Stripe and Paradigm have launched Tempo, a Layer 1 blockchain that fundamentally changes the rules of stablecoin transfers. This isn't just another blockchain project. It's a systematic attack on the incumbent payment infrastructure, executed by a team that knows how to build at scale, backed by the world's largest financial players.

Tempo solves critical problems that have hampered enterprise blockchain adoption: speed, transaction segregation, and predictable costs. With over 100,000 transactions per second (TPS) and sub-second finality, this is the first blockchain infrastructure for stablecoins operating at this performance level. And it's currently in private testnet with global partners including Visa, Deutsche Bank, OpenAI, and Shopify.

The Problem. Performance and Predictability

Current blockchain infrastructure wasn't built specifically for payments. Ethereum, Solana, and others were designed for general-purpose smart contracts, DeFi protocols, NFT marketplaces, and decentralized applications. They excel at these use cases. But when payment transactions compete for block space with NFT minting, gaming applications, and complex DeFi operations, the result is unpredictable fees and occasional congestion.

Tempo proposes dedicated infrastructure for stablecoin transfers, featuring isolated payment lanes (separate mempool), 100,000+ TPS throughput, sub-second finality, and transaction fees paid directly in stablecoins.

The Technology. Fundamental Transformation

Payment Lanes. Isolated Infrastructure for Stablecoins

Payment lanes are dedicated pathways that completely isolate payment transactions from other blockchain operations. This architectural separation ensures that payment processing doesn't compete for network resources with NFT minting, DeFi protocols, gaming applications, or other blockchain activities. For enterprises managing millions in daily transactions, this architectural isolation delivers consistent performance regardless of activity in other parts of the blockchain.

Sub-Second Finality. Technical Implementation

Tempo implements deterministic finality using Byzantine Fault Tolerant (BFT) consensus mechanisms. Once validators confirm a transaction, it achieves immediate and irrevocable finality. The transaction cannot be reversed or reorganized. This performance is achievable through Tempo's Proof of Authority (PoA) consensus model, where validator authority derives from institutional reputation rather than token staking.

Gas Fees in Stablecoins. Single-Asset Transaction Model

Most blockchains require users to hold two separate assets for transactions: the asset being transferred and a native token for gas fees. Tempo eliminates this requirement entirely. Transaction fees are paid in the same stablecoin being transferred. Sending USDC incurs fees paid in USDC. Sending USDT incurs fees paid in USDT. This functionality is implemented through an Automated Market Maker (AMM) integrated at the protocol level.

Reth. Technical Foundation

The technical foundation is Reth, Paradigm's open-source Rust implementation of an Ethereum execution client. Reth represents the fastest available Ethereum client implementation. EVM compatibility means existing Ethereum smart contracts and development tools can be used on Tempo with minimal modification.

Dankrad Feist. Ethereum Scaling Architecture Expertise

On October 17, 2025, Dankrad Feist announced he was joining Tempo as a senior engineer, transitioning from his full-time role at the Ethereum Foundation. Feist is a prominent figure in Ethereum development circles, known primarily for his work on scaling solutions—Danksharding, PeerDAS, and extensive work on Layer 1 scaling mechanisms.

His stated reasoning for the transition: "I believe that the real world moment is now, and I want to make sure we do not miss this window to touch normal people's lives everywhere in the world." His transition generated significant discussion within the cryptocurrency community. Vitalik Buterin publicly expressed support for Feist's decision.

The Global Ecosystem. Institutional and Technology Participation

Visa, Deutsche Bank, Standard Chartered, Revolut, Nubank, Shopify, DoorDash, Coupang, OpenAI, and Anthropic participate as design partners. The combined user base reaches hundreds of millions of users globally. Tempo's technical specifications—sub-second finality and low transaction costs paid in stablecoins—could support high-frequency micropayments required for autonomous AI agent transactions.

Leadership and Organizational Structure

Matt Huang serves as CEO of Tempo while maintaining his role as co-founder and managing partner at Paradigm. Since 2021, he has held a board seat at Stripe. Patrick Collison, Stripe CEO, has led Stripe's deliberate cryptocurrency strategy: 2018 (discontinues Bitcoin), 2021 (Matt Huang joins board), 2024 (restarts crypto support), February 2025 (acquires Bridge for $1.1 billion), September 2025 (announces Tempo blockchain development). This represents systematic vertical integration: Bridge plus Privy plus Tempo plus Stripe equals integrated infrastructure spanning from blockchain protocol to merchant payment acceptance.

