SUCCESS

The 83rd Venice Film Festival will feature projects from Danny Boyle, Werner Herzog, Martin McDonagh, and Alex Gibney, and an immersive project about the L.A. Fires

Wealth has always had two answers to a tax bill. Pay it, or move somewhere the bill cannot follow.

For most of American history the second answer was the expensive one. It required lawyers, accountants, a persuasive paper trail, and usually a second house in a state with better weather and a friendlier revenue department.

Then a third answer appeared, and it turned out to be the cheapest of the three. You do not pay and you do not move. You buy the election that decides whether the tax exists.

California is stress testing that idea this year. The state holds more billionaires than any other, roughly 200 of them, and a measure on the Nov. 3 ballot would take a one-time 5% slice of their net worth. Ninety percent of the money would fund the state health care program, according to NPR.

The exodus got the headlines. At least six billionaires established residency elsewhere before the deadline that mattered, including Google co-founders Larry Page and Sergey Brin, according to Fortune.

What the roughly 200 who stayed did next is the part that should hold your attention. And one of the six who left is bankrolling it.

Brin has given $102 million this year to Building a Better California, the group leading the campaign against the measure, according to data from the California secretary of state.

Brin spent $102 million opposing Proposition 40 but Berkeley polling shows billionaire tax still leads.FREDERIC J. BROWN / Getty Images

What $102 million buys in a ballot fight

I ran Brin’s spending against his own exposure, and the ratio is what makes this worth following.

His net worth stood at roughly $276 billion on Aug. 16, according to the Bloomberg Billionaires Index. A 5% levy on a figure that size lands somewhere near $13.8 billion.

Measured against that, $102 million is about three quarters of one percent. Priced as insurance, it is cheap.

Here is how the money and the polling stack up:

Brin has contributed $102 million this year to Building a Better California, including a $20 million donation in a recent filing, according to state campaign data reported by CPA Practice Advisor.

The union-backed campaign supporting the tax has raised just over $30 million, according to Fortune.

Proposition 40 led 48% to 41% among likely voters in a survey released Aug. 14, according to the UC Berkeley Institute of Governmental Studies.

Brin’s fortune has grown by roughly $26 billion so far in 2026, according to the Bloomberg Billionaires Index.

Building a Better California describes itself as a nonpartisan group focused on “supporting forward-looking ideas to improve affordability and quality of life,” according to the organization’s website. It says Proposition 40 would damage the state economy and cost the budget billions without lowering health care costs, according to its published FAQ.

Related: J.P. Morgan flags gathering storm in U.S. wealth taxes

The other side frames the same spending very differently. Brin “would rather spend $100 million to fund a shady opposition campaign” than pay the levy, said Service Employees International Union United Healthcare Workers West Vice President Debru Carthan, in a statement reported by Common Dreams.

Brin has explained his position in personal terms. He said he “fled socialism with my family in 1979” and does not want California heading the same direction, he told the New York Times in April, in comments cited by Fortune.

Why leaving California may not settle the bill

The mechanic that makes this fight strange is buried in the measure’s residency language, and it is the part most worth understanding if you have ever thought about moving for tax reasons.

Proposition 40 applies to anyone who was a California resident on Jan. 1, 2026. There is no proration for leaving in February, or June, or the week before the election. Net worth is then valued as of Dec. 31, 2026. “This retroactive residency date is likely to be challenged in court,” according to the California Budget and Policy Center. 

More Taxes:

A guide to the capital gains tax rate: Short-term vs. long-term capital gains taxes

Secretary Bessent cracks down on non-profit tax loopholes

IRS rules may fail to protect Americans from one AI tax risk

That date passed nearly eight months ago. Anyone who packed a moving van in December beat it. Anyone who moved in January did not.

So the spending makes sense even for someone who already left. A residency change is a defense you have to win in an audit. Killing the measure outright needs no lawyer at all.

Which is what Propositions 41 and 42 are for. Proposition 41 would bar new taxes enacted after Jan. 1 of this year. Proposition 42 would prohibit both wealth taxes and retroactive taxes outright, according to KQED. If a countermeasure and Proposition 40 both pass, the one with more votes takes effect.

Brin is not alone in funding that strategy. Ripple co-founder Chris Larsen, PayPal co-founder Peter Thiel, and venture capitalist Ron Conway have all spent against the measure, a pattern TheStreet tracked when Ripple quietly moved millions into the opposition earlier this year.

What the polling says about all that money

What struck me when I pulled the latest Berkeley numbers is how little the spending has moved.

Proposition 40 holds 48% support against 41% opposition among likely voters. The measure “is shaping up to be a closely fought contest,” said Institute of Governmental Studies co-director Eric Schickler, according to Berkeley News.

