0:09
Trump Drops Restrictions On Anthropic's Mythos and Fable Models
The Trump administration has lifted export restrictions that forced Anthropic to shut off public access to its Mythos and Fable models. After weeks of talks, Secretary of Commerce Howard Lutnick said Anthropic "has agreed to proactively detect and address security risks associated with the models; to work diligently with the U.S. government on protocols and standards and releases for Mythos, Fable and future models; and to inform the US government of any malicious activity." Access is set to begin returning July 1. TechCrunch reports: Anthropic had already publicly pledged to do much of this voluntarily, months before the export rule existed. That's part of why cybersecurity experts were skeptical of the restrictions in the first place. To them, the ban looked less like a security fix and more like leverage, a way for the Trump administration to punish Anthropic for its executives' public criticism of how the government, and the president's political opponents, might use the technology.
Mythos was originally made available to a select group of organizations beginning in April to allay concerns about its ability to identify and exploit vulnerabilities in software, while a version called Fable was released to the public in June with additional security guardrails. However, with Asian AI companies beginning to release their own AI models approaching Mythos-level capabilities -- among them Fugu and Tulonfeng -- the US government was under pressure to ease its restrictions on Anthropic to ensure that American AI could compete globally.
Last week, Lutnick cleared Mythos to be released to select customers approved by the White House. OpenAI's latest models were also released to a group of organizations approved by the Trump team, instead of the public. The Trump administration's erratic approach to AI policymaking has left companies across the industry with little clarity about what will govern future model releases. An executive order issued in June that signaled a desire to review models ahead of release was criticized by influential analysts like Dean W. Ball, who recently started a policy position at OpenAI.Read more of this story at Slashdot.
2:35
The Messy Middle - Pt 1
Paul Millerd walked away from the tidy world of McKinsey and BCG in 2017, swapping clear metrics and calendar blocks for the open‑ended chaos of freelance writing and consulting. The shift stripped away the scaffolding that once told him when to work and what counted as success, leaving a blank calendar, silent phone, and a lot of mental drag. He quickly realized the problem wasn’t a lack of discipline; it was hidden friction that pops up when the old structure disappears.
The piece maps six common sources of that friction—everything from unclear priorities to missing feedback loops—and pairs each with a simple fix. Those fixes stack into what they call the Messy Middle Operating System, a lightweight framework you can pull out whenever you feel stuck between the old role and the new identity.
Think of it as a set of habits that re‑introduce just enough structure to keep the momentum going without recreating the corporate cage you left behind. It’s less about forcing yourself to be ultra‑productive and more about nudging the invisible obstacles out of the way so you can keep moving forward.
3:51
AIEWF Daily Dispatch: Loops, Software Factories & Forward Deployed Engineers
Loops, loops and more loops. That word, loop, dominated conversations on day 2 of the AI Engineer World’s Fair — the first full day of keynotes and sessions. Perhaps knowing in advance what everyone would be talking about, AIEWF cofounder swyx titled his opening talk, “Loopcraft: The Art of Stacking Loops.”
swyx began by commenting on the evolution of AI engineering from 2022: from chat, to tools, to goals. “These days, we’re all about automations,” he added. “We’re all about cron jobs and loops.”
Allie Howe, a member of technical staff for Keycard, then introduced the main stage track for the day: Software Factories. She referenced Geoffrey Huntley’s influential article, “everything is a ralph loop,” a theory about turning an AI coding agent into a persistent worker by repeatedly restarting it against the same spec.
Pablo Castro from Microsoft then talked about Foundry, the company’s “AI app and agent factory.” He claimed that a “learning loop” occurs when people and agents work together.
OpenAI’s Alexander Embiricos and Romain Huet were next on, and they focused a lot on Codex, the company’s coding agent. One point they made was that using multiple agents via loops can result in enhanced productivity.
“There will be a lot of talk today about loops,” Embiricos said. “And if you can connect the agent to not only the work that you have to do, but why it has to be done, that’s how you can get the agent to start to begin much more work. And then if you can connect it to what you do afterwards, review and deploy, that’s how you help it land much more work.”
