The One Question bounty closes Thursday 10 December.

Join in
We Them Media logo We Them Media Support us

The State of Onchain UX. Why Crypto Still Feels Broken (How We Fix It) - Nnamdi Ernest

The picture loads from YouTube. The player loads only when you press Play. Turn on captions in the player if you need them.

Nnamdi Ernest says crypto has strong infrastructure but a poor user experience. He goes through unreadable wallet signatures, gas fees, phishing through hidden approvals, missing identity tools and a mismatch with how users think, with stories of a friend scammed in an airdrop and a lost transfer. He then covers fixes such as account abstraction, sponsored gas, transaction previews and starting newcomers on custodial accounts.

Not by We Them Media Channel: Web3Bridge. Published Sunday 26 October 2025.

Titles are the owner’s. It plays from their channel.

Transcript

This transcript may contain errors. This video belongs to Web3Bridge. Letters mark each change of speaker.

These are the speakers’ own words. We have not checked their claims.

0:02 A: Hello everyone, uh, good afternoon. I'm not going to be wasting too much of your time. I'll try and keep this, um, extremely short actually. So, um, my name is Ernest, Ernest Nnamdi. Sorry, my name is not on the slides. Um, at the moment I'm a blockchain engineer and a developer relations engineer for Morph. And previously I used to do DevRel for Celo and previously blockchain at ThorSwap. So today I'll be talking about the state of on-chain UX and basically how user experience sucks in crypto. Good. So I'm starting off with just a little vibe check. this statement saying we've built the financial infrastructure of the future, and yet it feels like it's, uh, internet from 1995. So we like to say the blockchain is the future of finance, and to that we've built like, uh, a really amazing technology. And we focused so much on like, um, say the back end of it, and then we did the front end zero justice. As a result, the user experience is actually quite terrible. For instance, you can teach your grandmother how to order food on Chowdeck, you can teach her how like, uh, book like an Uber ride, but you cannot really just in 5 minutes teach her how to swap, um, tokens on Uniswap. So, um, just to give you like a brief history of how far we've come. So we had Bitcoin, which gave us money without banks. We had Ethereum, which gave us programmable money. We had, uh, DeFi and NFTs, which gave us like a real use case for smart contracts, basically the Ethereum technology. And then we had rollups which came about to scale Ethereum. But now you see we've hit a roadblock, which is- hello, thank you. So basically we've hit the roadblock, which is the UX wall. So we are now at the point where we either fix the UX problem or we cannot progress any further. So I'll be talking about what's actually broken about user experience. So the first is wallet signing flows. At the moment, if you're doing- carrying out a transaction, say on Uniswap, or you're trying to say do a transaction on an RWA like Centrifuge, most often than not, you find out that you can't read the signature you see in your wallets, right? It's frankly unreadable. So you don't actually know what you're signing. And it's actually guesswork. And I'll be talking about this later on in this session. The second actually is gas. So imagine having to explain to someone new that you want to send USDT from one person to the other, but then you also need Ethereum to pay gas fee. It's like telling someone like, hey, you want to send money from Access Bank to GT Bank, but you need to have shares in Access Bank first to be able to do that. That actually doesn't make any sense. Now the third is phishing and scams. And this actually goes hand in hand with the first one, which is wallets, right? So I'm going to give like a story about my friend who used to like farm airdrops back in like 2022. So they were in this Discord channel and apparently they thought they were farming the airdrop. But what happened was the owners of those projects were actually farming them. So they told them that for them to receive the airdrop, each of them needs to have a minimum of $500 in their wallets. for you to be whitelisted for that airdrop, and they were going to check. So each of them started running around trying to keep a $500 balance in their wallets. And obviously, if you're a developer, you know it's quite easy for them to run a balanceOf function and check how many people have that specific amount. So they released a list of like, they called it a pre-whitelist, and she was celebrating that her name, or her address rather, was in that list. The next, they told all of them to basically sign up for the waitlist. But what they didn't know is behind that button was actually a function that, um, for those that are developers, you know that, um, there are 2 ways you can move money out of your wallet, right? There's the transfer function, which basically is signed by your private key, and there's the transferFrom function, which allows somebody else to move money from your wallet. But in order for you to do that, you need to grant approval. And for every approval, you need to say, who am I approving? The wallet address. And then the second argument is, uh, the amount you approve that person to spend. Now there's this thing malicious people do, which is they set the approve amounts to uint256 max, which is the maximum number, uh, Ethereum basically has. Basically speaking, is everything in your wallet. And what she didn't know is that, um, she was thinking if she was going to sign an approver function, it's going to write approver in your wallet. But that's not always