GM, frens! ☕️ 🔥 Ethereum’s next audit partner: AI

Fun fact: the universe runs on loops. The sun rises and sets because the Earth spins. Seasons repeat because we orbit. Even our own bodies follow 24-hour cycles whether we like it or not. Expansion, contraction, rotation, return. Nature rarely moves in straight lines.

Our industry likes to pretend it’s chaotic and unprecedented, but it mostly rhymes. Attention expands, capital contracts, narratives rotate and the same debates come back dressed slightly differently. The shape changes, the cyclic nature itself doesn’t 🌎️ 

Here’s what we’re looking at this week:

• 🤖 Ethereum’s next audit partner: AI

• 👀 Ambire’s S2 check-in: TVL up, users up, activity up

• 💰️ Hyperliquid drops $30M to shill crypto in Washington

• 😱 Quantum fear is turning into protocol questions and nobody loves that

Below is how our $WALLET bag is doing right now:

If you want to talk cycles without losing your mind over them, our Discord is open 🤠 

Ethereum’s next audit partner: AI

A lot of people swear the average consumer AI chatbot got a bit dumber lately.

Maybe it did, at least in the way it talks, the way it hesitates, the way it tries to butter you up while saying nothing of substance. That complaint could be true, but it’s probably also the least interesting part of what’s happening 🤔 

Because on the code side, the floor keeps rising. Not in the “look ma, the robot wrote a Hello World” sense, but in the serious professional sense where these tools were being judged on whether they can survive contact with real repos ⚙️ 

GitHub’s Copilot study line that keeps getting repeated is that participants finished tasks faster, with the headline figure landing around 55%. You can debate how “real” the task was, or whether it generalizes, but the direction of travel is the point: code assistance has crossed the threshold from toy to habit.

Now bring that energy into crypto, where the consequences of a mistake are not a failed build, but a public lynching over lost billions.. 💸 

So OpenAI and Paradigm just put out a benchmark called EVMbench. The idea is to stop arguing about LLM generated spaghetti and test whether an agent can actually do the full security loop on Ethereum style contracts 👇️ 

That means finding a vulnerability, showing it understands how it would be exploited in a controlled setting, and then producing a fix that preserves the contract’s intended behavior instead of “fixing” it by changing what the app does.

That’s why this matters for Ethereum specifically.

• Ethereum has the deepest smart contract surface area, the most reused patterns, the most historical failures to learn from, and the most motivated attackers. If you want a benchmark that means anything, you put it where the incentives are sharpest.

• A standardized test setup like this also helps separate two very different things people lump together: tools that speed up competent engineers, and tools that let greedy CEOs fire top tier devs in a rush without understanding what they’re doing 🤦‍♂️ 

And yes if we’re frank, AI can draft, refactor, suggest tests, explain an unfamiliar codebase and catch obvious footguns but it doesn’t grant architecture. It doesn’t grant taste. It doesn’t grant the discipline to keep interfaces clean. The unit tests still exist. Production still exists. For a reason. And attackers definitely still exist 🥷 

So the real difference isn’t when“AI replaces your brain” it’s when “AI supplements your brain” - like a power tool - a drill doesn’t replace carpentry, it might make a crappy carpenter a slightly better one 🪚 

Used properly, AI shortens the distance between intent and implementation. Used poorly, it shortens the distance between “it compiles” and “why is the balance zero”, as we can sometimes see 🙄 

And also in the meantime for Ethereum, some big portfolio people are seriously treating ETH like a “default holding” instead of a spicy side thing 👀 

Harvard’s management team, for example, trimmed some Bitcoin ETF exposure and picked up an Ethereum ETF position. A big signal that ETH is on the menu for serious portfolios now 👇️ 

Lately ETH’s whole story gets easier to explain, now that it’s actually usable for every type of situation.

Fees being low changed a lot. People stopped delaying basic stuff because it feels stupid to pay a toll for every click, and devs can build things without constantly designing around “how do we make this not cost a fortune” 🛠️ 

A lot of the scaling work from the last couple of years is finally showing up as a real day to day difference, and that’s the part that makes 2026 look decent.

You still need good apps, you still need disciplined devs who don’t treat security like a suggestion, but the chain itself is in a better place now, and moves like that ETF rebalance are basically the grown up version of noticing it 👀 

Ambire’s S2 check in: TVL up, users up, activity up

Two months into Ambire Rewards Season 2, the numbers look more like steady usage stacking up over time: around $100M TVL, roughly $1M in swap volume, and 5,000+ accounts that have already secured $WALLET rewards.

When you see all of the stats moving together it reads like people are actually using the wallet, leaving funds there and doing real actions instead of just hovering over a campaign page 🙃 

There’s also a simple psychological thing in the best sense of the word: once users start treating a wallet as the thing they open every day, trust stops being a marketing line and becomes routine. That’s how you end up with Ambire securing around $110M worth of user funds while the rewards season keeps running in the background 👀 

And with 30 days left and at least $100,000 in $WALLET still on the table for this season, the interesting part isn’t the countdown, it’s what kind of growth this reflects when people move from trying the wallet to actually sticking with it 🫡 

Hyperliquid drops $30M to shill crypto in Washington

Hyperliquid is putting real money behind a policy push in Washington. The Hyper Foundation is allocating 1 million HYPE tokens, valued around $29 million, to set up what they’re calling the Hyperliquid Policy Center (HPC) in DC. It’s supposed to be an independent nonprofit focused on research and advocacy for DeFi policy, with a particular eye on perpetual derivatives and onchain market structure in the US 👀 

This is the part of crypto most people avoid talking about because it’s not fun, and because it doesn’t fit into a clean meme. Policy work is slow, full of meetings, full of compromise, and full of people who will confidently ask questions that make you realize they’ve never used anything they’re about to regulate 🥴 

Perps are a good example of where reality and regulation constantly fail to meet.

