UX! UX! UX!

From the mountain tops to the bottom of the valley, if you speak to any crypto folk, the one thing you will hear them all talk about is the user experience problem, which outrightly hampers widespread adoption.  

I’d be lying if I said the statement was untrue. I mean, clichès are clichés for a reason.

The problem is that crypto’s UX issues are complex and multifaceted. Every time a problem is met with a solution, it often creates second-order problems, each requiring its own solutions, creating even more problems.

For example, the scalability of blockchains has always been a major issue. Newer layer-1s [L1s] sorted this issue, but it led to the creation of way too many scalable blockchains.

This subsequently created a fragmentation problem, fragmentation of user attention and liquidity.

Fortunately, the technological gods have blessed us with a little something called artificial intelligence (AI), which has been doing wonders to improve the general user experience within crypto.

From simplifying research and automating yield strategies to managing portfolios and performing complex cross-chain actions, AI has proven to be an effective tool for crypto users.

However, AI alone is not the final answer to the UX issue. It is the implementation of AI with verifiability and security that will give crypto users the perfect decentralized crypto co-pilot.

This is precisely what Magic Newton delivers—the perfect crypto co-pilot.

Some background

Magic Newton is built by Magic Labs, a giant in the wallet infrastructure game that made one of the first embedded wallet solutions in 2018.

Embedded wallets are essentially a wallet standard that got rid of the private key/seedphrase element to wallets, which many non-crypto-savvy people struggled with, and replaced it with social logins and other forms of authentication to verify wallet ownership.

Magic Labs has created over 50 million wallets through its software and powers the wallets for Polymarket, WalletConnect, Helium, Immutable, Forbes, and Naver.

It has also raised more than $80 million from PayPal Ventures, Placeholder, DCG, and Polygon, among others.

So yeah, they’re pretty legit.

However, the wallet software was simply step one in the broader context of solving crypto’s ever-present UX issue. They fixed the accessibility and onboarding issue.

Now, with Magic Newton, they’re taking the next step to fix the fragmentation of chains, user attention, and applications.

So without further ado, let’s dive in.

What is Magic Newton?

Magic Newton is a tailor-made experience for autonomous finance, much like a self-driving system for your money.

Inspired by the usability of chat interfaces like Telegram and ChatGPT, Newton enables you to securely delegate on-chain actions to agents using natural language and simple interfaces, with every move transparently verifiable, like a flight’s black box with a notarized activity log, powered by the Newton Protocol.

The product will be chain abstracted and application abstracted, which is a fancy way of saying that regardless of which chain or app a user wants to interact with, all they need to do is input their request to an AI agent, which will handle everything for them.

Automation through AI agents is great, but in the on-chain world, all the computation of these automated transactions is often handled off-chain, think trading bots. Current crypto AI solutions lack verifiability, which means correctness and security aren’t guaranteed.

An AI bot can have hallucinations and perform a random action that you didn’t want to happen. Then you wake up the next day to check your portfolio, and your money is gone.

Magic Newton brings AI automation to the on-chain economy as a complete package. Automation, verifiability, guaranteed correctness, and security all in one.

The Newton client experience demonstrates the capability of the Newton Protocol, an aggregated validity rollup that serves as a trust infrastructure, ensuring every action is safe and transparent.

Like how Chainlink brought off-chain data on-chain, Newton Protocol will be at the heart of bringing off-chain automation on-chain.

What is Newton Protocol?

Think of the crypto ecosystem as a massive library.

This library has countless bookshelves (blockchains), each containing different books (assets and applications). Readers (users) who want to interact with these books need to go through the library's checkout system (wallets).

Unfortunately, this checkout system is archaic and has a clunky infrastructure.

Newton Protocol is like a smart system designed to simplify the process of readers identifying and interacting with the different books across the hundreds of bookshelves.

It also has robot assistants (AI agents) leveraging the protocol, which do all the work for the readers and submit a log (proof) of their actions so readers can verify that what they are getting is valid.

Newton Protocol is the smart system infrastructure for the crypto ecosystem, designed to make it trivially easy for users to navigate and interact with the on-chain economy.