This segued to a presentation by Peter Steinberger, the “ClawFather” of OpenClaw, now working for OpenAI. He too was all-in on loops, noting that he designs loops to manage agents. He added that deciding what to pay attention to is his main challenge nowadays — and that the future is “better loops” to help solve this issue.
All this talk of looping led naturally to the concept of “software factories,” the subject of a presentation by Tereza Tížková from a company called Factory. She defined a software factory as “the whole loop, the whole lifecycle of developing software with autonomy.” She added that this doesn’t mean just coding, but also “collecting all the signals, reacting to user feedback [and] to logs, prioritizing what’s important, then orchestrating it all.”
Zach Lloyd from Warp also spoke about software factories; in fact, his thesis was that “software engineering will become factory engineering.” Loops in Lloyd’s framing were about improving the system.
In both Tížková and Lloyd’s talks, the emphasis was on having the agents doing the building for you. “You’ll be building the thing that builds the product,” was how Lloyd put it.
Afterwards, I went down to Warp’s booth in the AIEWF expo hall and spoke to Lloyd about software factories. I particularly wanted to know why Warp, which began as a CLI tool for developers, has pivoted into a ‘software factory’ platform where developers aren’t supposed to do coding anymore.
“The way to think of the factory is, like, pick your repos, pick the parts of the lifecycle that you want to automate, pick the ways in which you want humans to be brought into the loop,” Lloyd told me. “And different organizations [and] code bases will have different preferences for, like, do you fully automate code review [or] do you have humans do hard coding, stuff like that.”
I noted that the term “factory” might be offputting to many developers, since it implies mechanized rote work — much different from the creative era of coding we’ve just come from. Lloyd recognizes this is a challenge, but he argues software factories will become a new discipline of engineering — and that it still requires problem solving.
“For better or worse, the power of these systems is so great and the ability to accelerate is so strong that just writing stuff by hand...I don’t think it’s going to make sense for very much longer,” he said.
(For more from Zach Lloyd on software factories, stay tuned for a Latent Space interview to publish shortly.)
Related to loops and software factories, another theme from AIEWF today was the trendy new role of Forward Deployed Engineers. In an interview with Natalie Meurer, Head of Agent Engineering at Sierra, I established that FDEs are also sometimes called “agent engineers.” The main point is to help organizations adapt to agents, from a development perspective.
Meurer pointed out that a lot of the work of integrating AI into companies these days is in orchestrating agents.
“In practice, most customer-specific work takes place at the orchestration layer rather than in the models themselves,” she told me.
Cursor’s VP of Forward Deployed Engineering, Pauline Brunet, also ran a session today at AIEWF, in which she positioned FDE as part of the shift to software factories. “We partner with your organization to co-design and co-build your AI software factory,” she said.
9:14
From Sugar to Ethanol Fuel With a Little Microbial Help
In these trying times it seems appropriate to work through some ‘what if ‘ scenarios, such as the local gas station suddenly not having any more gasoline to sell you, or said gas station ceasing to exist altogether. In that case it can be incredibly useful to be able to create your own gasoline alternative in the form of ethanol. As demonstrated by [Hyperspace Pirate] in a recent video this process is fairly straightforward once you have procured an appropriate feedstock, such as here sugar (sucrose).
Although baker’s yeast (Saccaromyces cerevisiae) is more commonly associated with the production of ethanol-laced drinks, there’s nothing that says that you cannot distill out the approximately 10-15% ethanol that results from a yeast feeding frenzy and resulting waste products.
How to do this distillation step is explained in the video, with the mixture heated and put through a self-made reflux column to deal with the fact that the water/ethanol mixture is an azeotropic mixture, meaning that a lot of water is expected to make its way out of the condenser along with ethanol without this measure to condense as much of the water vapor before it can make its way to the top of the column.
Ultimately the conversion rate of plain white sugar to ethanol is about 54%, with the rest turning into CO2. With an appropriately converted combustion engine for running on 100% ethanol, it runs pretty well, though the final cost per liter of ethanol will heavily depend on your feedstock.