• The demand is clearly there, the volume is massive, and the market has been global by default for years, but the US framework still treats a lot of this like it’s an edge case 📜 

• If you’re a team building this stuff, you can either stay offshore forever and accept the tradeoffs, or you can try to help shape a path where “onchain markets” are understood as their own category with their own mechanics, instead of being forced into whatever legacy labels happen to exist 🧠 

Hyperliquid’s approach is basically: let’s try to be in the room early, with a structure that can speak policy language without flattening the tech into nonsense.

They’re bringing in Jake Chervinsky to lead the nonprofit as CEO, which is a signal that they’re taking the whole “translate crypto into what boomer regulators understand” job seriously.

There’s also the bigger meta point here. In crypto, we love pretending that dapps and shitcoins is the whole story, but the reality is that markets sit inside societies, and societies run on rules, enforcement, and incentives 🤓 

If the rules are written by people who don’t understand how decentralized systems work you get weird laws, the kind that creates accidental loopholes in one place and impossible requirements in another, i.e whatever the hell that clown show 🤡 was that we’ve had these past decades and then everyone spends years trying to route around it.

Quantum fear is turning into protocol questions and nobody loves that

Until recently “quantum” was still mostly a shortcut for getting a reply you can ignore, in the same bucket as “solar flare” “aliens suddenly attacking” type of FUD, we’ve all heard it and nodded politely, but then went back to refreshing our charts.

But the conversation is stopping being purely theoretical.

We’ve started seeing serious names attach themselves to the topic, and the suggestions got way more concrete. Not “someday we should think about it” but “here’s what we have to do” 🤔 

One of the biggest discussions this week was the idea of freezing older Bitcoin addresses as part of a future quantum upgrade.

• If some early address types are more exposed if quantum machines ever get good enough at pulling private keys from public information. If those coins can be swept by an attacker, it’s not just a few unlucky wallets getting drained. It’s a credibility event, plus a supply shock, plus a governance stress test all at once 🙅 

• Most people who actually work on cryptography will tell you there’s time, and they’re probably right. But markets and big players don’t wait for certainty. They price probabilities and they react to risk narratives long before the risk arrives.

That’s why it hits different when the people managing truly stupid amounts of capital start treating this as something to plan around. When you see big funds, ETF era players even entertaining the question, it stops being a niche discussion 🤷‍♂️ 

The proposed “freeze old coins” action is also a perfect example of why this topic has exploded 👇️ 

BTC is built to be permissionless. Freezing addresses, even in the name of protecting holders, sounds like a line you can’t uncross.

The minute you say “these outputs are unspendable now” you’ve introduced a social layer that looks a lot like governance by exception. Today it’s quantum. Tomorrow it could be something else. That slope is why people are extremely tense 🫠 

• At the same time, doing nothing has its own slope. If the threat model ever becomes credible and the chain hasn’t prepared a migration path, you end up with two bad options: rush a change that risks breaking things, or sit still and hope nobody builds the first working key extractor with criminal intent. Neither option fits Bitcoin’s culture of slow, conservative upgrades 🔨 

• So you get the compromise approach: build the upgrade path early, define what “safe funds” looks like in a post quantum world, and give people time to move.

• That’s where things like new output types and signature schemes come in. It’s not a promise that the threat is imminent. It’s infrastructure for a future where you’d rather not improvise 👨‍🔧 

Then there’s the other issue the community is outraged about: influence.

When large institutions hold a meaningful slice of supply via products like trusts and ETFs, they don’t need to “own” Bitcoin to exert pressure. They can shape the conversation, fund research, push timelines and generally make the projects feel different. If they decide the protocol has to move faster, that creates friction with the dev culture that is intentionally cautious and allergic to rushed consensus 👇️ 

And to be fair here: devs being conservative is a feature, not a bug. The base layer is not an app. It’s not a place where “move fast” wins. The fear is that corporate urgency doesn’t always respect that distinction, especially when everyone is starting to portray the risk as existential 🤷‍♂️ 

Other worthy reads

“100 Learnings from five years as a terminally online crypto degen” by Zeneca:

“WEB 4.0: The birth of superintelligent life” from Sigil:

“Former Binance Listing Manager Breaks It Down: The Three Core Variables That Drive Token Prices” from cryptodaoyi:

MEMES

That's all for now, frens.

We'll meet in a week! And remember, the market conditions are temporary, but our commitment to building a better Web3 is here to stay. Thanks for joining us, and we look forward to seeing you back next week. Cheers!

Yours, The 🔥 Team

Brought to you by Ambire: The Only Web3 Wallet That You’ll Need!