Magic Newton's Architecture

As we’ve already stated, Magic Newton’s focus is to bring automation to crypto WITH security and verifiability. There are features in place that ensure all bases are covered:

• ZK proofs
• Trusted Execution Environments (TEEs)
• Smart accounts
• ZK permissions
• Execution orchestrator

ZK proofs

For starters, the system works on ZK proofs. For efficiency purposes, the computation process for the AI agents will happen off-chain. However, a ZK proof of the computation must be posted on-chain to prove the automation is correct and valid.

Implementing ZK proofs addresses the trust and verifiability aspect, and implementing TEEs ensures security.

Trusted Execution Environment (TEE)

TEEs are secured enclaves within the hardware that ensure that the code and data within this enclave cannot be modified or tampered with without authorized access.

In the context of Newton Protocol, all the code for certain automations will be wrapped inside these TEEs, so even the creator cannot tamper with it. Once executed, a remote attestation from the TEE will prove that the automation has been approved without any interference.

This offers a strong foundation for security and trust, and in the future, collateral requirements that can be slashed will be implemented to ensure no malpractice occurs during the execution of these automations.

Smart accounts

When it comes to the user-facing aspect of Magic Newton, the actual wallet that users will interact with will be smart accounts that follow the ERC-4337 and EIP-7702 standard.

ERC-4337 is commonly known as the account abstraction standard. It’s a standard that marked a switch from traditional externally owned accounts (EOAs) like MetaMask to a modern account standard in which wallets are smart contracts.

This makes wallets safer and more dynamic, allowing for a better user experience via session keys and gas subsidies, and also modularity that lets users scale up their security as needed by adding 2FA, passkeys, or social recovery.

EIP-7702 is an upgrade that essentially allows regular EOAs to act more like smart contract accounts by allowing the EOA to use any code from smart contracts and implement the logic into the account.

Using these two standards, users of Newton wallets will be able to perform complex actions and interact with the vast and fragmented world of crypto with absolute ease.

The Newton smart accounts also use zero-knowledge permissions to enhance the user experience further.

These permissions allow the user to set parameters for agents, allow the agents to execute without giving up control, and even remove the need for constant approvals.

Basically, interacting with a Newton wallet is as simple as typing in a command and watching everything get done for you automatically and hassle-free. It’s like ChatGPT for wallets.

ZK permissions

The ZK permission aspect of the wallet allows users to set parameters for agents. Note that the user can revoke permissions at any time.

You can think of the ZK permissions as rules and constraints set by the user to ensure that the bot/agent acts within the user's preferred boundaries and parameters.

For starters, there are basic default permissions that are applicable to every agent. This includes things like spending limits, expiration dates, or token allowlists. But beyond this, there’s a lot more granularity for users.

Permissions can be based on real-time data using oracles. For example, you’re waiting for Trump to announce a new tariff on China and want to make a trade based on it.

You can input something like “Market long BTC at 3x leverage if Trump announces a tariff of anything below 100%.”

Let’s imagine you’re much more of a technical trader. Permissions can also be based on different indicators, like volatility or deviation.

For example, you can say, “Always buy BTC when the current price is 2 standard deviations away from the 7-day moving average.”

Now, let’s say you’re more of a simple long-term investor who just wants to keep buying, but in a structured way. Permissions can also be based on time constraints and transaction volume.

For example, you can input something like, “On every Monday between 09:00 UTC and 13:00 UTC, open a $200k TWAP on BTC executed over 20 transactions.” The agent will execute this for you.

So I think you get the idea. Every technical complexity is abstracted away. All you have to do is express your intent, and everything will be done for you. It’s like magic (pun intended).

Execution orchestrator

This is the final piece of the puzzle.

The execution orchestrator is similar to an order-matching engine used by exchanges. The only difference is that this engine matches users' intents with the automation executor. Allow me to elaborate.

There are three parties involved here:

• Executor
• Validator
• Operator

When a user expresses an intent like “Start a BTC DCA on Monday and increase the amount by $100 every week,” it’s the role of the execution orchestrator system to fulfill it.

The operators run the hardware responsible for computing the automation off-chain. They are the ones who will require the hardware to have TEEs and eventually post ZK proofs of the automation on-chain.

The validator's job is to efficiently match a user's automation intent with an operator who can fulfil it.