Seven points is not a comfortable lead. Measures sitting under 50% in August often lose, because undecided voters break toward no. But it is a lead, and it has survived a spending gap of better than 3 to 1, a fight TheStreet valued at $100 billion when the measure first qualified.

The more useful finding sits underneath. Voter awareness of Propositions 41 and 42 is low. That is the cheapest ground left to buy, and where the remaining money will go.

The politics are not clean on either side. Gov. Gavin Newsom opposes the tax, arguing it erodes the state’s revenue base over time, while the California Democratic Party officially supports it. The California Teachers Association is opposed. Sen. Bernie Sanders and Rep. Ro Khanna are in favor.

Why this state tax fight matters to your wallet

If you are reading this from Ohio or Georgia, the temptation is to file this under California problems. I would not.

Statehouses have been circling wealth and high-earner taxes for two years. Maine enacted a surcharge on income above $1 million in April, and Washington and Illinois have floated versions of the same idea. J.P. Morgan Private Bank flagged California’s measure as the most advanced example of a trend running well past one state, an analysis TheStreet covered when the measure qualified for the ballot.

The transferable lesson has nothing to do with billionaires. It is the residency snapshot.

Most people assume leaving a state ends its claim on them. It does not, and never really has. California audits departing residents on domicile, not mailing address, and the questions get granular. Where do your kids go to school. Where do you garage the car.

Proposition 40 just wrote that logic into a single fixed date, which is what makes it a template. If it survives the courts, expect other states to copy the date and skip the audit.

That is the outcome worth watching on Nov. 3, and it will not be settled that night. Whichever way the vote breaks, the lawyers are already booked.

Related: California’s billionaire tax ignites $100B fight

You’re in an aisle of a bookshop, airport store, or supermarket and have your hands full. You don’t bring out your phone; you glance at something and ask Siri to remember it.

That is the kind of interaction Apple may be preparing for its next generation of AirPods.

Unreleased AirPods with a camera appear to be employing a technology called Visual Intelligence in a demo movie in the macOS Tahoe 26.7 release candidate, according to The Verge. The clip shows someone wearing the glasses looking at a book while Siri says visual information can be saved for later.

That would push AirPods well beyond music, calls, and noise cancellation.

Apple (AAPL) currently sells AirPods Pro 3 starting at $249, according to the company, as Mashable reported. The leaked model hasn’t been officially revealed, and Apple hasn’t mentioned pricing, so investors shouldn’t assume the camera-equipped version will be the same price.

Still, $249 is a useful baseline for what millions of people already accept as the price of a premium Apple accessory.

The more important question is whether people pay more for headphones that could actually work as a second pair of eyes for Siri.

Apple may be turning AirPods into an AI interface

Reports say the cameras aren’t really for snapping images.

Rather, they seem intended to provide visual information to Siri and Apple’s artificial intelligence algorithms so the assistant knows what the user is looking at. This may lead to more natural encounters later.

More Apple:

Apple rewrites how Americans pay for iPhones

Bank of America doubles down on Apple stock ahead of earnings

Apple’s iPhone 18 AI bet hides a pricing trap

A traveler might inquire about a sign without pulling out a phone. A shopper may tell Siri to recall a product. A blind user could be presented with extensive descriptions of the environment.

Those use cases are still a bit speculative, since Apple has not launched the product. But the leaked demo gives investors a better feel for what Apple is trying out: AI that understands the real world without forcing the user to look at a screen.

That is important, since AirPods are already in a highly coveted spot. They are worn on the body and already include microphones. Plus, users keep them in for hours.

Apple says the AirPods Pro 3 now offer up to eight hours of listening time with active noise cancellation, plus heart-rate tracking and hearing-health capabilities.

Adding visual intelligence would move the product from “headphones” to wearable computers.

Apple may be giving AirPods a pair of eyes.VALERIE MACON / Getty Images

Human appeal of camera-equipped AirPods creates privacy problem

The promise is convenience.

The worry is that a camera in an earbud is much less noticeable to other people than a phone held up or smart glasses on someone’s face.

That might present a difficult trade-off for Apple.

Although The Verge observed that the leaked design doesn’t reveal whether there’s any kind of recording indicator, it pointed out that the cameras seem designed for low-res visual context rather than photography.

Apple has made privacy a big part of its brand, and the company would certainly face issues about when the cameras are on, what data is saved, and if passersby can detect whether the system is interpreting the environment.

That may be as important as the gear.

While consumers may adore the concept of asking Siri what they’re seeing, they might feel quite differently about everyone around them wearing invisible cameras.

What consumers should know

$249: Current starting price of AirPods Pro 3.

Up to 8 hours: AirPods Pro 3 listening time with active noise cancellation.

Visual intelligence: Feature shown in the leaked camera-AirPods demo.