With the full costs of the electric heater of the distillation column taken into account – at 2.57 kWh/L – as well as the cost of the off-the-shelf sugar, [Hyperspace Pirate] with his Florida kWh cost of $0.12 paid around $2.62/L, or $9.91 per gallon. Even with how much prices at the gas pump have shot up recently, you’d pretty much need to find a free source of feedstock and otherwise optimize the process for it to make much sense, even in this economy.
That said, it’s crazy that the world of Mad Max doesn’t run on ethanol. If tomorrow a certain bubble were to implode and the global economy fell apart as a result, producing bioethanol would seem to be a highly marketable skill.
11:39
Why managing a family’s digital footprint requires a holistic strategy
It should come as no surprise, but modern Australian households have long evolved past the point of a ”family computer”. That relic of a bygone millennial era – the family’s sole access point to riding the information superhighway. Back when people used to say things like “surfing the web” and for that matter, the phrase “riding the information superhighway”. Today, your typical Australian family manages an entire ecosystem of connected infrastructure: smartphones, remote work laptops, school-issued tablets, gaming consoles, and an expanding list of smart home IoT (Internet of Things) devices. This can make creating watertight digital security for an entire household tricky. As these multi-device ecosystems scale, they introduce a wider attack surface for external threats while complicating internal privacy management.
This is why at PCMag, we recommend trying to consolidate your household’s digital security into a single suite as much as possible. Look for trusted security suites that offer specialised family plans. Something with parental controls, preferably a VPN, room to grow as new devices are added, and a tiered account system that allows administrator privileges. You want your entire family to have access to essential services such as antivirus, password managers, identity theft protection and encrypted backup. But you also need to be conscious of the dangers of a single access point holding the “keys to the kingdom” when it comes to your entire family's personal information. Having a single administrator avoids needing to leave passwords just laying around the house, and a high device-count limit means you don’t pay for countless different individual subscriptions for everyone.
On top of your standard personal devices, as more and more smart home technology integrates into local networks, securing the whole home requires more than just antivirus on your laptop. In a modern home, every single device connected to your Wi-Fi is a door. Bitdefender handles this by providing high-volume licensing via their Family Plans — covering up to 15 or 25 devices under a single subscription, depending on the selected tier. Dedicated local clients provide automated threat detection across Windows, macOS, iOS, and Android systems, preventing a compromise on a single mobile device from pivoting laterally across the home Wi-Fi network.
A common point of contention in shared household security plans is data overlap. When multiple family members share a single account configuration, personal data boundaries can become blurred. For this, we need security software that treats household members as distinct digital entities. Rather than forcing the family to share a single administrative profile or master password, each member should receive their own isolated digital space.
As an example, Bitdefender Family Plans generally handle this well. With a structural isolation ensures that a teenager's personal logs, search queries, and credentials remain private, utilising unique vaults within the Password Manager and separate Anti-Tracker configurations. Concurrently, parents can secure sensitive work emails, financial profiles, and banking credentials using the isolated, sandboxed Safepay browser environment without exposing administrative keys to the rest of the household.
The modern internet can be a hostile environment for a child, filled with social media pressures, cyberbullying, explicit content, and algorithmically optimised apps designed to maximise screen time. No parent has the time or the energy to hover over their child’s shoulder every minute they are using a screen. It ruins the relationship and is practically impossible when kids need devices for schoolwork. Instead, modern digital parenting requires a shift toward automation—setting healthy, invisible boundaries that run passively in the background.
For example, Bitdefender integrates a Parental Control module directly into its central management console, Bitdefender Central, allowing parents to configure distinct rulesets customised to each child's age group. The objective is to create a safe digital playground. You want a system that automatically filters out explicit web categories, enforces a hard bedtime by pausing internet access, and flags suspicious links or phishing scams before a child can click them, all while giving them the freedom to explore the web safely within those borders.
When you look at this massive list of requirements—protecting smart TVs, isolating work from play, giving teens their privacy, and keeping children safe—doing it manually on a device-by-device basis is a recipe for burnout.