Assessment. Specialized Infrastructure Development

Tempo represents infrastructure development targeting specific requirements of payment-focused blockchain applications rather than general-purpose blockchain platforms. The project combines experienced leadership, institutional participation (direct involvement from major financial institutions as design partners and likely validators), operational infrastructure (functional system in private testnet), technical differentiation (architecture choices specifically targeting payment use cases), and strategic integration (positioning within Stripe's broader cryptocurrency strategy).

Blockchain-based payment infrastructure continues developing across multiple approaches. Payment-specialized infrastructure like Tempo optimizes specifically for stablecoin settlement. Enterprise adoption will likely depend on alignment between architectural characteristics and specific institutional requirements.

The financial technology landscape is experiencing a seismic shift. As you read this, a new digital payment infrastructure is being built—one that fundamentally changes how humans interact with commerce.

This isn't about replacing human decision-making. It's about giving people unprecedented control over how they delegate and monitor transactions executed by AI agents on their behalf. Crypto fintech is emerging as the crucial bridge, the gateway that allows humans to remain in the loop while enabling AI agents to handle the execution.

The transformation isn't "humans vs. machines" or even "humans replaced by machines." It's about crypto fintech becoming the control layer, the interface where human intent meets machine execution. Where we, as humans, define the parameters, set the limits, grant permissions, and maintain ultimate authority, while AI agents handle the searching, comparing, negotiating, and transacting. Always human-in-the-loop, but with a fundamentally new way of delegating and controlling the transactions that AI agents execute on our behalf.

The Convergence. When AI Meets Payments

This shift is happening because three simultaneous revolutions are converging: the emergence of capable AI agents that can act autonomously within defined parameters, the maturation of blockchain technology that enables programmable, transparent money, and the urgent need for payment systems that let humans delegate transactional authority without surrendering control.

Google is rolling out the Agent Payments Protocol (AP2) with backing from over 60 industry leaders. Coinbase is breathing life into the long-dormant HTTP 402 status code through their x402 protocol, enabling direct stablecoin payments where humans approve once and agents execute within bounds. Meanwhile, Stripe is masterfully blending crypto infrastructure with traditional banking rails through their acquisition and expansion of Bridge, creating a seamless bridge between two worlds that were once thought irreconcilable.

On the AI front, Sam Altman demonstrated at OpenAI's Dev Day how AI agents will independently execute and complete tasks in external applications. Anthropic's Model Context Protocol (MCP) has filled a critical gap, connecting and energizing the fintech, e-commerce, and software development communities that are now racing to implement agents for everyday businesses.

The Great Blending. Custodial Meets Self-Custodial

Stripe Bridge. The Custodial Approach. Stripe's strategy with Bridge represents an elegant solution for bringing stablecoins into mainstream commerce while maintaining the user experience that traditional finance has perfected. By acquiring Bridge for $1.1 billion, Stripe gained a platform that handles the complexity of blockchain infrastructure entirely behind the scenes. Bridge operates in a custodial model, holding user stablecoin balances (USDC from Circle or USDB from Bridge itself) and enabling seamless conversions between stablecoins and regional fiat currencies.

Coinbase x402. Preserving Self-Custody. In stark contrast, Coinbase's approach with x402 and CDP Wallets maintains the crypto ethos of self-custody while still enabling the seamless experiences that modern commerce demands. The x402 protocol revives HTTP status code 402 "Payment Required," turning it into a production-ready standard for stablecoin payments that work directly over HTTP. What makes x402 revolutionary is its simplicity. A server responds with a 402 status, including payment details. The client signs the payment using their own wallet—mone.my or MetaMask—and retries the request with proof of payment. Settlement happens on-chain in seconds with zero facilitator fees, and users maintain complete control of their private keys.

When combined with Google's Agent Payments Protocol, x402 enables something extraordinary. Users can grant spending authority to AI agents through web3 signed "mandates" without ever surrendering custody of their funds.

Google's Master Play. Agent Payments Protocol (AP2)

Google isn't merely building another payment system—they're establishing the lingua franca for agentic commerce. AP2, developed in collaboration with over 60 partners including Mastercard, American Express, PayPal, Coinbase, and major Southeast Asian platforms, provides a framework for secure, auditable payments initiated by AI agents.

The protocol addresses three critical questions that autonomous payments raise: Authorization (did the user actually permit this?), Authenticity (does this reflect genuine user intent?), and Accountability (who's responsible if something goes wrong?). AP2's answer is a system of "mandates"—cryptographically signed, tamper-proof credentials that create an irrefutable audit trail.

In the coming days and weeks, we'll witness Google Search transform into an agentic conversational application, one that doesn't just find information but can act on it—including making purchases on your behalf.

