The Commoner

Tuesday, September 22, 2026

Sup, freaks.

A few thoughts on why I think people are dramatically undervaluing bitcoin, followed by Muse running into Amazon, the people who want to police AI, and a security update Lightning operators should read.

WHAT TO EXPECT IN THIS NEWSLETTER

• Marty's Bent: Why bitcoin's stupid simple rule set becomes more valuable as the world gets harder to predict.
• Signals: Meta's Muse and Amazon's resistance, who gets to police AI, Strategy's capital allocation, the Fed's supply-shock problem, and a Core Lightning security release.
• Freedom Tech Corner: Prepare a payment in BTCPay while keeping the signing keys on your hardware wallet.
• The latest bitcoin market and network readings in the Data Snapshot.

Marty's Bent

Keep It Simple, Stupid

I think people are dramatically undervaluing bitcoin right now, freaks. Particularly when you consider what is happening outside of bitcoin. We are watching AI advance at a pace that makes it hard to keep up with what these tools can do from one week to the next, a geopolitical order that is becoming more fractured, and governments piling more debt on top of promises they were already struggling to keep. Apparently this is the environment in which we are supposed to get bored with a monetary system because its rules are too simple.

As I wrote yesterday, I think those simple rules are going to be coveted as the world gets more chaotic. There is an incredible amount of creative destruction ahead of us. I'm taking my first spin with Muse this morning and am excited about getting tools like this into more people's hands. I don't think people have come close to wrapping their heads around what they mean for the businesses and institutions we interact with every day. A company can spend years building a product, hiring people and convincing customers to pay for it, only to discover that a growing amount of the work it charges for can be done another way. Being excited about that possibility doesn't make it particularly easy to determine which companies will be worth owning on the other side of it.

This would be a lot to work through on its own. We're going to be working through it while countries are competing more aggressively for energy, industrial capacity and influence over trade. The move toward a more multipolar world adds another layer of uncertainty for anyone trying to plan beyond the next quarter. Relationships people took for granted become less reliable. Supply chains that looked efficient become a liability. Governments respond to all of this while carrying enormous debt burdens and making promises that require even more spending. I'm having a hard time looking at this combination and coming away with the conclusion that I should put more faith in their ability to manage my money.

Then you look at bitcoin. The exchange rate moves around, but the issuance rules are public, there is a limit of 21 million, and anyone who wants to can run the software to verify that the rules are being followed. You can hold your own keys. You can send bitcoin without getting a central account administrator to approve the transfer. If a government gets itself into trouble, it cannot call an emergency meeting and make the node you are running accept extra coins. You don't have to predict which company wins the AI race or which government makes the right strategic decisions to understand why this is useful.

The fact that people look at these properties and complain that bitcoin isn't doing enough is pretty incredible. Jason Calacanis described bitcoin last week as a "boring store of value, trusted as that" before comparing it to CDs in the age of Spotify and DVDs in the age of Netflix. Being trusted as a store of value is apparently an underwhelming accomplishment. I would argue that getting people to trust that their money will operate on terms they can understand and verify is a rather important accomplishment. Especially when the alternatives are managed by people who keep changing the terms.

There is plenty of sophisticated engineering that makes bitcoin work, and plenty of work left to do on the tools people use to interact with it. What I value is the ability to understand the monetary system underneath those tools. The flashiness of another protocol doesn't do much for me if it comes with more things that can be changed, more people who need to be trusted, or a monetary policy that needs to be managed. I have enough moving parts to keep track of. I would like fewer of them involved in my savings.

Don't get bamboozled by the chaos or the flashiness of the latest technology. Keep it simple, stupid.

SIGNAL

Amazon would prefer you do the shopping yourself

I've been taking my first spin with Meta's Muse this morning. It's the first AI tool I've recommended that people simply download and start using. Giving people an agent with its own cloud computer and browser is a pretty big step beyond asking a chatbot questions. Amazon says it has blocked Muse, citing identification and credential concerns. As BuccoCapital points out, having an agent do your shopping creates an uncomfortable problem for the company that wants to control what you see while shopping. If an agent can compare products and make the purchase for you, Amazon has less control over what you see along the way. I'm interested in how that fight develops as more people start using these tools. Muse also puts a lot of trust in Meta: it operates the cloud computer, training is opt-out, and its Confidential VM is still forthcoming.

Who gets to set the rules for frontier AI?

DataRepublican's latest essay takes aim at Effective Altruism's influence over AI governance and promises a network map. You don't have to wait for that map to find proposals worth paying attention to. Dario Amodei is calling for embedded third-party evaluators, naming METR as an example, alongside government-supported coordination over frontier development. METR publishes funding disclosures and conflict policies, including restrictions on frontier-company funding and financial interests. Those policies deserve scrutiny if evaluators are going to help set limits on frontier AI development. Who pays them, how are their standards chosen, and how would a developer challenge a bad assessment? I don't want government-backed coordination to give the largest labs and their preferred evaluators a way to keep competitors out.

Strategy is buying both sides of its structure

Strategy bought 950 bitcoin for $75.7 million during September 14–20, bringing its reported holdings to 846,000 bitcoin. It also spent $174 million buying back STRC preferred shares. Both purchases came out of its USD Cash balance; it reported no shares sold through the at-the-market program that week. There's more going on here than adding bitcoin to the pile. Strategy is also buying back part of the structure it uses to finance that pile. Separately, $57.4 million from its USD Reserve went toward preferred dividends and debt interest. Worth keeping track of if you own the securities. There are obligations attached to them that don't come with holding bitcoin in your own wallet, and management's decisions about those obligations are part of the bet you're making.

The Fed can't produce the missing supply

Chicago Fed president Austan Goolsbee gave a speech in London yesterday arguing that persistent supply shocks can't simply be looked through. His explanation is worth reading alongside yesterday's oil discussion with John. Monetary policy works on demand. If the missing supply doesn't return, bringing demand into line with it means accepting less output and employment. These are Goolsbee's views, not a new FOMC decision, but they lay out the problem pretty clearly. People are already paying more for the things they need, and the central bankers' tool for dealing with that is to make it harder for people to spend. Getting more oil and refined products to market seems like a much better way to improve people's lives.

Core Lightning operators have an update to make

If you're running Core Lightning, take a look at v26.06.8. The maintainers strongly recommend upgrading as soon as practical. Alongside ordinary bug fixes, this stable release addresses responsibly reported vulnerabilities. One changelog entry describes a channel-state problem where a unilateral close after a splice could broadcast a revoked commitment and expose channel funds to a penalty. There are also authorization and crash fixes. Some tests are being withheld temporarily to give operators time to upgrade before more detail becomes available. Read the operator notes before touching your node, particularly if you've run development builds: their newer database schema prevents a simple downgrade to the 26.06 series. This is worth making time for rather than leaving in a tab you intend to get back to.

⚡ FREEDOM TECH CORNER

DATA SNAPSHOT

As of September 22, 2026, 12:17 p.m. ET. MVRV and realized price are daily observations dated September 21 UTC.

Sources: CoinGecko for price, rolling 24-hour change and market cap; mempool.space for network estimates; Alternative.me for daily sentiment; Bitcoin Lab for MVRV and realized price. MVRV compares market capitalization with realized capitalization. Realized price values coins at the prices when they last moved.

If you know someone who would enjoy The Commoner, forward this to them.

Stay frothy,

Marty

Marty Bent · TFTC · Nostr

YouTube

Podcast