
Pinned Tweet
Personal agents are everywhere (Muse! Instinct! Hermes! GrokBot! Dots! Duet!).
And, increasingly, they're running into walls – similar walls to the ones banks threw up during the Open Banking fights of the past decade, as fintech users struggled to connect their apps to their bank-held financial data.
It's time for Open Agentic Commerce.
In Dodd-Frank Section 1033, the backbone of Open Banking, the definition of “consumer” reads: “an individual or an agent, trustee, or representative acting on behalf of an individual.” Congress obviously wasn’t contemplating AI agents when it wrote Dodd-Frank — but technology changing doesn’t obviate the underlying principle: consumers should be able to choose who, or what, transacts on their behalf.
Banks and companies like Amazon have real, legitimate concerns about agents becoming vectors for fraud or unauthorized transactions. But those are arguments for building portals for secure access, not for prohibiting agents altogether. Open Banking offers a useful lesson here, in that secure, standardized interfaces are preferable to forcing third parties into credential sharing or screen scraping.
We propose some basic guidance that relevant regulators (FTC, OCC, CFPB) could consider to help the market to coalesce around open norms before incumbents obtain sufficient lock-in, freezing the current landscape to their advantage.
With standards like these established, agents can make decisions that better serve the consumer, and will index on things like quality or price, rather than being funneled into a constrained set of options.
The push for Open Banking resulted in better experiences and lower prices for consumers. As agents become principal commercial actors, it is imperative that we lay the groundwork now so that an agent acting on your behalf is equally capable – if not more so – of delivering value and cost savings to you as you would be acting on your own.
Read our proposal below:
Glorious future
Now live on polimon.zak.co
Polimon: policy monsters
Battle your way through think tanks, subcommittees, and lobbying firms across DC as you try to stop the evil Team Moat and their regulatory capture
Have fun
One thing if CME is doing market research to support their litigation against CFTC on perps — but the fact this is from someone on their retail marketing team indicates they’re trying to launch a perps product of their own.
Which then begs the question, doesn’t that entirely undercut the argument CME is making in their lawsuit, and bolster that of the CFTC’s? (E.g. that CME just wasn’t ready for market so is suing to slow things down)
🤠⚡️🤖
Swarm riders and Swarm-native companies
AI-use even among highly AI-native people is power-law distributed.
There seems to be a new class of people — call them “swarm riders” — who are capable of consistently commanding large swarms of agents to productive ends.
This is a new kind of managerial skill, analogous to the skill it takes to manage large organizations of people, but wholly different in nature and of course more technical.
Based on the swarm riders we know (including many at @Tempo and @Paradigm), these people are capable of productive token spend that is 1000x-10000x higher than median software engineers, and 100x higher than top 5% Claude Code users.
The art of swarm riding is different from software engineering and more akin to machine learning, where practitioners grapple with black box systems: defining reward functions and evals, transmuting non-verifiable problems into verifiable problems, running experiments, etc.
Swarm-native companies are a new class of company founded by and populated by swarm riders. These companies are organized around swarms in a way that conventional companies will struggle to replicate, as the cost of internal human coordination is sand in the gears of an effective swarm. Effective swarm-native companies will define the right reward functions and evals, remove functional and bureaucratic barriers, and then let the swarms cook.
We’ve gotten to see a few swarm-native companies up close — reshaping financial rails @Tempo, building a new gaming platform @Spawn, cyber swarms at Cathedral Systems — and it’s clear these teams wield a significant advantage over pre-AI teams.
Some questions I’ve been wondering about:
1) How fast will swarm riding diffuse through the economy? The implications for token spend could be significant.
2) Can swarms be effectively adopted by existing companies or will it generally track the formation of new swarm-native companies?
3) What becomes scarce as swarm capabilities themselves commoditize (through frontier models becoming better at swarming, swarm capabilities built into harnesses, etc)? Cultures and teams that can define reward functions well and otherwise get out of the way?
1
29
1,603
Power access is fragmented, infra permitting is sporadic—and yet compute demand is universal and unyielding.
The relentless team of @OverReactor1776, @adamblake, & @nickporadek built @mavriaX to attack this problem head-on.
It’s time to manufacture the intelligence factory 🇺🇸⚡️
The next phase of the AI race will be won or lost in the physical world. America’s frontier researchers are regularly working miracles, but they’re limited by atoms and electrons. That’s why we founded Mavria.
We're turning bespoke construction projects into repeatable infrastructure products that can be manufactured at scale, like cars, computers and even rockets are today.
