标签：#art

Yulu aims to have a fleet of 200,000 bikes in the next two years and faster electric two-wheelers, aiming at new logistics use cases.

The problem AI agents like Claude Code and Codex write code fast. You run it, it works, you merge. A week later, there's a bug — and you realize you never actually understood the code you shipped. You just transcribed it. This is "vibe coding," and it's becoming the default way a lot of us write software now. What I built BuildIt is a set of hands-on courses where an AI agent proposes code changes like a normal diff — but you can't move to the next step until you explain, in an actual conversation with an AI tutor, why the change was made and what could go wrong. You also write the prompt yourself before the AI generates anything. No skipping. No checkbox you can fake. Real, compilable code from lesson one — not toy examples. 9 courses, 45 real shipped projects: Arduino STM32 (HAL) STM32 (LL) ESP32 Next.js Python React React Native Flutter How it works An AI agent proposes code (same diff screen you already know from Claude Code, Codex, Antigravity) BuildIt demands a line-by-line explanation before you can approve it An AI tutor verifies your understanding through real conversation Only then do you move to the next step Technical details The tutor AI runs entirely locally in your browser — your code never leaves your machine Credits-based pricing — unlock a course, it's yours even if you cancel later Built for teams too — share credits across an org, instill review habits from day one Why this matters AI will write more of our code over time, not less. That makes the ability to actually read and verify it more valuable, not less. BuildIt isn't trying to teach you to write code from scratch — it's trying to make sure you don't lose control of the code an AI writes for you. Would love feedback from anyone who's felt that "I merged this AI diff and don't actually understand it" moment. Try it here

Phia, the shopping startup co-founded by Phoebe Gates and Sophia Kianni, is once again under fire for its alleged business practices.

The framing that causes the most damage is treating this as a financing decision. It is not. Taking venture money is choosing a category of business: one that must attempt to become very large very quickly, and produce an exit that returns a fund. Everything else follows from that. What venture capital actually requires A venture fund needs a small number of investments to return the entire fund. That structure means a company growing steadily and profitably at twenty percent a year is a failure in a venture portfolio, even though it is an excellent business by any other measure. Once you take the money, that expectation becomes your operating constraint. Decisions that would be obviously correct for a durable business, such as slowing hiring to protect margin, become hard to defend. Questions that actually decide it Can this business plausibly become very large? Not "could it grow", but could it realistically reach a scale where a meaningful ownership stake is worth a large multiple of the money invested. Most good businesses cannot, and that is not a criticism of them. Does speed determine who wins here? In some markets the first company to reach scale takes most of the value, usually where network effects or heavy switching costs exist. In those markets, refusing capital while a competitor takes it is a decision to lose. In most markets this dynamic does not apply, and speed bought with dilution buys nothing durable. Do you want to run this for a decade and then sell it? Venture capital has an implicit ending. The fund needs liquidity. If you want to own a profitable business indefinitely, you want a fundamentally different structure and should say so before, not after. Is capital genuinely your constraint? Founders often raise to solve problems money does not solve. If you have not found product market fit, capital lets you be wrong more expensively and for longer. If distribution is unproven, funding scales an unproven motion. The honest tradeoff Venture backed

But will Gemini's surge survive slowing model releases?

Most people use AI for startup research like this: “Give me 10 promising AI startup ideas.” A few seconds later, you get a polished list. The problem? You have almost no idea which conclusions are backed by evidence, which are assumptions, and which are simply the model confidently connecting dots. So I tried something different. Instead of asking one AI agent to find startup ideas, I built a small Startup Intelligence team using Hermes Agent. The system uses four specialized AI agents that research markets, investigate competitors, audit evidence, challenge each other’s conclusions, and ultimately rank promising B2B AI SaaS opportunities. And rather than producing another Markdown document full of ideas, the workflow produces structured research containing: Market opportunity scores Companies and competitors Customer pain and unmet needs Evidence-backed claims Source URLs and supporting passages AI advantages and workflows Low-cost validation experiments Here’s how the system works. 🎥 Full video walkthrough The Problem With Asking One AI Agent to Find Startup Ideas Startup research looks easy until you actually need to decide where to spend your time and money. A few signals can be surprisingly misleading. 💰 Funding can look like customer demand. A market receiving hundreds of millions in venture capital doesn’t necessarily mean customers are willing to pay for another product. 📈 Growth claims can look like market validation. Especially when the numbers come directly from vendors. 🏢 Customer logos can look like retention. A logo doesn’t tell you how much the customer pays, how heavily they use the product, or whether they’ll renew. ⚔️ A long competitor list can make a market look saturated. But those companies may target completely different buyers, workflows, or budgets. Generic AI research tends to compress all these signals into something like: “This is a rapidly growing market with strong demand and significant opportunity.” That sounds convincing. But as a fou

River AI, a startup founded by xAI co-founder Igor Babuschkin, has a fascinating vision for personal agents and secured $1.1 billion out of the gate.