true, right? That only happens if the developer decides to seek approval by calling the approve method directly. But if you use ERC-2612, you actually won't see approve, you will just see the function signature. So all she saw was just gibberish, 0x71222BB. She didn't care, she just clicked sign. And as she signed it, that's how all of them lost their initial balance. So that was like a welcome to Web3 for, um, everyone in that Discord group. So the next, um, problem is actually identity. Um, in Web2, we take, um, identity for granted because, you know, we go on any app and no matter the application we open up, we know there's always going to be that button, sign in with Google or sign in with Apple. But yet, despite how long we've spent in Web3, we've not had anything quite on that level. We've had some progress, we've had, um, ENS, we have like identity protocols, but nothing on that level at the moment. And then you can sum all of this to just one big problem, which is the mental model problem. So you see, the fundamental, fundamental issue is, um, mental models. See, users have, um, a very simple mental model. I want to be able to do X. They don't want to think about the complexities involved in all of that. Um, you see, another personal example, uh, last weekend I tried, um, using a product. Basically, you could send crypto and then get naira to your bank account. So I sent crypto and they told me transaction successful, but I didn't see anything in my bank account. So I reached out to them and they were like, uh, NIBSS or whatever was having issues. So I should just wait, they can't credit me. Now, even as someone who's been in crypto since 2016, that wasn't funny to me, right? So my own was, if NIBSS had issue, why even accept my money in the first place, you know? So, but these are things that, uh, you have to look at it from the way, uh, from a user's perspective. I don't actually care what happens in the interim between when I sent my money and when that money gets to my bank account. All I care is that I gave you crypto I should get naira in my bank account. And we need to have the same approach also in Web3. Because what we have at the moment is when a user says, I want to send money, the blockchain goes ahead to show the user every single state in the blockchain. For instance, the user shouldn't need to care about what nonce is. The user shouldn't need to bother about things like transaction fee, gas fees, and all of that. But I will talk about this more soon enough. So how do we fix it? We fix it by keeping the superpowers of the blockchain, but also removing the unnecessary details, right? So users shouldn't care about gas. Users shouldn't care about seed phrases or chains. They just want apps that solve real problems. Like, let's take, for example, a prediction market app, right? What is important is that you should be able to go to a prediction market dApp, look for like a prediction. Like let's pick something that's very likely to happen. For instance, Arsenal will win the Premier League. Anyone with common sense will vote yes, right? So you could go in and put, say, $1,000, $5,000, and you could say, yes, Arsenal is going to win the Premier League. What the user shouldn't have to worry about is oracles, oracle manipulation, and how that data is being fetched. He has no business with the price feed. He shouldn't know which Oracle provider that platform is using or the smart contract. All he knows is he should be able to click yes or no to a specific prediction. So, and we have at the moment 2 kits for all of this. So we have account abstraction, obviously, which brought about smart wallets. We have gasless UX, which brought about like sponsored gas. We have cross-chain invisibility, which is what LayerZero and Axelar currently does. Uh, we have human-readable identity, which is some of the work, um, ENS is doing. And then we are beginning to see projects like Rabby build, um, better safety warnings into their applications. Now, another thing I think we could do is progressive decentralization. So here's a not-so-controversial take. Um, in Web3, we like to say not your keys, not your wallet. which is basically like, um, something we use to drive people to use, um, custodial wallets, right? Or non-custodial wallets, sorry. Um, so we want people to be in charge of their own private keys. We want people to be in charge of their own, um, security. But this could be a problem for someone who is new because even people who are experienced face issues with this. Another personal story, uh, In 2022 December, I wrote the whole year in review thing, and the last 2 paragraphs was dedicated to how I was crying because I woke up, I read a tweet from my friend, uh, Olanet, and he was basically talking about nobody should keep money in exchanges, move your money out of Binance, blah blah blah blah blah. And I had just woken up, so I said, fine, here's what I'm going to do. Let me do like a test transaction. Let me move some money out of Binance to my wallet. So then I moved some Ethereum, around, uh, $6,000, and I went to the bathroom. So as I was in the bathroom, I kept waiting for the notification that says, uh, my money has arrived in my wallet, but I didn't see it. So eventually I paused whatever I was doing and I started refreshing MetaMask, and what I saw was transaction history. So when I clicked on the transaction history, I saw that the money came in and left in the same second. So going back, I started obviously after like maybe 3 hours of panicking, um, I started looking at what could have been the cause. Now for the developers here, most often than not you have a MetaMask extension in your laptop and then you use it carelessly