September: Reports suggest Apple could potentially unveil the product as early as its fall hardware event, though Apple has not confirmed that timing.

It’s easy to miss the bigger picture.

Apple doesn’t need to replace the iPhone with AirPods with cameras. It just needs to limit the number of times users have to take the iPhone out.

That’s a whole different product ambition.

The iPhone put a computer in your pocket. Now Apple may be studying whether AirPods could literally put AI in front of your eyes without putting another screen in your face.

Related: Apple may finally have a $2,000 reason to enter foldables

If wise people avoid problems by learning from the mistakes of others, then OpenAI’s latest ChatGPT update may be the wisest move the company has ever made.

Meta Platforms is in the news as the tech giant goes on trial over accusations that it purposely got teens and children addicted to its social media platforms. The company could potentially face over a trillion dollars in fines over the allegations.

Meta is accused of knowing that its platform was detrimental to teens’ mental health for years while it publicly espoused that the opposite was true.

On Aug. 18, OpenAI unveiled its plan to keep teens safe on its own platform while being at least somewhat honest about the potentially detrimental effects its large language model can have on young users.

While OpenAI has already faced lawsuits over teen deaths, including a case where ChatGPT allegedly advised a teen to take his own life, it is at least using this opportunity to tighten up its own liability.

OpenAI childproofs ChatGPT

On Aug. 18, OpenAI revealed it is making extensive changes to ChatGPT in order to keep teens and young people from harming themselves through its platform.

In its announcement, the company recognized that students today will be the first generation to grow up with artificial intelligence as an everyday part of life. And it says it recognizes the responsibility that places on OpenAI.

Related: OpenAl is reinventing ChatGPT to justify its $1 trillion IPO

According to OpenAI, while the vast majority of students are using AI in the classroom, only 16% of high school leaders say that all of their students are learning the technical knowledge to understand it.

So, in addition to starting partnerships with CodeAI to help students and educators better understand ChatGPT, OpenAI announced the launch of ChatGPT for Teens, which the company describes as “a dedicated experience designed with learning at the center to help teens think critically, deepen understanding, and use AI with confidence.”

What changes does ChatGPT for Teens make?

ChatGPT for Teens will be just like the original platform, except for the built-in mechanisms that OpenAI says will help protect kids.

Those mechanisms include features that will give parents more control over the experience and more guidance for the teens on how to use the technology safely.

“Parents and educators want young people to be ready for a future shaped by AI—and they are right to expect that to come with care,” Dr. Allison Mishkin, Head of Child Development, OpenAI, said in the company announcement. 

“Our goal is not simply to help more students use AI, but to help them ask better questions, think critically about the answers they receive, and use these tools responsibly,” Mishkin continued. “We look forward to working alongside CodeAI, parents, and educators to learn, improve, and help young people build those skills.”

More AI:

Nvidia just made a move Wall Street wasn’t ready for

Microsoft just took sides in AI policy fight

OpenAI just disclosed something genuinely alarming

Those features include “study mode,” which uses “guiding questions, scaffolding, metacognitive prompts, and knowledge checks” to make sure students aren’t using ChatGPT to just cheat.

ChatGPT for Teens also features homework reminders, quizzes, learning visualizations and “study hours” where parents can make study mode a default for however long the feature is on.

OpenAI says it will also share more of its safeguarding work publicly, starting with new under-18 evaluations that they’ve added to its system cards that show how its models perform against teen-specific standards in “challenging sensitive-content cases.”

Meta in spotlight over lack of teen controls

OpenAI may have been planning to roll these features out this week for a while, but the timing of the announcement is noteworthy considering AI rival Meta is going to trial this week over its own child-related issues.

Four states – California, Colorado, Kentucky and New Jersey – are seeking up to $1.4 trillion in penalties against Meta, which they claim violated consumer protection laws, including the Children’s Online Privacy Protection Act.

Meta’s last-ditch effort to stop the trial was rejected last week by the Ninth Circuit Court of Appeals, which ruled that Meta did not have the “immunity” from liability its defense asserted under Section 230 of the Communications Decency Act.

Meta contends that the district court’s denial of its Section 230 immunity is immediate grounds for appeal should it lose the trial.

“On the eve of trial, Meta has resorted to seeking an emergency stay in the appellate courts. Meta’s latest effort to get out of taking accountability has failed — again,” said California Attorney General Rob Bonta.

The lawsuit was originally filed in 2023.

Meta filed a motion to dismiss in 2024 that was rejected and also attempted to obtain a summary judgment in June that would have terminated the lawsuit.

OpenAI launches ChatGPT for Teens.hapabapa / Getty Images

What is Meta accused of doing?

