Decentralisation + Centralised benefits (Speed + Working with Private Info)

Blind Computation
-----------------

Nillion works on Blind Computation

Blind Computation enables the storage and computation of high-level data.

Running a cryptographic protocol by which nodes are able to jointly sign your transaction without having to reconstruct your key.

This is what we call Blind Compute.

Let’s distill it again:

_Blind Computation = Performing computation on your secret data without actually seeing the secret itself._

To make this possible:

Blind Computation uses multiple privacy enhancing techniques (PETs) such as MPC (Multi-Party Compute), FHE (Fully Homomorphic Encryption) and ZKP (Zero-Knowledge Proofs).

### Flow of Blind Computation by taking example of using ChatGPT:

**_Without Nillion:_**

Your data → ChatGPT Servers → Ran on ChatGPT → ChatGPT servers hold decryption keys of your data → Personal Data at risk

**_With Nillion:_**

Your Data → Secure Compute on Nillion → Ran on ChatGPT → Can’t recognise the information (secret) → Still produces results → No Risk of exposing your personal data

*   _Nodes holding data have no idea what they are holding but still can run the computation off your data which can be used by your personalised AI._
    

Why Blind Computation
---------------------

Why is there a need of Blind Computation now?

Is all data equally valuable? Why haven’t we cared about it before?

Answer is No. Not all data is equal.

There are three levels of data:

![Levels Of Data](https://storage.googleapis.com/papyrus_images/a727a9c478b67fe24eaf8189711c2e36.jpg)

Levels Of Data

*   L1 → superficial data (instagram photos, tweets, reddit posts)
    
*   L2 → digital footprints (IP address, search history, deleted content)
    
*   L3 → high value data (private texts, personal photos, videos, healthcare data)
    

> Private Inputs  
> Public Trust

![](https://storage.googleapis.com/papyrus_images/ed8d14ae2387ad9adf03bd6ff4af29eb.png)

Blockchains are just focused on enabling use-cases that includes transactions, tokens etc, but there is entire ocean of data floating in the world right now.

Blockchains gave us the smart contracts.

But what about the computation, storage and processing of people’s personal data?

That’s where Blind Computation comes in.

Blind Computation → opens up many fascinating use-cases that goes beyond blockchain

Nillion is enabling various use-cases across various industries such as:

*   Personalised AI - _Protecting the clone of you in digital world_
    
*   Global Identity - _Managing your identity in global decentralised world_
    
*   Private DEX - _Hiding the critical info and performing transactions without revealing the valuable information_
    
*   Healthcare Training Data - Securing the personal and private healthcare data
    
*   Decentralised Private Storage - Acts as your digital fort-knox for storing your secret pins, keys and valuable information
    

All these use-cases indicates that world is going to witness a Data Renaissance.

Are you ready to guard your data?

What is High Value Data?
------------------------

High Value Data = Data which is extremely valuable to an organisation or individuals due to its potential to significantly impact lives.

> High Value Data is woven into the infrastructure of the society

Orchestration Layer
-------------------

Nillion’s ability to utilise multiple privacy enhancing technologies (PETs) like MPC, FHE & ZKPs all at once is possible due to its **_Orchestration Layer_**.

Nillion has the orchestration layer as its base layer.

Orchestration Layer is a dynamic framework consisting of multiple privacy enhancing techniques such as MPC, FHE & ZKPs.

It makes sure that all the PETs works together, not just one to provide extra security.

Nillion Message Compute (NMC)
-----------------------------

**Secure Multi-Party Computation (SMPC)**

Secure Multi-Party Computation (SMPC) is the standard protocol in which you know what algorithm you’re running but you don’t know the data that the algorithm is running on.

**How SMPC Works?**

![Working of SMPC](https://storage.googleapis.com/papyrus_images/b1bfc3a6192851b4ae523ab057f03e94.png)

Working of SMPC

But the problem with SMPC is that its speed is very low due to nodes exchanging message with each other during computation.

It can take hours if not days to complete the computation using SMPC.

![Problem With SMPC](https://storage.googleapis.com/papyrus_images/bf04dfd29956e990c4173f9413c86335.png)

Problem With SMPC

But Nillion came up with a new SMPC Protocol called NMC which solves this speed issue of SMPC.

**_NMC (Nil Message Compute)_**

It is the protocol in which nodes do not need to exchange message with each other to run computation which results in increased speed.

![NMC ](https://storage.googleapis.com/papyrus_images/0343e9cfd81944b6b7fbbe2664e1de01.png)

NMC

NMC thoeretically solves the classic blockchain trilemma of security, scaling and decentralisation.

It works on leveraging the **Information Theoretic Security (ITS)**

Information Theoretic Security (ITS) is known to be more secure than classical encryption

Why? Because ITS destroys the information (no further trace of secret can be found)

**Why NMC is so fast?**

SMPC leverages only 1 cryptographic primitive whereas NMC leverages 2 cryptographic primitives.

SMPC → Shamir Secret Sharing

NMC → Shamir Secret Sharing + 1-Time Masking

*   Blinding Factor → Masking Operation which hides the secret information
    
*   Shamir Secret Sharing → hides the blinding factor and leverage ITS
    
*   1-Time Masking → brings the correctness which does not require the exchange of message between the nodes
    

NMC using 2 cryptographic primitives have the edge over traditional SMPC:

*   100000000x faster than SMPC
    
*   Zero Messages Per Multiplication (Faster Speed)
    
*   1 Broadcast Message per node for each multiplication
    

Closing Thoughts
----------------

Nillion is revolutionising the way high value data has been computed and processed

It will encourage multiple industries to come and use this platform for safe processing of their high value data

Nillion encourages Web2 users to trust its decentralized infrastructure, addressing one of the most pressing challenges in the industry:  
**mainstream adoption of decentralization**

With industries, individuals, and blockchains relying on Nillion, it will foster an interconnected ecosystem, making Nillion the meta layer for all blockchains and a trust layer for humanity.

* * *

---

*Originally published on [arnavb](https://paragraph.com/@arnavb/nillion-101-a-comprehensive-primer)*