The Real Impact on Traditional E-commerce

For traditional e-commerce businesses, the writing is on the wall. The Model Context Protocol isn't just another technical standard, it's potentially a new distribution channel. Possibly the most important one emerging since mobile apps. With ChatGPT's growing share of product search and Google's transformation into an agentic platform, implementing MCP isn't optional—it's existential.

The Self-Custodial Future. Agents That Shop While You Sleep

Imagine this scenario: You open your mone.my wallet and sign an "Intent Mandate"—a web3 signed authorization that says "Find me Nike Air Max 270 in size 42, maximum price $100, from verified retailers only, valid for 30 days." You approve a spending allowance of 100 USDC from your wallet. Then you close your laptop and go to sleep.

Your AI agent immediately gets to work. It queries MCP servers from dozens of retailers, comparing prices, checking inventory, and validating merchant reputations. At 3 AM, it finds a perfect match, an authentic pair for $89 USDC at a verified retailer. The agent validates that this meets all your mandate constraints, generates a "Cart Mandate" with the specific purchase details, and executes the payment directly from your wallet using the x402 protocol.

When you wake up, you have a confirmation: product purchased, $89 USDC spent (saving you $11), order confirmed, tracking number provided, and complete blockchain receipt showing exactly what happened. All while you were sleeping. All from your self-custodial wallet. All within the parameters you set. This isn't science fiction—this is what AP2 + x402 + MCP enables today.

The Path Forward

The transformation happening right now will reshape e-commerce, redefine fintech, and create entirely new categories of businesses. The companies that will succeed are those that understand a fundamental truth. We're not just adding AI features to existing products. We're reimagining commerce itself for an era where your digital representative can negotiate, transact, and optimize on your behalf—24/7, across global markets, all while you maintain complete control.

The future of commerce isn't human-to-business. It's human-delegating-to-agent-to-business, powered by programmable money, operating at the speed of code, with humans always in control. And that future is already here.

CalCompute is back like a boomerang. California wants to democratize AI – but what does that really mean?

Just a few days ago, on September 29, 2025, California Governor Gavin Newsom signed SB 53 – the Transparency in Frontier Artificial Intelligence Act. The legislation introduces new transparency standards for operators of the largest LLM models: requiring public disclosure of security protocols, reporting of critical incidents, and protecting whistleblowers at AI companies.

AI regulation, while still controversial, is slowly becoming a global standard. Europe is sharpening its teeth – the AI Act is gaining force, and EU countries are gradually implementing its provisions. California, historically a pioneer in tech regulation (remember CCPA or net neutrality?), is once again charting its own course in the face of federal decision-making paralysis.

But there's one element of SB 53 that intrigues me far more than yet another compliance requirement – CalCompute.

A story that keeps coming back

CalCompute isn't a new idea. It first appeared in February 2024 as part of a far more controversial bill, SB 1047, authored by State Senator Scott Wiener. That bill was a lightning rod in the tech world – it aimed to impose strict regulations on creators of "frontier AI models," meaning models whose training costs exceed $100 million.

SB 1047 required: pre-deployment safety testing, audits by independent firms, penalties reaching up to 30% of model training costs for repeat violations, and legal liability for creators – including open-source developers.

Reactions? Mixed. Elon Musk supported it. Nancy Pelosi called it "more harmful than helpful." Meta's Yann LeCun warned it would kill open-source models. Anthropic (creators of Claude) was cautiously positive, but with reservations.

In September 2024, Governor Newsom vetoed SB 1047. The reason? Too restrictive, too much legal uncertainty, risk of pushing innovation out of California.

But there was one thing in that bill everyone agreed on – CalCompute.

What exactly is CalCompute?

CalCompute is a project to create a public computing cluster – a supercomputer that would be accessible to: academic researchers, startups without budgets for multi-million dollar infrastructure, community groups working on AI in the public interest, and institutions focused on disaster forecasting or medical research.

Why does this matter? Training advanced AI models requires massive computing power, costing tens, often hundreds of millions of dollars. This creates a natural barrier to entry – only tech giants (OpenAI, Google, Meta, Anthropic) can afford it. Smaller players, even with brilliant ideas, are outmatched from the start.

CalCompute aims to change that. The idea is simple: democratize access to computing power so AI isn't the exclusive domain of a few Silicon Valley corporations.

The consortium is set to be established at the University of California, comprising representatives from academia, labor unions, AI experts, ethicists, and public interest advocates. By January 2026, they're expected to present a detailed plan: how much it will cost, how large the cluster should be, who will have access, and under what terms.

Why did CalCompute survive while SB 1047 failed?