Back in 2023, a coalition of 33 attorneys general led by California’s Rob Bonta filed a lawsuit in the U.S. District Court for the Northern District of California alleging that Meta “designed and deployed harmful features on Instagram and Facebook that addict children and teens to their mental and physical detriment.”

According to Bonta, “Meta has been harming our children and teens, cultivating addiction to boost corporate profits.”

The lawsuit claims that Meta created a business model “focused on maximizing young users’ time on its platforms” while it also used “harmful and psychologically manipulative” features to keep them addicted. It also claims that Meta misled the public about the safety of those features.

According to the lawsuit, those harmful and psychologically manipulative features include:

Dopamine-manipulating recommendation algorithms.

“Likes” and social comparison features known by Meta to harm young users.

Audiovisual and haptic alerts that incessantly recall young users to Meta platforms.

Visual filter features that promote body dysmorphia.

Infinite scrolling designed to discourage users’ ability to self-regulate.

For its part, Meta says the case misrepresents the lengths the company goes to to protect children on its platform.

“We work hard to keep people safe on our platforms and have been transparent about the challenges of identifying and removing bad actors and harmful content,” Meta said.

“We remain confident in our record of protecting teens online and will continue to defend ourselves against claims that misrepresent the facts,” the company stated.

Related: Meta stock battered by biggest legal threat yet

Cerebras is betting that its specialized chips will outpace traditional graphics processing units as AI agents become more widespread.

Traditionally, when you talk about McDonald’s and gas, you’re not referencing filling up at the pump.

The fast-food giant, however, has made a deal to give consumers discounted gas in partnership with Shell. That move is part of the chain’s efforts to deliver broader value to customers without necessarily having the lowest prices.

“We’ve listened to customers and adjusted along the way with a relentless focus on delivering leadership in value and affordability, and our efforts are working. In the U.S., we launched McValue at the start of the year, which drove immediate incrementality, and then we relaunched Extra Value Meals in September,” CEO Christopher Kempczinski said during the chain’s fourth-quarter earnings call.

Now, the chain has decided to leverage its loyalty promotion to offer members a meaningful discount on gas. That partnership could allow the franchise to grow its business without further lowering prices while driving customers to fill up at Shell stations.

How the McDonald’s gas deal works

Shell, which has more than 12,000 U.S. gas stations, according to ScrapeHero, shared the news of the partnership on its LinkedIn page.

“Eligible MyMcDonald’s Rewards members can redeem 1,500 points for 50¢/gal off at participating Shell stations. Running August 12 through September 12, this limited-time national offer is designed to attract new Shell Fuel Rewards members, drive site visits, and generate incremental gallons,” the company shared.

The offer, however, is only available to new Shell loyalty program members who enroll through the company’s app.

“This promotion brings together two iconic brands with a shared goal: delivering more value to customers while fueling growth for our business. We’re excited to welcome new customers to the Shell Fuel Rewards program and drive more members, more visits, and more gallons,” the gas giant added.

C-Store Dive sees this partnership as a smart way for Shell to add new customers.

“With consumer sentiment continuing to fall and plague convenience retailers, this promotion offers Shell a direct connection to McDonald’s nearly 210 million 90-day loyalty members, creating an opportunity for more sign-ups and repeat visits to its fuel pumps and c-stores,” the website reported.

McDonald’s has one of the largest loyalty programs in the world. Shutterstock

How the McDonald’s loyalty program works

To use the McDonald’s loyalty program, called MyMcDonald’s Rewards, you need to download the company’s app.

“Earning rewards points is very easy, simply download our app and agree to participate in MyMcDonald’s Rewards. Present the 4-digit code before ordering, or get points automatically when you order in the app,” the company shared on its rewards program FAQ page.

Earning and redeeming points is fairly simple once you do that.

“For every dollar you spend on eligible products, you will receive 100 points. You can start redeeming your MyMcDonald’s Rewards when you have 1500 points,” the company added.

The new gas offer, while it’s only a one-time-use program, could keep customers away from Costco, at least for one fillup.

“When low on gas, consumers choose gas stations based on cheap gas (56%), location (52%), ease of entering and exiting (37%), cleanliness (25%), and high-quality gas (25%),” according to a Bludot survey.

Gas prices top $4 a gallon nationwide

After a week where prices dropped, the national average for a gallon of regular gasoline is back on the rise.

“Today’s (August 13) national average is back up to $4.07 after dropping to $4 on Monday (August 10). Crude oil prices are once again in the $80 per barrel range amid continued uncertainty along the Strait of Hormuz,” according to AAA. 

Slowing sales were not enough to keep prices down.

“While gasoline demand is down, crude oil prices are keeping pump prices higher than normal for this time of year. So far, this is the highest August on record when it comes to the national gasoline average,” added AAA.

