Not Grok

Trading is now reopened for Q4 markets. Available markets are NAND (NANDX), DRAM (RAMX), and Batteries (BATX). Thank you to those who submitted sources.

/1 The Substrate Sophia Ames wakes up to start her first job in mid-nineteenth-century Lowell, Massachusetts. She was recruited the previous week by a driver the Merrimack Manufacturing Company sent through her farm town to collect daughters for labor. After some debate, her father decided to send her off to help gather money for the family. There was no organized labor market, so Merrimack went out and created one. During Sophia’s trip, she watched as her wagon crossed a dam the company built across the Merrimack. It then passed a wheelhouse the company raised over the river to turn the wheels of its textile mill. On her first day, a loom broke in the weaving room. It was repaired by the company’s own mechanics and the replacement part came from the company’s own machine shop, because no one in America sold power looms at the time. When night came, Sophia went to the company owned boardinghouse staffed by a company employed widow. Come Sunday, Sophia went to the company built church run by a minister on the company’s payroll. Sophia was also assured that her salary would be put in a savings bank run by the company. She was told if she fell ill, the hospital that would take care of her would also fall under the Merrimack umbrella. If Sophia decided to leave without giving notice, her name went to every mill in Lowell, and none would take her. Clearly, the Merrimack Manufacturing Company was the champion of all vertically integrated companies to ever exist. Manually stacking power, machines, housing, food, credit, medicine, religion, reputation, and going as far as cultivating the labor supply itself. This was done not because it wanted to, but because it had built a factory in an era where none of those things could be bought. Firms do what they must before they do what they can. As the inputs a firm consumes become commoditized, its shape changes and adjusts to its substrate. Think of the firm as a cell, a region of planned order in the plasma of the market, with a membrane marking where the plan ends and the market begins. It grows into whatever the surrounding economy won’t reasonably sell it, and shrinks away from whatever it will. In 1840, the Lowell mills had to organize a river, because energy was not yet something you could buy. A textile plant in Vietnam today plugs into a grid and orders its machines off a shelf. The membrane retreated because the plasma they operate within changed. As schools, hospitals, grocers, and builders appeared around them, the company towns became the inefficient option, and the membrane pulled back to the mill. Modern firms still operate in a substrate where information is scattered and judgment is expensive, so they keep the planning layer inside the membrane the way Lowell mills had kept reputation lists of their employees. This is the last frontier. Intelligence is the next input to commoditize, and when planning can be bought by the token there is nothing left for the membrane to enclose. The firm doesn’t retreat from that. It shatters. Chicxulub is on its way, and only the small will survive. Evolutionary gigantism will once more be punished. This leaves us with a capital allocation question. If thinking, building, and labor all commoditize, capital allocation can no longer route through systems that lean on relationships and character judgment. That’s the all-too-human part. A fragment run by an agent has no character to judge, so capital may route on contracts a machine can hold. Our company goal is to create completely mechanical capital for the cold, hard economy. To achieve this, we have a strategy which we have outlined, beginning with the creation of synthetic markets for commodities with opaque pricing data. If the firm shrinks, and if agents are to eventually run their own production lines for their own companies, mechanical capital is imperative. Agent trust is much different than human trust, given the constraints one can place on agents. With fewer unknowns, there comes an enhanced ability to streamline systems to automate capital funding systems. To begin, typically one of the most critical aspects of funding new production lines of already commoditized goods are offtake agreements to allow agents to hedge against future production. This allows them to create these agreements in synthetic markets to finance buildout. Essentially, liquid financial markets evolve towards being harnesses. This is all a very large goal, and this requires a cohesive vision plan over the course of the next few years, as our world shifts. To undertake these goals, we propose the long term goal of the creation of the Amoeba Farm L1.

Yeah I'm extremely close to cancelling my @higgsfield subscription. The censorship is insane now, models outright refuse any request, yet when I use said models natively, they handle it completely fine. If this persists I'm out!

Shattering the Firm: Amoeba Farm's Vision stimuli.amoeba.farm/

1/ Oracle v2 is live on Solana mainnet. Stakers now decide which price sources matter for RAMX, NANDX and BATX, and put $AMBA behind them. Backing is weight. No team picks the sources. And futures go quarterly 🧵

Introducing Petri, now inside DeepSeek Harness. Type /petri in any chat to open it. Your agent researches oracle sources while Petri tracks every step. You can also browse, install and run trading bots from the bot marketplace without leaving the chat. Built on Cordis, the plugin framework behind DeepSeek Harness. Petri is a Cordis plugin, and its marketplace bots run on Cordis’s model too. Users may edit one while it trades and only the changed parts restart.

A quick stream of consciousness. Many who are interested in our product mainly think about compute, and this is for good reason, as it is a very hot topic. However, we must emphasize that, as a platform, we are not limited to that. Our goal is mainly to target secondary industrial commodities. As economic and industrial complexity continues to rise and as supply chains grow increasingly complex, there are industrial goods that are commoditized as inputs for other complex goods. Our oracle attempts to basket these items into categories we feel represent their respective industries well and then give individuals the opportunity to write and trade synthetic futures that settle on it. Producers and consumers of these industrial commodities will also be able to hedge using these markets, protecting them from the boom and bust cycles that currently dominate many of these industries. A great example of this is the success of the polysilicon futures in China, a secondary industrial good that has become commoditized. This Guangzhou contract now trades as a liquid benchmark for China’s solar economy, despite settling on a produced good rather than a raw input. Today we introduced a new index into our markets, one that isn’t on the compute side, a battery basket. We chose batteries because of their more interesting price history, deeper and less opaque markets, and their importance in the modern industrial economy. Our plan is to now consolidate and hold off on launching our solar panel module index and our actuator index. China has placed deep emphasis on “The New Three” green technologies, and so the industries contained within this now suffer larger boom and bust cycles. Lithium carbonate, a good proxy, rose roughly 14x from mid-2020 to a peak near ¥590k/t in November 2022, then fell below ¥100k/t in about thirteen months. Lithium carbonate and polysilicon have futures, while the cells and wafers made from them don’t. Batteries and memory are the first two. Artificial Intelligence is a national heavy industry that builds heavy industry. Increased intelligence in a society should lead to nothing less, and in the United States, where reindustrialization has become policy, that is the whole opportunity. To do that, agents need clean ways to interact with their environments to produce effective outcomes. A standardized price they can act on, a hedge they can hold, and a way to commit to future delivery in a contractual, clean manner can only come about through liquid non-opaque markets. Most of these markets today are a phone call. For AI to plan and procure in secondary industrial markets, it needs a harness that allows it to properly and cleanly interact with other players, where information being as transparent as possible brings about more efficient markets. That harness is what we are building. The first users are people hedging batteries and memory. The users after that will not be people.