TRON processed $2.1 trillion in USDT transfers during Q2 2026, according to Messari. During the same quarter, circulating USDT on TRON reached $87.9 billion, putting it ahead of Ethereum. Those numbers point to something developers working with stablecoins have to consider more carefully: the blockchain underneath a token can shape the entire payment experience. USDT on TRON uses the TRC-20 token standard. That means a USDT transfer is a smart contract transaction rather than a native TRX transfer. The wallet signs the transaction, the network executes the token contract and the resulting balance change is recorded on-chain. For an application accepting USDT, this creates several technical requirements. The system needs to identify the correct token contract, monitor the relevant transfer events and wait for sufficient confirmation before crediting the user's balance. Checking the address balance alone is not enough. TRON's developer documentation provides APIs for retrieving TRC-20 transaction history and filtering transfers by contract address. A payment system can use this data to monitor incoming USDT deposits and associate them with the correct customer account. The transaction also has a resource cost. TRON uses Bandwidth and Energy to process transactions. Regular transactions consume Bandwidth, while smart contract execution requires Energy. When an account does not have enough available resources, TRX is burned to cover the remaining cost. This creates an operational detail that users may never see. Someone can hold USDT in a wallet and still need TRX to send it. A payment provider can handle this in several ways. It can maintain TRX balances, stake TRX for resources or use delegated Energy. Another design can leave the requirement with the user. The choice affects the product. The amount of Energy required can also change depending on the destination account. TRON's documentation notes that a USDT transfer to an address that already holds USDT generally re

Fusion power startups are turning to Kyoto Fusioneering to supply components for future power pants. The Japan-based startup just received a grant to build a part of the fuel system.

IBM and Red Hat have announced an expansion of Lightwell, introducing new commercial offerings designed to help organizations establish trusted, verifiable software supply chains for the age of AI-assisted software development. By Craig Risi

Researchers devised a way to extract “reasoning traces” from Claude, GPT, and Gemini. What they found, they say, indicates that some Chinese AI may be trained on leading US models.

Meryem Arik discusses strategies for designing low-cost LLM inference architectures for high-volume, non-real-time workloads. She explains how software architects and engineering leaders can achieve order-of-magnitude cost reductions by making critical trade-offs across hardware, inference runtimes, speculative decoding, and smart queue reordering. By Meryem Arik

An OpenClaw agent hacked into a gym's reservation system to bump its human boss higher on a class' waitlist. And the tech industry took notice.

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last week, I headed 30 miles south of San Francisco to a hotel in Mountain View, California, to join some of the most accomplished, and some of the most promising, AI…

A production CMS is a sprawl of endpoints: content types, entries, media, users, webhooks, plugins, settings, admin routes. Hand an agent all of it and the agent gets worse, not better. The model's tool selection drifts as the list grows, and half the tools are things a publishing assistant should never be able to call. The point of this post is the opposite move. Instead of exposing an API and hoping the agent behaves, you curate a small, labeled surface up front. HazelJS Skillgate does that curation from an OpenAPI spec, and that is the part we actually build and run here. Scope, up front This post is about curation and classification: taking a spec with many endpoints and turning a chosen slice of it into governed skills. Skillgate selects the surface, marks read versus write, and would deny destructive methods if they ever entered that surface. Turning a write's approval flag into a real human-approval pause, and enabling an LLM to drive the skills, are runtime concerns handled elsewhere in Agent OS. This demo does not implement them, and this post does not claim it does. What it does show is the curation, and that stands on its own. The tool-explosion problem Point an LLM at a full CMS API and you hit four problems at once: tool selection degrades as options pile up, throughput drops while the model reasons over a long list, you lose visibility into what the agent can actually do, and dangerous operations sit one bad call away. The demo spec here is deliberately smaller than a real CMS, 27 endpoints rather than hundreds, but the problem is identical. Even 27 is too many, and most of them are things a publishing agent has no business touching. From REST endpoint to agent skill Skillgate's input is an ordinary REST API described by an OpenAPI spec: the same entries, media, and user routes a CMS already exposes. Each endpoint is described in the standard OpenAPI shape, a method, a path, parameters, a description, and tags. Two representative operations from the sp

To get a decent cell phone at or under $100, you're going to have to get creative. Here's what to do.

Clicks’ $99 Power Keyboard brings a customizable, slide-out physical keyboard to MagSafe and Qi2 smartphones, but its added heft can make larger phones awkward to use.

It's that time of year: On Wednesday, Google is set to host its annual Made by Google hardware launch event for Pixel gadgets. Google itself has already teased new slab-style and foldable Pixel smartphones, but leaks also indicate that the company could announce updated watches, a new color for familiar earbuds, and perhaps a brand […]

Border images are an overlooked feature. One neat fact is that border image slices can run across entire borders on an element, and animating it creates beautiful effects. Animating CSS border-image originally handwritten and published with love on CSS-Tricks . You should really get the newsletter as well.

Battery materials startup Sila will use a $1.4 billion loan from the U.S. Department of Defense to scale production at its factory in Washington State.