National average gas prices:

August 13 National Average: $4.07 

One Week Ago: $4.06 

One Month Ago: $3.87

One Year Ago: $3.15 

McDonald’s sees loyalty as a key sales driver

“In digital, we’ve built the industry’s largest customer platform with nearly 220 million active loyalty users, and we’re now among the largest loyalty programs in the world,” Kempczinski said during its second-quarter earnings call.

He talked extensively about the loyalty program driving increased visits during the Q2 2025 call.

“In the U.S. alone, on average, the same customer visits 10.5 times in the year before joining the loyalty program and then 26 times in the year after joining,” he said.

The Shell deal is not the first time the chain has offered rewards that go beyond its own menu.

“They are earning points in the app and using them to unlock exclusive deals. And thanks to our recent partnership in the U.S., customers were able to extend rewards to new experiences like the Snapchat+ subscription with premium features,” he added.

Related: Major retail meat company closes plants, lays off over 3,200

Block, the technology company founded by former Twitter CEO Jack Dorsey that owns Square, Cash App and the music streaming service Tidal, is open-sourcing Berd, a desktop application it originally built to give its own employees a single environment for working with AI agents across different models, tools and projects.Berd is a locally installed graphical desktop application rather than a browser-based workspace. It is available now on GitHub under a permissive Apache 2.0 license — meaning anyone can use, modify and redistribute it, including commercially — with free downloadable builds for macOS, Windows and Linux. The repository reached version 0.6.2 on Aug. 18, its seventh public release, and lists 91 contributors.“Berd is desktop-first because much of its value comes from working directly with projects, local files, tools, repositories, and agents running on or connected to the user’s computer,” Brad Axen, Head of AI Capabilities at Block, told VentureBeat in emailed responses.A daily work surface for AI agentsBlock positions Berd as a “daily AI work surface”: a single place where users can start chats, attach files or folders, choose agents and models, work inside persistent projects, configure AI providers, manage skills and extensions, review session history and build automations.The design goal is not merely convenience. According to Berd’s product specification, users should always be able to tell which project, files, agent, model, provider and session state are actively shaping a conversation. That emphasis on visible operational state is what Block says separates Berd from a generic chatbot wrapper. Configuration — providers, extensions, skills, automations, projects — is treated as part of the workflow rather than buried in an administrative layer, and the interface is designed to expose failures, unavailable providers, and loading and streaming states plainly rather than smoothing them over with assistant-style friendliness.Berd grew out of a practical problem inside Block. Employees were already working with capable agents — Block’s own Goose, Anthropic’s Claude Code and OpenAI’s Codex — but the experience around them had become fragmented.“We had capable agents through goose, Claude Code, and Codex, but working across them meant navigating different interfaces, configuration systems, and ways of managing context,” Lucinda Bell, Technology Communications at Block, said in an email. Berd, she said, gives Block teams “one consistent desktop application across models and harnesses.”For enterprises, that layer may matter as much as another incremental gain in model capability. Persistent projects mean users can return to an established collection of files, instructions and agent configurations instead of reconstructing context for each task. Block also explicitly designed Berd to make agentic work accessible beyond engineering, letting people start with a conversation and add tools, context and structure as the work requires.Block is not positioning Berd as a mobile product. Axen said the company’s open source collaboration platform Buzz already provides a mobile experience and is Block’s preferred environment for work on the go.Block thinks agents should look different when they are differentBerd takes an unusual approach to a problem most enterprise AI products leave implicit: how users tell one configured agent apart from another.Instead of treating every agent as another blank chat window, Berd gives agents roles, instructions, skills, tools and recognizable visual identities. Block created collections of animated characters — its flagship designs are called “Gloopies” — and Berd’s public site showcases predefined personas such as Berdy, Pushback, Choosey, Copycat, Tinker and Wildcard, each built around a different style of work: Pushback plays devil’s advocate on drafts, Choosey helps narrow decisions, Copycat learns to write in the user’s style.The visual layer is not meant to substitute personality for substance. Block’s own formulation, from a company blog post on Berd’s design, is functional: “The avatars make the agent recognizable. Its role, skills, and tools make it useful.”That distinction could grow more important as workers juggle several specialized agents at once. A recognizable identity serves as visible shorthand for an otherwise invisible package of instructions, tools and permissions — agents appear on the workspace as distinct animated characters rather than identical entries in a chat list, so users can tell them apart before opening a conversation.The product specification guards against the characters turning Berd into a novelty. It describes the intended personality as “focused, capable, companionable,” while explicitly rejecting toy-like presentation, oversized decorative UI and interfaces that hide operational state behind friendliness.An orchestration layer, not another model or agent runtimeArchitecturally, Berd is neither a new foundation model nor a new agent runtime. The application is built with Tauri 2 and React 19 — frameworks for building lightweight desktop apps with web technologies — and communicates with Goose, the open source agent framework Block introduced in January 2025, through a standard called the Agent Client Protocol (ACP). Goose runs quietly alongside Berd as a bundled “sidecar” process