@mac_eth asked how to fix the $SURPLUS disconnect. So I went looking at what's actually worked On June 7 I wrote that $SURPLUS was the strangest chart on Base: product usage at all-time highs, token at the bottom, and nothing connecting the two. This week Mac posted something better than another metrics update he asked how to close that gap. Two ideas on his list: buying tokens based on how the business does, and converting the token into actual shares of a company. Same disclosure as last time: I hold the token, I'm down on it, and I'm not a developer or a lawyer. I read what's public and check what I can onchain. That's it. So. Do his two ideas hold up? Idea one: the project buys its own token This is the playbook everyone runs now. Hyperliquid spent around $644M last year buying its own token with trading fees close to half of all such purchases in crypto. Optimism announced its own program in January. Pump fun burned $370M of PUMP in one go 36% of circulating supply then cut its purchases from 100% of net profit down to 50%. The predictable demand that had been holding the price floor got halved overnight. Two problems for Surplus. First there's nothing to buy with. Hyperliquid funds purchases from fees. Surplus charges nothing. $0 revenue is Mac's whole position, he repeats it in every single update. Buying tokens "based on business results" with no revenue means buying them with runway money. The same runway he sold 2% of supply to secure. That's the project eating itself. Second even funded buying isn't the cheat code it was a year ago. By early 2026 the market had seen this playbook so many times that announcements barely move prices anymore. It amplifies revenue that exists. It can't replace revenue that doesn't. What survives from this idea: write the rule now, run it later. A public commitment "the day this product charges anything, X% of it automatically buys the token" costs zero today, breaks no promises, and turns "$0 revenue" from a dead end into a countdown. Idea two: turn the token into shares This one got a lot more real recently, so worth being precise. In January US regulators said it plainly: putting a stock on a blockchain doesn't change what it is. A token that gives you a piece of a company is a share, with all the paperwork shares come with. Then in March Nasdaq got approval to trade tokenized versions of listed stocks, and firms like Securitize started offering tokenized shares for US companies. The rails are getting built for real this time. Here's what the rails don't fix: you can only convert a token into shares of a company that exists. SURPLUS started as a community launch through Bankr the community picked the name, launch fees went to Mac. I couldn't find any public trace of a company with an actual cap table behind this token. So the path would be form the company, structure the equity, do a compliant offering, then run the swap. For a solo dev with no revenue, that's months of lawyer invoices before a single new line of product ships. Not impossible. Just the worst possible first move. What's not on his list Both of Mac's ideas try to push value at the token from the outside. The thing the token is actually missing is a job inside the product. Right now you can use Surplus every day and never touch SURPLUS. That's the disconnect. All of it, in one sentence. Three jobs the token could do: Hold it, pay less. Keep some SURPLUS, your inference gets cheaper or your requests jump the queue. Demand starts scaling with the exact number that keeps hitting highs transactions. Sellers post it as a deposit. Anyone listing surplus credits locks some SURPLUS first; deliver bad service, lose the deposit. This one quietly solves a second problem too why buyers should trust anonymous sellers at all. For a resale marketplace that's the existential question anyway. Payments destroy it. Pay for inference, the tokens used get destroyed; a fixed amount gets created for the compute sellers on the other side. More usage means more destroyed than created. The supply itself starts tracking the product. And here's the part where I correct my own post. Last time I flagged the contract's built in ability to create 2% new tokens a year as a risk a switch sitting in the owner's hands. Unexplained, it still is one. But it's also exactly half of that third mechanism, already written and deployed. The creation side exists. Add the destruction side and the "risk" becomes the engine. I missed that the first time. Reading a contract as a threat list makes you skip what the same code could be used for. What I'd want to see, in order of cost The written rule about future revenue free, could happen today. Then hold for discount, the smallest build. Then seller deposits. The destroy and create design last, when volumes justify touching supply mechanics at all. Shares go in a folder labeled "after the regulation settles," not on a roadmap. One thing none of this changes: when I last counted, SURPLUS had around a thousand holders and thin liquidity. Any of these mechanisms sets the direction of the link between product and token. None of them sets its strength. The chart stays violent either way for a while. But "violent and connected to the product" is a different bet than "violent and connected to nothing." The second one is what holders own today and Mac just said out loud that he knows it. We'll see which way he goes. I'm still holding mine.