Senator Wiener learned his lesson. After the SB 1047 veto, Governor Newsom convened a group of experts (including AI "godmother" Fei-Fei Li from Stanford), which released a report in June 2025. The conclusion? AI regulations are needed, but they must be empirical, flexible, and based on a "trust but verify" approach.

SB 53 is exactly that version 2.0 – taking the least controversial elements from SB 1047: whistleblower protections, CalCompute, transparency requirements (instead of harsh penalties). And it leaves the door open for evolution – California's Department of Technology is required to annually update the regulations based on technological developments and stakeholder consultations.

Sounds great. But does it really?

The idea of AI democratization is brilliant. A world where not only OpenAI or Google decide what models are created and what they're used for is a better world. CalCompute could: level the playing field for smaller players, enable research on AI in the public interest (medicine, climate, crisis management), and balance excessive concentration of power in Big Tech's hands.

But there's also a flip side. What happens when the government starts deciding who gets access to computing power and who doesn't? Who gets a slot on CalCompute's calendar – and according to what criteria? Could the concept of "AI in the public interest" become, in practice, what those in power deem a "desirable" use of technology?

And there's an even more fundamental question: CalCompute could become the de facto standard for what AI can and cannot be. If the public cluster gains dominance as the training infrastructure for AI, then the government – not the market, open-source community, or researchers – becomes the Oracle defining the boundaries of technology.

What's next?

SB 53 is just the beginning. This is pioneering regulation – the first of its kind in the US – but the world is watching. If CalCompute succeeds, we might see similar initiatives in other states or countries. If it fails – it will serve as a warning of how easily good intentions can morph into government overreach.

I'm a proponent of AI regulation. But I'm also skeptical of simple solutions to complex problems. CalCompute sounds good on paper, but the devil – as always – is in the implementation details. California is experimenting. We're watching.

Just a few days ago, Google quietly published something that might turn out to be one of the most important milestones in the evolution of AI and Web3. On its official blog, the company announced the Agent Payments Protocol (AP2).

The very first words they used set the stage: Authorization. Authenticity. Accountability.

It almost reads like the opening lines of a Web3 whitepaper. The core values of decentralized technology are now being adopted as the foundation for a new open standard designed to allow AI agents to transact safely on behalf of their users.

Agents Enter the World of Transactions

Up until now, AI agents have mostly been about productivity, creativity, and automation. They could write, code, generate images, and optimize workflows. But commerce? That was always the missing piece.

With AP2, agents gain the ability not just to identify themselves and prove authorization, but to actually execute financial transactions. This is the moment when interactions between AI agents move to a completely new level: they can now buy, sell, and settle payments in real time.

And here's where one detail from Google's announcement becomes especially important. Among the more than 60 organizations backing AP2 is Coinbase, bringing its X402 protocol into the mix. X402 is designed for micropayments and API monetization, giving AI agents the ability to pay "as they go" instead of locking into subscriptions. Combined with stablecoins like USDC, this means an agent can autonomously spend a few cents for an API call, access to market data, or compute power instantly, securely, and without human intervention.

Why AP2 Matters

Picture this: you ask your AI agent to organize a vacation. It finds the best flights, books a hotel, and arranges a rental car within your budget. Sounds convenient—until the moment comes to pay. Until now, a human was always required to enter card details or approve the transaction.

AP2 changes that by solving three critical problems:

Authorization – The agent can only act within the boundaries of your explicit, cryptographically secured consent.

Authenticity – Every request for payment matches your true intent, preventing errors or manipulation.

Accountability – The system makes it clear who is responsible if something goes wrong: the user, the agent, or the merchant.

In short, AP2 provides a universal language of trust for transactions between AI and the financial system.

Web3 as the Natural Backbone

If this all sounds familiar, it's because blockchain and Web3 have been building exactly these mechanics for years: immutable public ledgers, programmable money, smart contracts that guarantee enforceability of agreements.

It's hard not to see how much AP2 borrows from the Web3 playbook. The difference is that now, these ideas are being implemented by one of the largest players in tech—and aimed directly at making AI agents economically autonomous.

Stablecoins and Coinbase X402

One of the most exciting aspects of AP2 is its native support for stablecoins, but that alone isn't enough for full autonomy. To really enable agents to pay, settle, and monetize without human friction, you need a payments protocol built for machine-to-machine, real-time, microtransaction scenarios. That's where Coinbase's x402 enters.