and handles the underlying agent loop; Berd handles projects, sessions, context, agents and configuration on top.For readers unfamiliar with it, Goose is a model-agnostic agent framework and runtime, originally focused on software development workflows. It connects large language models to files, commands and outside systems — including through the Model Context Protocol (MCP), an open standard for wiring AI systems to external tools and data — so agents can take actions rather than merely generate text. Goose works with proprietary model providers as well as locally run open models. Block has since contributed Goose to the Agentic AI Foundation, the Linux Foundation body it helped establish with Anthropic, OpenAI and others in December 2025, giving the framework a vendor-neutral home alongside MCP.Axen drew the boundary explicitly: agent harnesses such as Goose, Claude Code and Codex manage the loop between a model, its context and its tools, while Berd provides the consistent desktop environment around those harnesses. In the public release, users configure whichever providers and harnesses they want rather than inheriting a Block-selected model stack.The project’s README adds an operational detail that matters for anyone evaluating the software: Berd does not simply pull whatever Goose build happens to be current. Its build process pins a specific Goose backend version through a lockfile, verifies the cached binary matches that pinned version, and packages it with the app. Developers can explicitly substitute another Goose binary for local testing. For organizations inspecting Berd, that creates a defined version boundary between the desktop application and its default agent backend rather than letting the backend silently drift.How Berd compares to Codex, Claude Code, Cursor and GooseThat separation matters because the agent-tooling market is already converging on products that manage increasingly autonomous AI workers. OpenAI describes its Codex app as a “command center” for agents, with separate project threads and multiple agents operating in parallel. Claude Code, Anthropic’s agent for reading codebases, editing files and executing commands, now spans the terminal, IDEs and a desktop app. Cursor describes its latest release, Cursor 3, as a unified workspace for software development with agents, including handoffs between local and cloud agents and multi-repository workflows. And Goose itself remains available directly, as a free command-line tool and desktop app for technically comfortable users.Berd’s differentiation is therefore less about inventing the agent workspace than about making that workspace open source, multi-harness and potentially useful outside coding. Every other product in the table below is either tied to a single vendor’s models, priced as a subscription, or aimed squarely at developers — and in most cases all three.ProductMakerPricingModel supportDistinguishing featuresBerdBlockFree and open source (Apache 2.0); users pay only their own model provider costsAny provider the user configures, across multiple agent harnessesDesktop workspace over agent harnesses rather than a harness itself; local-first data storage; visual agent identities; persistent projects; enterprise custom distributions; designed for non-engineers as well as developersCodex appOpenAIIncluded with paid ChatGPT plans, from Plus at $20/month to Pro at $200/monthOpenAI models only“Command center” for coding agents; multiple agents running in parallel across separate project threads; cloud and local executionClaude CodeAnthropicIncluded with paid Claude plans, from Pro at $20/month to Max at $100–$200/month; also usage-based via APIAnthropic Claude models onlyDeep autonomous coding — reads codebases, edits files, runs commands; available in terminal, IDEs, desktop and web; usage caps have drawn developer criticismCursorAnyspherePro at $20/month; Ultra at $200/monthMultiple frontier modelsAI-native code editor; Cursor 3 adds a unified agent workspace with local-to-cloud agent handoffs and multi-repository workflows; developer-focusedGooseBlock / Agentic AI FoundationFree and open source (Apache 2.0); bring your own model, including free local modelsAny provider, or fully local models via tools like OllamaThe model-agnostic agent framework and runtime underneath Berd; CLI and desktop versions; can run entirely offline with no data leaving the machinePortability is part of Block’s pitch. Axen said much of the underlying work in Berd is deliberately kept portable: folders, Git repositories, files and plain-text instructions remain accessible outside the application; skills are file-based; and Berd agents can be exported and shared between Berd users with their instructions and settings intact. He pointed to ACP and MCP as part of Block’s direction toward “clear, user-owned formats and open protocols wherever possible.” That does not make every piece of state interchangeable across every harness, but it makes portability an explicit architectural goal rather than a side effect of using local files.Local-first data, telemetry and enterprise distributionsBerd is designed around a local-first data model. Axen said conversation history is stored on the user’s device in the local Goose session database, while credentials live in the operating system keychain by default.That does not mean all AI processing happens locally. “Local-first does not mean that no data ever leaves the device,” Axen said. When a user asks a configured model to work on particular context, the relevant prompt and material are sent to that model’s provider — so the actual data path, and the contractual terms governing it, depend on which provider an organization chooses.Telemetry — the usage data software sends back to its maker — is disabled by default in official Berd distributions. If a user opts in, Block says it collects a random installation identifier and a limited set of predefined usage events, while excluding prompts, messages, files, source code, credentials, names, email addresses and IP addresses from the telemetry payload itself (Block and its service providers still process IP addresses for transmission, security and rate limiting). Users can switch telemetry off again in settings. Block cautions that third-party forks can modify or redirect the telemetry system, so enterprises evaluating a customized build need to review that build’s actual behavior rather