16:43
Gemini Spark can now automate your Mac, even when you're not looking
So Google's Gemini Spark, which is their AI assistant that can do tasks in the background, is now available on Macs. What's interesting is that it can automate tasks on your computer, even when you're not actively using it. This means it can perform actions like organizing files or sending emails without you having to lift a finger. It's an extension of what Spark can already do on the web and mobile devices, but now it has access to your computer's capabilities.
This change is pretty significant because it allows Spark to integrate with your computer in a more seamless way. For example, it can use your Mac's keyboard and mouse to interact with apps, which opens up a lot of possibilities for automation. It's not just limited to simple tasks either, Spark can also use your computer's processing power to handle more complex jobs.
I'm curious to see how people will use this feature, because it has the potential to really change the way we work with our computers. Instead of having to manually perform repetitive tasks, Spark can take care of them in the background, freeing up time for more important things. It's a subtle but powerful change that could have a big impact on productivity.
It's also worth noting that this update is only available to AI Ultra subscribers, at least for now. It'll be interesting to see if Google expands access to this feature in the future, and how it will affect the way we use our computers.
18:23
AMD's new CPU core type just got unlocked on Linux before it was even announced
I’ve been poking around the Linux kernel lately and stumbled on something that feels like a backstage pass. A recent patch landed that adds a brand‑new AMD core classification—essentially a “low‑power efficient” mode—right before AMD even hinted at it publicly. The kernel now treats a subset of cores as if they’re built for background chores, dialing down voltage and clock speed automatically, which means laptops could squeeze out extra battery life without the user ever noticing.
What’s odd is how the change slipped in: the developers used a codename that matches AMD’s internal naming for their upcoming efficient cores, and the scheduling logic was already tuned to shift lightweight threads onto those cores. It’s the kind of detail you’d expect only the chip maker to know, yet it showed up in the open‑source tree weeks before any press release.
Because the patch is already merged, any distro that pulls the latest kernel will start recognizing those cores tomorrow. You won’t see a flashy UI option, but the OS will quietly reshuffle work, keeping the heavy lifting on the performance cores and letting the efficient ones nap when they can. It’s a subtle shift, but it hints at how tightly Linux and hardware vendors are syncing up these days.
Bottom line: if you’re on a recent Linux build with an AMD processor, you’re already benefiting from a feature that AMD hasn’t officially announced yet. It’s a neat reminder that the open‑source community sometimes gets a glimpse of the roadmap
20:07
New Florida Law Bans Local Net-Zero Emissions Policies
An anonymous reader quotes a report from Inside Climate News: A new state law limits Florida communities' aims to offset greenhouse gas emissions that are warming the global climate and intensifying disasters such as hurricanes. Specifically, HB 1217 prohibits local governments from pursuing net-zero emissions goals. At least 10 cities and counties have implemented such policies, including Fort Lauderdale, Miami, Orlando and Leon County, where Tallahassee, the state capital, is located. But the new law will not necessarily upend these policies, said Bradley Marshall, senior attorney at Earthjustice, an advocacy group. "It's certainly meant to scare municipalities and local governments from trying to do things to further net-zero policies," he said. "Now, its exact impact and what it exactly prohibits is probably up for some debate. Things that are adjacent to it -- emissions reductions and even climate change reduction policies -- on their face will not run afoul at all of a ban on adopting a net zero policy."
The measure requires local governments to submit an affidavit annually to the state Department of Revenue verifying compliance. Gov. Ron DeSantis, a Republican, signed the measure on April 22, Earth Day, and the law will take effect July 1. It states that "net zero policies, carbon taxes and assessments, and emission trading programs are detrimental to this state's energy security and economic interests and inconsistent with the energy policy and the environmental policy of this state." [...] HB 1217 also prevents local governments from purchasing items such as vehicles or appliances based on the fuels they use or production of the items. Local governments may not participate in carbon-trading programs or use public funds to support other organizations with net-zero policies. Cities and counties also may not charge a tax or fee tied with carbon emissions. "This bill is definitely part of a larger coordinated push by the political enablers of the fossil fuel industry to obstruct any tools -- legal or legislative tools -- to hold the industry accountable for its contributions to climate change," said Laura Peterson, senior analyst at the Union for Concerned Scientists, an advocacy group. "Florida is really on the front lines. So I imagine the governor is taking this step because he sees what's coming down the pike. It's not getting better. So I can only assume that this is an effort to satisfy some of the pressures that he's getting from donors and from his party to protect the industry. And he's doing it at the expense of his constituents."Read more of this story at Slashdot.
