FlashFeed

Full text

OpenAI reveals first branded hardware, the Codex Micro, a programmable macro pad built with keyboard maker Work Louder

Codex Micro seems to be based on Work Louder's Creator Micro 2's layout, mapped to Codex coding-agent shortcuts

The move reinforces OpenAI's Codex offering as one of its mainstay areas of focus by allowing developers the ability to perform tasks or interact with AI faster

OpenAI's first branded piece of hardware is not a long-anticipated consumer device it is building with ex-Apple design chief Jony Ive , but rather a programmable macro pad called the Codex Micro.

The keyboard, which consists entirely of macro keys designed to "supercharge people's Codex usage," according to an OpenAI spokesperson at the AI Engineer World's Fair, is reportedly a collaboration between the iPhone creator and the custom macro pad creator Work Louder.

With OpenAI's developer-centric account on X indicating that the full launch of its hardware foray is expected on July 15, the AI giant seems to be pulling out all the stops to ensure it becomes a well-received add-on for the developer community.

A simple rebadge or a sign of things to come?

The as-yet-pending release 'Codex Micro' seems to be inspired by Work Louder's existing Creator Micro 2, a compact macro pad that offers thirteen mechanical keys, a joystick, a rotary encoder, and touch controls, arranged across programmable layers to power users needing faster or more fine-grained control over AI-assisted coding tasks.

The move is understandable for OpenAI in terms of both securing a victory with developers and brand recognition, and essentially testing the waters on how it would handle a hardware launch for the company's upcoming AI device for more general-purpose users.

It can also, to a degree, be seen as OpenAI essentially acknowledging that its earlier stance of narrowing its focus to 'nail' its core business might be one the company is willing to make exceptions to, especially when it comes to coding tools or enterprise use-case hardware.

OpenAI's CEO of applications, Fidji Simo, reportedly told staff that the company was looking to deprioritize areas outside its core focus to allow it to lead where it mattered.

In 2025, OpenAI shipped the Sora video app, the Atlas browser, ecommerce features inside ChatGPT, advertising work, and hardware efforts, a "series of startups" approach that insiders said had produced organizational confusion and constant reshuffling of scarce compute, distracting it from a truly centralized goal.

Hardware, in other words, was explicitly on the list of distractions. A physical keyboard is arguably as clear a violation of that directive as one could possibly design.

(Image credit: Shutterstock / Mehaniq) OpenAI is also reeling from a smaller-than-expected gap from competitor Anthropic and its Claude models in the areas where its GPT models do compete. This can perhaps be attributed to Anthropic's much narrower focus, which caters specifically to coders and enterprise through its Claude Code and Claude Cowork offerings.

One can argue that OpenAI's move isn't one that distracts it from its core focus, but rather complements it, even as R&D and integration for the most part is something that Work Louder will commit to.

It allows the AI juggernaut gets to test out both the marketability of an OpenAI-branded hardware product and appease developers and founders with a low-effort play even as they have increasingly been considering tools from Anthropic and Google as well as other AI solutions providers.

None of OpenAI's previous concerns may apply here; the exercise does not consume compute, it caters to a key audience for OpenAI, with Codex assisting 5 million weekly users as of June, and it does not meaningfully engage an engineering team as some of its other projects do.

With OpenAI and Anthropic slated to IPO soon, both are locked in a race to secure as many active users as possible to justify their valuations, even as they vie to build the most powerful models to cater to various industries, including defense, telecommunications, pharmaceuticals, and software development, to name a few.

OpenAI's move might just be a sign of things to come, as it leverages ChatGPT's massive brand recognition to develop marketable, revenue-generating solutions such as a custom macro keyboard, even as it is loath to spend any of its engineering or compute resources on anything but the most important of its tasks, even as enthusiasts continue to wait for the release its upcoming collaboration with legendary Apple designer, Jony Ive.

Full text

Presented by Red Hat

At VentureBeat's recent AI Impact event, where the discussion centered on what separates enterprises that scale agentic AI from those that stall in pilot mode, Brian Gracely, senior director of portfolio strategy at Red Hat, detailed what companies actually run into once agents reach production.

He dove into cost discipline, the security blind spots unique to autonomous systems, and the organizational friction that determines whether agent adoption spreads beyond early champions.

Enterprises are overestimating how far behind they are on AI agents

Many enterprise leaders, especially those following industry keynotes and AI announcements, worry that they’re already falling dangerously behind competitors deploying agents at scale. But according to Gracely, much of that anxiety reflects a misconception about how quickly organizations learn once they begin building. Teams often move up the learning curve far faster than they expect.