than assume Block’s defaults carry over.The public repository builds a self-contained, general-purpose distribution that does not depend on private package registries or enterprise credentials — technical teams can build and evaluate it without access to Block’s infrastructure. But Block has also created what it calls “distribution seams”: defined points where organizations can overlay managed provider settings, private agents, runtime configuration, optional companion tools, their own update channels and their own signing and publishing infrastructure, all without adding private material to the public source tree.Block’s own deployment shows how an enterprise can put governance around that architecture. Axen said Block runs its managed distribution on approved devices using approved model-provider infrastructure, managed configuration and governed tool connections. “The application can remain open and inspectable, while an organization decides which models, tools, permissions, and policies are appropriate for its environment,” he said.One caveat: “enterprise distribution” should not be read as a complete enterprise management plane. The README does not describe capabilities such as single sign-on, centrally enforced administrator policies, organization-wide permission controls, data-retention rules or audit logs. It establishes that companies can customize and distribute Berd; it does not by itself explain how those installations are centrally governed after deployment. Berd should be understood as endpoint software installed on users’ computers, not a centrally hosted SaaS product.Pricing, models and a closed contribution modelBlock has not announced a subscription price for Berd, and there is no indication one is coming. That does not make agent usage costless: organizations still pay whatever model, API or provider charges they incur. The economic pitch on Berd’s site is flexibility — users can match the model to the job rather than paying for an expensive frontier model on every task. As the site puts it: “Some tasks need the big brain. Most don’t.”Neither Block’s announcement nor the README provides a definitive matrix of supported model providers or recommends particular models. Internally, Block does not treat any one model as a permanent default: Axen said the company’s managed distribution exposes employees to “a curated mix of frontier and open-weight models,” with the catalog changing as models improve. “The best choice depends on the work,” he said.There is also a governance distinction developers should note. Although Berd’s source is public and Block solicits issues and feedback, the repository does not accept outside pull requests — external code contributions are automatically closed. The README says Berd is built by “a small team at Block” and directs outside participants toward well-formed, reproducible bug reports instead. Users are free to inspect, build and fork the code, but Block retains a tightly controlled upstream. For companies considering Berd as infrastructure rather than an experiment, that makes Block’s long-term maintenance and support plans worth clarifying.Block is not presenting Berd as a revenue product today. Axen said the first return on investment comes from Block’s own use of the software — making internal agent work more consistent while teaching the company how employees actually use agents — and that open-sourcing it adds leverage by letting outsiders inspect, adapt and give feedback on the architecture. A commercial layer around enterprise deployment may eventually follow. “Supporting organizations that want to deploy agents at work is an area of interest for us,” Axen said, adding that Block will share more when there is more to share.From private agent work to multiplayer workflowsBerd supplies the individual desktop experience; another Block project, Buzz, focuses on shared rooms where people and agents collaborate against a common, searchable record.Buzz is Block’s free, Apache 2.0-licensed collaboration workspace for humans and AI agents. Built on the decentralized Nostr protocol, it combines familiar team features — channels, threads, direct messages, voice and media sharing — with agents that hold their own cryptographic identities and defined permissions, and that can participate in conversations, code review and approved automated workflows. Unlike Berd’s local, single-user focus, Buzz is explicitly multiplayer, and can be self-hosted or used through Block-hosted infrastructure. It is likewise model- and harness-agnostic: teams can bring agents powered by Goose, Claude Code, Codex or others.Axen framed the split between the two products partly as a question of trust boundaries. “A local agent may have access to your filesystem and credentials,” he said. “An agent in a shared Buzz room has its own identity and access scoped through the relay.” Berd begins with one person and comparatively privileged access to that person’s machine; Buzz begins with multiple participants, explicit identities and shared-workspace permissions.The more consequential roadmap detail is that Block does not intend the two products to remain separate. “Going forward, our focus is Buzz, and we plan to bring the best parts of Berd’s single-player experience into it,” Axen said. Block wants work to begin privately and become collaborative without users switching tools or rebuilding context, with cross-device continuity — including mobile via Buzz — part of that ambition. In the meantime, Axen said, Berd remains available as an open source desktop application, but Buzz is the product Block encourages people to download and adapt.What enterprise adopters should take awayThe short version for IT and security teams: conversation history stays local by default; credentials use the operating system keychain; context sent to a model leaves the machine on whatever terms the configured provider offers; official telemetry is off by default and excludes prompts, files and source code; and organizations can package their own governed distribution — though central management tooling is not yet documented.The larger idea behind the release may matter more. Models and coding agents are becoming increasingly interchangeable components inside larger workflows. Berd is Block’s bet that the surrounding context — the projects, skills, tools, identities and persistent working environment — becomes the stable layer instead.