Our first product is Terrain, a factory-built data center engineered for speed and simplicity of deployment. With standardized 5 MW blocks, it integrates power, cooling, controls and infrastructure needed for AI compute. Terrain 01 is in production in Houston today.
Instead of sourcing and integrating a series of disjointed modules, customers get an integrated, factory-tested product. It’s the difference between building your own PC vs. buying a MacBook.
We’ve raised $130M, including a new $100M round led by @paradigm at a $1B valuation. We’re also backed by @amppbc, Critical Frontier, @Liquid2V, @DreamVenturesVC, and others. We're putting that capital to work securing critical supply chains, scaling manufacturing, and expanding our team.
Congrats to @PluralEnergy on their partnership with the state of Alaska — Alaska has long shared its resource wealth with its residents, the new InvestNorth platform will make it possible for Alaskans to share in its infrastructure buildout too
Alaska and Plural just launched InvestNorth, a new platform for financing infrastructure in Alaska.
Bloomberg (@business) has the exclusive: bloomberg.com/news/articles/…
(Third-party reporting; statements and figures are Bloomberg’s own.)
Securities offered through Plural Brokerage LLC, member FINRA/SIPC.
Amazing work by the @NousResearch team
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.
Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line.
After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs.
The values in an open model are one fine-tune away.
Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
A really great overview of what the near-future might portend in Washington by @JBSDC
A Forecast of the Near Future:
As we close out this remarkable week, discussions are shifting to the impacts of the election on the lame duck, the new committees, and the topics of early bills and funding fights. I want to pan up out of the day to day business and offer some predictions on more macro aspects of politics.
- The new Democratic Majorities that take power in January will not be divided primarily by ideology, whether someone is on the left or in the center. They will instead be divided by seniority. The chairs and their staffs will want to use this moment to achieve long held goals, such as root and branch oversight of corporate America, assuming the freshman will go along. This large freshman class will, like the freshman class after Watergate, realize they have their own different views and will resist being told what to do. In the end, the Chairs probably end up generals without an army.
- There are few deals with the Trump Admin. The Trump WH is not cowed by the results and decides the midterms were a function of the MAGA base being betrayed. While blame is shifted early on (rightfully) to the Iran War as a cause, hardliners inside the WH successfully argue that the base was let down by the lack of more deportations. The WH is disinterested in doing deals with Democrats generally and there are multiple shutdowns. The public sees Washington again as broken.
- The 2028 campaign for President starts early, with the first candidate to declare on the Dem side announcing an exploratory committee before Christmas. The Dems have a massive field, including one statewide elected official who only is sworn into office in January 2027. But unlike 2019, there is a clear S tier of just 3-4 people. The primary is less toxic and contentious than in 2019, as many in the party now realize the policy bidding war was a mistake and need infighting. The absence of Bernie, Biden, and Warren or clear successors to any of them makes the race less about the past and more about the future. On the GOP side, VP Vance chooses not to run but former Secretary of State Rubio does. He does a surprising challenge from the right on both the war and inadequate commitment to hardline immigration policy. The primary is short for the Dems and long for the GOP.
- The next two years of politics are defined, after a short burst of focus on broad oversight, by AI, which continues to pick up speed. The need to have technical knowledge of the subject to do effective oversight and legislation on it favors younger members with less seniority in both parties, reducing the power of the Chairs and Rankers. Technical staff grow in value versus lawyers for the first time in a generation, and surprising “gangs” emerge in the Senate to tackle aspects of the topic for the first real time in a decade. 2028 is the AI election, marked by at least one spectacular hack of major infrastructure during the last 90 days.
- When folks look back on our politics in 2030, it is clear to all that the Trump era, the Long 2016, ended on November 3, 2026, which also marked the end of the Dems long 2016 primary. 2029 also is the first year of an administration in which Dems do not feel compelled to fight over the choices of 2008-2009, whether they win or lose the election, marking the end also of the Long 2009. Trump is gone to Mar a Lago and seen like all former presidents as an artifact of a prior time.
- The Administration that takes office in 2029 experiences a ChatGPT like moment in robotics in its first 9 months. It is a surprise to all in DC.
1) this group's website went live in the last 24 hours, formed out of nowhere (like the group) for the express purpose of opposing equity perps
2) the lawyer on this complaint is Megan Barbero, Gary Gensler's SEC general counsel
I love tradfi regulatory protectionism!
The Capital Markets Alliance (CMA), whose members include Cboe, Citadel Securities and Healthy Markets, has filed a lawsuit challenging the CFTC’s approval of perpetual futures on individual U.S. stocks.