That rapid progress creates a different challenge, however. As agent usage expands, AI costs rise just as quickly, turning cost management from an engineering concern into a recurring boardroom discussion.

Agentic AI usage is orders of magnitude higher than during the chatbot era, making AI costs a growing concern for enterprises. At the same time, organizations are becoming increasingly aware of their dependence on a small number of model providers. According to Gracely, that combination is driving many enterprises to explore alternatives that give them greater control over costs and infrastructure.

"The two or three top providers are already telling the market that they're losing money, and they're trying to go public to make up those gaps," he explained. "At some point, the dependency on that means you're either going to buy at a very high-cost level, or you're going to figure out alternatives to control what you're doing."

Right-sizing AI models is the fastest lever for cutting agent costs

The biggest cost issue is that enterprises overspend by defaulting to the most capable model available regardless of task complexity.

"If I'm simply trying to resolve an insurance claim, I don't need to know about the history of Western civilization in my model, I don't need to know World Cup soccer scores," Gracely said.

Semantic routing is the mechanism many companies use to make that judgment automatically, classifying requests and sending each to a model sized for the task without requiring users to choose, while infrastructure techniques like caching repetitive queries cut how often a request needs to reach GPU compute at all. Together, he said, these tools remove the assumption that efficiency and innovation pull in opposite directions.

"There's a lot you can do at a GPU infrastructure level, and quite a bit you can do in terms of flexibility of models," he explained. "Those give excellent choices in terms of the levers you're trying to pull, whether you need efficiency or you need innovation. That shouldn't be a binary choice."

The financial discipline needed for token spend is similar to the FinOps practices that took years to mature in order to take control of cloud compute spending. Those underlying frameworks will transfer even as the vocabulary changes, Gracely said, especially as organizations push for internal education on model selection so teams stop defaulting to the most prominent option for tasks that don't need it.

"The same way we first had to teach the financial people what an EC2 instance is and what an S3 bucket is, you're going to have to start explaining tokens to them," he said. "We don't always need a Rolls-Royce. We don't always need caviar, because we're trying to do basic types of things."

Patch speed is now critical as AI tools find vulnerabilities faster

AI-powered vulnerability discovery is forcing enterprises to rethink how quickly they can identify, validate and deploy patches. Long-established patch management cycles may no longer be fast enough in an environment where AI can uncover — and attackers can exploit — new vulnerabilities much more quickly.

"Most companies are probably going to have a window of somewhere between seven and 14 days to stay ahead," he said. "There are groups, Red Hat included, that are going to build patches for these, but the embargo window is going to be short."

AI is also changing what defenders need to look for. Rather than simply uncovering isolated critical flaws, AI security tools can identify combinations of seemingly minor vulnerabilities that become dangerous only when chained together. As both software complexity and vulnerability discovery accelerate, Gracely argued that the ability to rapidly manage and update software is becoming a strategic capability rather than simply an operational one.

Subject matter experts and compliance teams decide whether agents scale

In the end, organizational adoption comes down to the need for deep, sustained involvement from the subject matter experts whose knowledge the agent is meant to encode, which makes earning their buy-in a prerequisite rather than an afterthought.

"You have to think about the incentives, what you do for people who participate in this work so they don't feel threatened that it's going to take away their job, and how you incentivize people in the long run to cooperate with that innovation," he said.

Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com .

Full text

A team of conservationists has discovered and restored the tombstone of Sebastian, a formerly enslaved man who died free in 1729.

"That discovery is likely one of the oldest gravestones of a free Black person in America," Michelle Wu, the mayor of Boston, said in a July 4 speech . "It's been there all along. We just had to go look and share the story."

The stone marker is inscribed with Sebastian's chosen name, Boston, and features an image common in 17th- and early 18th-century tombstones in New England: the death's head . The symbol includes a stylized skull flanked by wings, which were possibly intended to symbolize spiritual resurrection .

On the tombstone, a simple, five-line inscription reads:

Original Inscription

HERE LYES Y e

BODY OF BOSTON

AGED 70 YEARS

DEC D FEB Y Y e 28

1728

Modern English

HERE LIES THE

BODY OF BOSTON

AGED 70 YEARS

DECEASED FEBRUARY THE 28

1728

In 18th-century American writing, the word "the" was often abbreviated with a Y-shaped letter form called a " thorn " representing the "th" sound. Tombstones also frequently abbreviated "deceased" and the month. Somewhat confusingly, before 1752, the American colonies regularly used the Julian calendar , as did Britain, in which the new year started on March 25. Boston's year of death is listed on the tombstone as 1728, but his February 28 death was actually in 1729 in our modern Gregorian calendar.