23:01
A new Linux driver gets us one step closer to daily driving the OS on an M3 MacBook
The Apple M3 MacBook has been pretty stubborn when it comes to accepting Linux on its hardware. The community has only just managed to boot Linux off an M3 MacBook thanks tothe Linux 7.2 kernel. However, that's currently as far as you can feasibly get, as the input hardware does not support Linux at all.
23:29
[AINews] Sonnet 5 today, and Fable 5 tomorrow
In separate announcements, Sonnet 5 was released today, and Fable/Mythos 5 were approved to be released again after some work with the government. The primary discussion around Sonnet 5’s efficiency was a damper on the excitement, driven by tokenizer changes and 3-6x more turn taking in benchmarks:
Our newest staff writer is reporting on the ground from AIE, and you can catch swyx and other keynote speakers on the stream today:
AI News for 6/29/2026-6/30/2026. We checked 12 subreddits, 544 Twitters and no further Discords. AINews’ website lets you search all past issues. As a reminder, AINews is now a section of Latent Space. You can opt in/out of email frequencies!
Anthropic launched Claude Sonnet 5 as its new default mid-tier frontier model, with immediate rollout across Claude, Claude Code, API, and ecosystem partners.
Anthropic officially announced Claude Sonnet 5 as “our most agentic Sonnet yet,” emphasizing planning, browser/terminal tool use, and autonomous execution that previously “required larger and more expensive models” (@claudeai)
Anthropic’s developer account said Sonnet 5 offers top-tier coding and tool-use performance at Sonnet pricing, with a 1M-token context window, and is the new default in Claude Code for Pro users and available on the Claude Platform including API and Managed Agents (@ClaudeDevs)
Anthropic kept the standard list price at $3/M input tokens and $15/M output tokens, but introduced a promotional rate of $2/M input and $10/M output through Aug. 31 / Sept. 1 depending on the post (@kimmonismus, @ClaudeDevs, @ArtificialAnlys)
Sonnet 5 surfaced first through leaks and client-side sightings: leakers claimed knowledge cutoff January 2026, $2/$10 promo pricing, and a 1M-context variant before launch (@kimmonismus); users then reported it appearing in the model selector, Claude Code 2.1.197, Anthropic GitHub, and finally going live in accounts including Germany (@kimmonismus, @scaling01, @scaling01, @kimmonismus)
Anthropic simultaneously expanded platform support around the launch: Claude Desktop on Linux (Ubuntu/Debian beta) with Claude Code/Cowork/chat on paid plans, though Computer Use was not included in that Linux release (@ClaudeDevs, @ClaudeDevs)
Anthropic also shipped Managed Agents updates—streaming session deltas, per-session overrides, webhook events, reverse pagination, credential injection scoping, and an observability tab with token/tool metrics—making the release as much platform/integration story as raw model story (@ClaudeDevs, @ClaudeDevs)
The launch was preceded by a large rumor cycle centered on Sonnet 5 + Fable 5.
Earlier app-string sleuthing suggested Anthropic was preparing to put “Fable 5” behind a separate usage-credit system billed outside existing plans, with identity verification language appearing nearby; that fed speculation that access would be gated and more regulated than existing plans (@kimmonismus)
This triggered concern that Sonnet 5 might launch as the widely accessible but weaker companion to a stronger, more restricted Fable 5, possibly with regional access issues, especially in Europe (@kimmonismus)
Additional rumor posts tied a potential Sonnet 5 release directly to a Fable 5 re-release, with some users explicitly saying they assumed Sonnet 5 would “at least” come with Fable news (@kimmonismus, @kimmonismus)
After launch, that expectation went unmet.