The group argues the products fall under the SEC’s exclusive jurisdiction and is asking a federal court to overturn the approvals.
equity markets are more volatile than ever. making it more efficient for institutions to hedge is paramount. credit to @ChairmanSelig and the @CFTC for their work
Today, equity perps can begin trading for the first time on @CFTC registered exchanges.
This move reflects my commitment to ensuring that America’s markets remain the most innovative, liquid, and well-regulated in the world. Instead of being forced to turn to offshore, unregulated venues to access equity perpetual futures, Americans may now access equity perps on CFTC-regulated markets subject to robust safeguards.
I appreciate the continued partnership of @SECGov and @SECPaulSAtkins as we work to further harmonize our agencies' regulatory initiatives.
2
18
1,337
“Huh, feels like there hasn’t been a new blockchain development advance in a while”
* @gakonst insta-appears with a new devmog*
stablecoins are for everyone (even for those whose lobbying group decides to sue their regulator)
The OCC wants community banks to help shape the future of payments and is encouraging the development of payment stablecoin capabilities within the community banking sector. Interested community banks are encouraged to consider applying to become, or for a subsidiary to become, a permitted payment stablecoin issuer, as applicable. These opportunities complement the OCC’s Community Bank Comeback initiatives to right-size regulation and tailor supervision so community banks can grow, compete, and continue serving their communities. OCC.gov/GENIUS
16
31
1,167
Was thinking this AM about what a combined entity could mean
One current vision for Tesla is for customers to opt-in to using their cars as RoboCabs when not in use. Disrupt Uber and then some
Taking that a step further though, imagine:
- you set your GrokBot an IRL task, e.g. “clean my gutters”
- the GrokBot then “rents” use of an Optimus humanoid, whether from a home (where it’s been made avail, like a Tesla car), or from maybe a local Optimus hub
- the GrokBot then directs (or rides on board!) the Optimus to come to your house and clean said gutters
- Optimus returns to its home upon completion, and GrokBot compensates the Optimus owner (whether consumer or corporate) with an agentic payment
Would basically turn an agentic harness into a new age TaskRabbit
Elon Musk today on a potential merger between SpaceX and Tesla:
"Both Tesla and SpaceX have formed special committees at the request of shareholders. Because I'm a conflicted party I can't be in on these discussions. I don't think we should assume that a merger is a certainty. Do I think it would be a good idea? Yes. But at the end of the day, shareholders have to be supportive of this. The special committees have not reached an agreement on any transaction, and it doesn't mean they will, but there is a lot of be gained from cooperating on SI, chip design, robots, mass or orbit, connectivity. I think it would be a good idea."
The GENIUS Act is arguably the first financial regulatory law written to foster innovation, as much as it was to develop regulatory safeguards.
We @paradigm appreciate @USComptroller and the @USOCC’s hard work to deliver these regulations without delay, while honoring that very ethos behind GENIUS.
It is a historic day for the American financial system, and financial innovation.
At the @USOCC, we have worked diligently and expeditiously to deliver the GENIUS Act final rule ahead of schedule to support President Trump's vision and @SecScottBessent's leadership to grow the economy and lead the global digital currency revolution.
Today marks the birth of a new industry having a federal framework for payment stablecoins in the digital age. It also brings us back to our original mission when the OCC was charged with the responsibility for organizing and administering a system of nationally chartered banks and a uniform national currency.
The global demand for compute is insatiable, but (particularly as the last few weeks in the market have shown) compute infra buildout is not linear, nor is demand.
Enabling labs, neoclouds, and investors to hedge their compute exposure will deliver greater certainty to the market, and unlock the next leg of global growth.
The @CFTC has asked how best to approach this burgeoning compute derivatives market — we teamed up with @andromeda_ai to answer.
Our take:
- these markets are nascent, so defining a market structure now is premature. But establishing basic benchmarks can help these markets to develop
- physical delivery can help anchor price
- fixed-date and perpetual derivatives serve complimentary roles, and are both well-suited to this underlying commodity
- compute fits easily within the existing CFTC framework, and does not need a bespoke regime
We appreciate @ChairmanSelig seeking comment from the market on these important questions, and look forward to engaging with the Commission as these markets develop.
Today, @paradigm and @andromeda_ai responded to @CFTC’s request for comment on compute derivatives.
The market for compute is big, and only getting bigger. But unpredictable prices mean higher costs, slower builds, and a diminished AI stack.
Compute derivatives fix that.






