Kelly Thomas, the director of the Historic Burying Grounds Initiative for the Boston Parks and Recreation Department, told WBZ News Radio that she identified the gravestone during a restoration project at the Granary Burying Ground , the city's third-oldest cemetery. Established in 1660, the cemetery includes over 5,000 graves, including the final resting places of Declaration of Independence signers Samuel Adams and John Hancock; the U.S. Revolutionary War officer Paul Revere; and Crispus Attucks, a sailor of mixed African and Indigenous ancestry who is thought to be the first American killed in the Boston Massacre.

"I was reviewing the photos of headstones, and then I noticed that the stone only had one name," rather than a first and last name, Thomas told WBZ News Radio, suggesting that Boston was either enslaved or free when he died. Based on a search of historical records, Thomas determined that Boston's enslaved name was Sebastian and that he also went by Bastian.

RELATED STORIES

'It's a huge deal': Archaeologists discover second cannonball from the Battle of the Alamo, and it was likely fired by Texans

8-year-old African American boy from Colonial Maryland found buried with white Colonists, and it's unclear if he was enslaved

Lost Indigenous settlements described by Jamestown colonist John Smith finally found

A surprising amount of information was recorded about Boston and his wife, Jane Lake. In 1701, they had their daughter, also named Jane, baptized at First Church in Boston, where they regularly attended worship services. But while the two were married with children, they were also both enslaved and lived in different households, historian Gloria Whiting wrote in a 2016 study .

Boston was likely freed in the early 18th century, some time after the 1702 death of John Waite, who had held him in slavery. By 1708, a list of free Black people included the name Boston, and he had gained a reputation in the city as a hardworking handyman, according to Whiting's study. He had been freed for roughly 30 years when he died in 1729 and was well-known enough to have an obituary published for him in the New-England Weekly Journal , a rare occurrence for enslaved or freed Black people in colonial America.

Wu also said that, in recent weeks, archaeologists in Boston had discovered musket balls and gun flints from the infamous Battle of Bunker Hill that took place on June 17, 1775. The battle was fought by British Crown forces and New England troops in the first stage of the American Revolutionary War.

See if you know what these mysterious artifacts are with our archaeological fragments quiz!

Full text

Bitcoin Magazine

Polymarket Turns On Instant Bitcoin Deposits Via Lightning Network, Powered by Spark

Polymarket, the crypto-native prediction market, has begun supporting instant Bitcoin deposits over the Lightning Network . The feature uses infrastructure from Spark , a Bitcoin protocol built for payments and stablecoins.

In a post on X, Spark told users they can deposit BTC to the platform with more speed and more privacy than the older method offered.

The move extends a funding push that started in October 2025, when Polymarket switched on standard on-chain Bitcoin deposits. Those deposits carried a wait: most on-chain Bitcoin transactions need three to six confirmations, a window of 10 to 60 minutes, before a platform credits an account.

The on-chain route carried a higher minimum deposit, a reflection of bridging costs. For a trader who wants a position on a live market, both the delay and the fee are a cost.

Lightning and Spark close the gap. Spark validates a Bitcoin transaction at the moment it broadcasts, checking for double-spend risk, fee adequacy, and replace-by-fee flags.

JUST IN: The world's largest predictions market Polymarket now accepts Bitcoin Lightning deposits! pic.twitter.com/CxOObnbyJ2

— Bitcoin Magazine (@BitcoinMagazine) July 7, 2026

The protocol credits the deposit in under a second and absorbs the confirmation risk, a design Spark markets as zero-conf.

Polymarket does not have to manage confirmation thresholds or run its own Lightning nodes; a single Spark SDK handles on-chain, Lightning, and stablecoin rails.

Spark keeps deposits self-custodial. Each wallet ties to the user’s own keys, so the protocol, not Polymarket, carries the operational load, and users retain control of funds until a trade.

Spark counts wallet providers such as Breez, Xverse, and Cake among the teams building on the same rails, and Tether chief Paolo Ardoino has praised the protocol as a route to programmable Bitcoin over Lightning.

Polymarket’s boom over the years

Timing matters for a company in a growth phase.

Founded in 2020, Polymarket rose to prominence during the 2024 U.S. presidential election and has added Chainlink oracles, earnings markets, and a fresh contest with regulated rival Kalshi.

Faster, cheaper funding lowers the barrier for the Bitcoin holders who make up a large share of the crypto audience, and it hands Polymarket a fresh answer to a rival that has pressed it on volume.

This post Polymarket Turns On Instant Bitcoin Deposits Via Lightning Network, Powered by Spark first appeared on Bitcoin Magazine and is written by Micah Zimmerman .