Higher education’s hidden admissions crisis is operational overload.

Smart decision-making means assessing options quickly and refusing to do anything that could capsize the ship — no matter how much you like the opportunity.

Inflation has reshaped where Americans shop, and nowhere is that shift more visible than in the off-price apparel aisle.

As grocery bills climb, gas prices creep back toward $4 a gallon, and everyday essentials eat up a bigger share of household budgets, shoppers who once bought without blinking are now comparison shopping for basics like never before.

That behavioral pivot has turned discount retailers into one of the most closely watched corners of the consumer landscape, with every earnings report and every foot-traffic reading treated as a referendum on how deep the squeeze on American wallets really is.

As a longtime fan of TJ Maxx, I’ve noticed during my visits that the stores are more crowded, even on weekdays, than they were five years ago. People need a way to save, and off-price giants are all chasing that value-hungry customer.

TJ Maxx, Marshalls, Ross, and dd’s DISCOUNTS have historically coexisted by carving out slightly different niches, but inflation has blurred those lines. Off-price as a whole has spent the past several years pulling traffic share away from traditional apparel retailers, with the entire off-price category comfortably outpacing it.

Now, new visit data from Placer.ai have revealed a new leader in the race.

The rise of Ross Dress for Less

Visits to Ross Dress for Less surged 16.4% year over year in the second quarter, while sister chain dd’s DISCOUNTS grew a robust 8.4%, Placer.ai noted. That momentum accelerated into summer, with year-over-year visit growth picking up further at both banners in June and July.

The traffic numbers lined up with the financials: Comparable-store sales at Ross climbed 17% in its most recent quarter, blowing past Wall Street’s guidance. That kind of growth in a single quarter isn’t just outperformance; it’s a signal that Ross is actively pulling shoppers from somewhere.

Related: Ross Stores CEO doubles down on change that may deter customers

The numbers tell us exactly where they’re being pulled from, in fact: TJ Maxx and Marshalls. The data show that foot traffic to TJ Maxx essentially hovered around last year’s levels, with same-store visits still down slightly heading into July, Retail Wire reported.

TJ Maxx still comfortably beat the traditional apparel retailers bleeding customers by mid-single digits. But for a chain that has ridden the value wave for years, Ross’ growth surge invites uncomfortable questions about what shoppers are finding there that is drawing them away from TJX brands.

Comparable-store sales at Ross grew 17% in its most recent quarter, exceeding Wall Street’s guidance.Jason Armond / Getty Images

Ross, TJX serve different customer needs

The retailers do target different customer bases. TJ Maxx and Marshalls have long skewed toward a higher-income customer shopping at somewhat higher price points, Retail Brew noted. This means someone who isn’t just hunting for the lowest tag but who also wants the thrill of the find: a designer handbag or a recognizable label tucked between the racks.

Social media users regularly post their finds on TikTok, leading followers to go on their own treasure hunt in hopes of finding similar deals.

More Retail:

Dollar General copies Costco’s playbook with a discount twist

Pepsi and Coca-Cola bet big on soda Americans say they want

Iconic supermarket chain closes more stores and facilities

That treasure-hunt positioning, along with the option to shop TJX’s stores online, has been a winning formula for the brand. But it also leaves the chains more exposed when household budgets tighten. Shoppers may rein in the more discretionary, nice-to-have side of their spending, or decide the drive to the store isn’t worth it with gas prices climbing.

Ross, by contrast, has built its identity around the shopper for whom price is the deciding factor above all else, brand recognition included, noted Felicia Mckenzie, Amazon senior product marketing manager, in a LinkedIn post. Neither Ross Dress for Less nor dd’s DISCOUNTS even operates a digital storefront, so every one of those visits represents a shopper choosing to show up in person for the deepest discount available.

In an inflationary environment, that stripped-down positioning is proving more resilient.

Still, both TJX brands showed real signs of life heading into back-to-school season, with Marshalls posting a modest visit increase in July and TJ Maxx’s same-store visits narrowing to just 0.6% below last year, a reminder that the treasure-hunt model hasn’t lost its pull entirely.

But as long as inflation keeps reshaping household budgets, the retailer offering the deepest, simplest discount looks best positioned to continue winning the foot-traffic race.

Related: TJ Maxx has a secret weapon Walmart and Target can’t match