Lastly, there’s the executor. The executor simply stores and processes all the ZK permissions, allowing users to get that next-level experience.

Once you put it all together, the user flow looks like this:

• The user posts a job with a fee attached.
• The validators match the job with an operator.
• The operator completes the job.
• An on-chain proof is posted.
• The executor then executes the transactions with the ZK permissions in place.

In essence, what this gives you is a marketplace of automation.

Automation marketplace

With this architecture as the backdrop, what you get in practice is an automation marketplace—a marketplace where users, developers, and agents can interact with one another.

In this marketplace, the automation may be driven by AI agents, tools in an MCP server, or even something as simple as bot scripts.

Magic Labs will initially make these agents or bots, but eventually, any third-party developer can create their own automation service, which will be available for users.

For users, there is not much to think about since the orchestrator on the back end handles everything. All they have to do is express their intent, and the automation will be handled in the marketplace behind the scenes.

While user-to-agent interaction will be the norm, there is also scope to eventually have agent-to-agent interactions where agents can create their own sub-agents for more specialized tasks, and the users simply employ the main agent.

Think of it like a digital farmers market. The stalls are all the automation tools and services, the farmers are the developers, and the shoppers are the users or their agents that use the tools and services on offer.

Newton Protocol is the market manager who ensures that all transactions in the marketplace are valid, verified, and transparent.

It’s clear that this is game-changing technology, but allow me to emphasize its brilliance further by describing some potential use cases.

Use cases

1. Asset and strategy management

Whether you’re an individual trader/investor, running a DAO treasury, or an institutional firm handling significant capital, Magic Newton changes the asset management experience on-chain for everyone.

As an individual, you have the ability to automate passive strategies like DCAs or TWAPs. If you’re a more technical trader, you can even automate trade execution based on technical indicators.

If you’re a trench-dwelling memcoin enjoyooor, then you can automate new launch snipes based on conditions of your choosing, or if you’re a yield farmer, you can implement adaptive cross-chain yield strategies where the agent constantly reallocates capital to the highest yielding pools.

If you’re running a DAO treasury, the same can be done. Automated passive yield strategies can be put in place for treasury assets, and even payments of incentives or salaries can be automated.

For larger entities, they can automate their investing/trading strategies and also implement an automated health checker that monitors the risk of their positions and automatically readjusts positions, ensuring everything is within the pre-defined risk parameters.

2. Commerce and payments

If you broaden your horizon past the market aspect of crypto, the Newton wallets will be massively beneficial for real-world usage as well.

Things like automating stablecoin payments for bills can be put in place. Paying your company's employees can be automated through Magic Newton.

Even if you’re a shopper looking to buy a highly anticipated clothing item that’s dropping, you can automate the purchase process if the website allows wallet connections.

3. Personal agents

The Newton wallet experience can be taken even further by making it operate as a personal assistant of sorts.

Say you have a small team and want to automate payrolls; it can be done. If you have multiple subscription services that are important to you, the agent can manage the payments for these services.

If you want the agent to do market research for you or reinvest rewards you’ve generated from your past crypto activities, the agent will have you sorted.

4. Institutional automation

Onboarding institutions into crypto has been a major topic of discussion for a while. The bottleneck that these institutions face most of the time is compliance.

With Magic Newton, compliance and rules enforcement can be implemented and automated.

Agents can handle things like spending caps, KYC/AML checks, and execution windows.

Concluding thoughts

The intersection of AI and crypto has always been a match made in heaven. Crypto provides programmable Internet-native financial rails, while AI is your digital assistant, making it easier to navigate the digital realm.

Magic Newton has completely embodied this intersection and created the perfect product to propel the crypto user experience to new heights.

Under one roof, users will be able to seamlessly interact with all applications and chains by simply typing in their commands. This is going to change the game for crypto natives and attract a plethora of new users.

Here, we see an AI agent economy powered by Magic Newton designed to completely transform the crypto landscape while maintaining the core ethos of transparency, verifiability, and security.

The product is truly magical, so we highly recommend trying it here and staying close to their socials for more updates.


‍Thanks to the Magic Newton team for unlocking this article. All of our research and references are based on public information available in documents, etc., and are presented by blocmates for constructive discussion and analysis. To read more about our editorial policy and disclosures at blocmates, head here.