because obviously it's testnet. But what had happened is I imported one of those wallets to the MetaMask on my phone to be able to test a project I was building. So after testing it on my phone, I had forgotten that the wallet on my phone is compromised. So when I sent that $6,000, I had basically donated Christmas gifts to someone else. So this is why I said, um, maybe users shouldn't immediately start with, um, their own wallets and private keys and all whatnot. Maybe we start with like a custodial approach. It's fine you leave your money in Binance for now at the beginning, right? Because obviously the fear is what if Binance goes down? FTX has gone down anyway. But if Binance goes down, if Binance goes down, we have a much bigger problem basically. But I think yes, at the beginning, users should be able to keep their money in exchanges, get familiar with the ecosystem, learn security tips. Then as they progress, we could have some sort of like a managed approach. where they have some sort of like, um, a split between non-custodial and custodial. And obviously, as they get more experience in the industry, they could go into like full sovereignty. So, um, this is what I think transaction states should obviously look like as we, um, progress into building more user-friendly applications. So at the moment, what we have is you sign a transaction, you wait and you hope. For instance, in my case, right, when you send a transaction and you wait and you hope that that money actually arrives at the destination, and then maybe you'd find success. Sometimes you won't find any success. But a better approach would be to preview transactions, confirm it, track the state, and complete it. So we should be able to see what we are signing, which is already possible today, but a lot of applications have not yet adopted this. rather than seeing a very weird hex or decimal signature, you should be able to see exactly what you're signing. You are giving this application approval to do XYZ. The next is error handling that doesn't suck. Again, how good is my slides? Error handling that doesn't suck. Sorry. This is a problem most developers would face. I don't think users will face it a lot. A lot of time when you're say building blockchain applications, let's say you have a contract that inherits another contract that inherits another contract, and maybe there are 5 functions being called between all of them, and then you run it. Let's say you run a test script and then you just see execution reverted. Now, at what point, a particular point, did the execution revert? Is it at the first contract? Is it at the second contract? Is it at the third? What specific function is failing? we can't tell. Most often than not, for those that use Foundry, they start doing -vv. That doesn't work, add another v. That doesn't work, you add another v. Then after 5 v's, you've run out of, uh, you run out of the limits for it. And then most often than not, you need to go to- it's not possible. You need to go to, uh, the Explorer and actually look through, um, the log trace. And looking through the log trace is something, say, someone who spent a lot of time building smart contract applications will be able to do because a lot of the log traces, they represent functions in terms of like their signature. And obviously these signatures are not in, say, clear words. For instance, the signature for transfer is in numbers actually. So moving on quite quickly is gas UX innovation. So I'm just going to breeze through it. We already have platforms like Biconomy that help platforms set up prepaid transaction bundles. We have gasless transactions where protocols could sponsor transactions for their users. We have gas abstraction where you'd be able to pay gas fee in whatever token. So rather than you're sending USDT and you need to pay gas fee in ETH, you'd be able to send USDT and pay gas fee in USDT also. Batch transactions, which obviously was brought about by account abstraction. You'll be able to bundle transactions together. And then moving quickly to emerging patterns, we could see what is currently working. So at the moment, we have one-click trading available by platforms like Jupiter. We have intent-based architectures from platforms like Uniswap, social authentication from like Web3Auth and Dynamic. We have embedded wallets. And then we have chain abstraction by LayerZero and Axelar. So the backend is busy. It shows that we are not only making progress on the frontend but also on the backend, right? So in 2021, a proposal for account abstraction went out by Vitalik, and that's what brought about smart wallets. In 2024, another proposal for Smart EOAs. Basically what 7702 does is it temporarily makes every EOA address- think of EOA address as your normal wallet- it converts it into a smart account, and then it could revert back to being an EOA. And then, uh, 8004 is, uh, US-focused, but you basically allow for people to integrate even AI into blockchain, uh, operations. So focusing on the metrics that actually matter: time to first success, error recovery rates, mobile completion rates, and return usage rates. Speeding through this. So finally will be the future and basically where we are headed. So where we are heading is where a world where crypto actually feels like Web2, where onboarding takes seconds, not hours. where mobile native experiences are actually everywhere and where blockchain complexity has actually been abstracted. I'm going to stop here today because they are giving me- but thank you all so much for your time.

More from this hub

Something wrong, or a video you want off this page? Write to wethem.xyz@gmail.com and it is fixed or taken down within 48 hours.