What is x402 exactly? Internet-native payments using stablecoins over HTTP. It resurrects the long-unused HTTP 402 ("Payment Required") code so that when a client (which might be a browser, app, or AI agent) requests a paid resource, the server responds with 402 including pricing info, supported tokens (e.g., USDC), destination wallet addresses etc. After the 402 response, the client sends a payment payload (signed), then repeats the request with a header (e.g., X-PAYMENT) including the payment info. The server or a facilitator then verifies the payment on-chain, settles it, and returns a normal 200 OK with a payment receipt header.

Because AP2 is about establishing a standardized framework for agents to transact (authorization, authenticity, accountability), x402 fills in the payments piece: it's the stablecoin facilitator in AP2's ecosystem. Agents can not only talk and coordinate via AP2's standards but also pay each other for services, data, compute, etc., without human intervention. Micropayments by agents for resource usage become practical. Real-time monetization: content providers, SaaS, and API providers can charge exactly for usage (pay-per-use).

A New Opening for Web3

For years, the Web3 industry has been searching for its next big narrative after ICOs, IDOs, DeFi, metaverse, and gaming. AP2 may be it. Autonomous AI agents, empowered by stablecoins and smart contracts, represent a real business use case—not speculation, not hype. This is the kind of adoption that could redefine decentralized finance and finally bridge Web3 into mainstream commerce.

Not so long ago, the future of AI felt like chatting with a clever machine. We typed prompts, the model replied, and for a moment it seemed like magic. But if you've ever tried to use an LLM for real work—shipping code, running research, or managing projects—you know the truth. Conversation alone doesn't get the job done. It's like asking a colleague for advice when what you really need is a team that rolls up its sleeves.

That's where the story begins to shift. Instead of managing every step through prompts, we are learning how to delegate—not to a single model, but to a swarm of autonomous agents working together. What once sounded like science fiction is quickly becoming a practical reality. And just as graphical interfaces once replaced command lines, agent swarms are poised to transform how we work with digital tools, run experiments, and even build entire systems.

A Short History, Told Quickly

It began with chatbots—scripted if/then trees that pretended to converse. Then came LLM chat: far more natural, but still reactive. You asked; it replied. Useful, but cognitively heavy—you had to break work down into steps, write long prompts, and micro-manage outputs.

The next step is different: agents that plan, choose tools, and act on your behalf. A recent GAO report frames it clearly. Agents can operate autonomously to accomplish complex tasks and make time-critical decisions, potentially reshaping entire workflows.

The Agent Leap (and Why It's Happening Now)

Think of an agent as a digital teammate. It designs its own workflow, uses APIs, tools, and services as its "hands and eyes," and adapts when things change. Modern stacks give agents the ability to break a request into steps, call other services, retry when something fails, and report back with results.

IBM describes them well: agents automate multi-step goals by deciding, problem-solving, and executing—not just predicting text. And when you group them, you get agents coordinating as a team. One plans, another executes, a third reviews, a fourth checks compliance. It's not one giant model trying to do everything—it's a division of labor, closer to how real human teams work.

How Swarms Actually Work (Without the Hype)

The structure is surprisingly familiar:

Role specialization. Agents are assigned narrow jobs (fetch data, write copy, test code, check compliance).

Orchestration. Either a "lead" agent delegates tasks (hierarchy), or peers coordinate directly (decentralized).

Shared memory. Agents write to and read from a common workspace so context isn't lost.

Fault-tolerance. Timeouts, retries, and redundancy keep progress moving even if one fails.

This pattern is showing up in data engineering, product development, and operations. Powerdrill.ai calls swarms "intelligent middleware"—a layer between human intent and fragmented digital infrastructure, turning APIs, databases, and microservices into coherent outcomes.

From Prompting to Delegating

Here's the real transformation: prompting was a form of micro-management. You told the model exactly what to do, step by step. Delegation flips it around: you set the goal, the swarm handles the steps, and you step in only to review and steer.

That changes the cognitive load. Instead of thinking in instructions, you think in outcomes. The interface shifts too: from chat boxes to delegation surfaces—spaces where you define goals, constraints, and success criteria, while seeing the plan and progress unfold in real time.

Human-AI interaction research has long argued for this: keep humans first, design for transparency, and build systems that augment rather than replace. Delegation aligns with that principle.

The Future of Work (and Why It's More Human)

The irony is clear: the more agents automate, the more human skills matter. When machines handle execution, humans focus on framing problems, setting priorities, making trade-offs, and telling the story of outcomes. Delegation doesn't erase our role—it sharpens it.

Researchers and policymakers agree: the future is less about replacement and more about rebalancing. Some tasks vanish, but new ones—like orchestrating swarms, defining guardrails, and aligning outcomes with values—become central. The key skill isn't writing clever prompts anymore. It's learning how to delegate well.