People buy video walkie-talkies to facilitate short-distance communication among children and / or family members. Aaron Christophel recognized an opportunity to bring DOOM onto one of these low-cost gadgets. These devices have small color screens, built-in cameras, microphones, speakers, and rechargeable batteries. Models retail online for between fifteen and twenty euros and rely on the TXW818 system on a chip to function. This chip has processing capacity comparable to some wireless modules and supports external memory, as well as four megabytes of PSRAM.
Christophel began by closely inspecting the electronics, noting that the devices differed: some had two megabytes of external flash memory, while others had four. Stuffing chip markings and scrambling flash memory is a typical practice that manufacturers use to make it difficult to meddle with their goods. Not to mention the specific tools required to delve deep into things: a USB to UART converter, a Blue Pill board posing as a J-Link clone, and a box called a profiler that allows you to monitor power consumption as devices perform different things.
Extended Jobsite Range: 2W output reaches up to 300,000 sq ft or 25 floors under optimal conditions, keeping large, dispersed crews reliably connected
IP67 Weatherproof Build: Rated for full water and dust immersion with anti-slip housing that survives drops up to 2 meters, built for harsh outdoor…
Secure Team Channels: 22 FRS channels and 121 privacy codes eliminate interference, with auto-squelch and up to 18 hours of battery life per charge
The majority of the project included reverse engineering. Christophel dumped the original firmware and examined it with Ghidra, an app that is quite useful for this type of task. The software development kit that came with the device did not include drivers for the screen or camera, so Christophel had to create his own. Things only got tougher from there, since the stock firmware disabled the debug interface immediately after booting up. Christophel needed to find a means to keep it on, so he devised a mechanism that allows him to maintain access to the debug capabilities by connecting specific capacitors and communicating with the flash chip at startup.
With that resolved, Christophel began work on a new firmware, which did a variety of cool things such as initialize the screen and make the buttons functional. He also created a function that allows the device to recognize the amount of flash memory it has on board and alter its settings accordingly. Things got a lot more complicated when he attempted to port the game DOOM over. The original game data file alone takes up approximately one and a quarter megabytes, which was a problem for devices with only two megabytes of flash. Christophel had to reduce the file size to five hundred kilobytes. When the device boots, the firmware unpacks this data into PSRAM, which is exactly what is required to run the game.
The gameplay is rather straightforward, with you using the walkie-talkie buttons to move ahead and backward, as well as to turn around. When things begin rolling, an entertaining visual effect appears in the center of the screen, which is simply the front camera feed with the player’s face staring back at themself. Given the limited hardware, the performance is adequate, and the little screen does an excellent job of displaying the action, allowing you to have some fun with the device. Christophel has made the whole source code available on GitHub, so if you’re intrigued, you can now build the binaries and try it for yourself. [Source]
The 2026 World Cup is rolling out two layers of technology that most of its 10 million visitors will actually touch: a consumer-AI layer led by Google, and a biometric-identity layer that turns a fan’s face into a ticket. This is the quieter half of the tournament’s tech, the half aimed at fans rather than threats.
Across 16 host cities in the US, Canada, and Mexico, neither layer comes with a robot dog attached. Both, though, are likely to outlast the final.
Google’s Gemini goes to the World Cup
Google has made Gemini and its Pixel phones official sponsors of several national teams, among them France, Argentina, Morocco, Iraq, Turkey, and the United States. Pixel is the official phone of the French squad, which is also using Gemini for team communications.
For fans, Google is pushing tournament features across Search, Maps, Waze, and the Gemini app: live score tracking, AI-generated tactical diagrams, and match highlights the app assembles on demand. It is also making its AI Mode Pro visuals free over the summer, timed to the tournament. For Google, the World Cup is a global launchpad for Gemini, dressed as fan service.
At the gate, the bigger change is biometric. At Gillette Stadium near Boston, fans can opt into facial recognition that links their face to a digital wallet, so they enter and pay without a ticket or card. Several venues are testing similar face-based entry.
Advertisement
Around the stadiums, cities are wiring up wider surveillance. In Seattle, officials connected stadium-district CCTV and automatic licence-plate readers to a Real-Time Crime Center, after a public fight over when the cameras switch on and whether they would track immigration status. None of this is unprecedented: Qatar ran the 2022 World Cup with around 22,000 cameras across eight venues. The new element is the consumer-facing pitch, that handing over your face is simply faster.
This biometric layer sits alongside the more visible security hardware TNW has already covered, the robot dogs, hunter drones, and AI cameras, but it is the part fans will queue up and opt into themselves. And the core technology is fallible: facial recognition is something independent studies have shown misidentifies women and people of colour more often than white men, and which TNW has long flagged as a civil-liberties risk once it scales.
Even the referee is now a camera
The AI reaches the pitch too. FIFA’s body-worn ‘Ref Cam’, trialled at the 2025 Club World Cup, is now written into the Laws of the Game and will be available in every match, with selected moments fed to broadcasters and stadium screens. FIFA’s partner Lenovo is using AI to clean up the footage, claiming up to 50 per cent less motion blur from a sprinting referee.
The pitch is transparency. The effect is one more live, AI-processed feed in the broadcast.
Advertisement
The bill falls on the fans
More than 120 civil-society groups, including the ACLU and Amnesty International, have issued a travel advisory for the tournament. They warn of racial profiling, device searches, social-media screening, and facial recognition, and advise some travellers to remove face-unlock from their phones before flying.
In February, ICE said its agents would play a ‘key part’ in tournament security.
The face-payment systems are, for now, opt-in. The question the tournament leaves open is what happens on 20 July, the day after the final. Stadium facial recognition, licence-plate networks, and AI video analytics rarely disappear when the crowds do.
The World Cup is where the softer half of this infrastructure gets normalised, in front of 10 million people, as the price of getting through the gate.
This article is crossposted from IEEE Spectrum’s careers newsletter. Sign up now to get insider tips, expert advice, and practical strategies, written in partnership with tech career development company Parsity and delivered to your inbox for free!
There is no shortage of people telling recent engineering graduates that their degree was a mistake and that AI is coming for their jobs before they even land one. I respectfully disagree.
I have been a software engineer for 12 years, done well over 100 interviews on both sides of the table, and run Parsity, an AI engineering program. A few patterns emerge consistently in who actually breaks through in today’s job market. Here’s why I think the job market isn’t as dire as it looks, and what I would do if I were looking for my first tech job.
The Numbers Need Context
The Federal Reserve Bank of New York recently placed unemployment for recent CS graduates in the United States at 6.1 percent, with computer engineering graduates at 7.5 percent. Compared to philosophy majors at 3.2 percent and art history graduates at 3.0 percent, those figures look alarming. They require more context than most headlines provide.
Advertisement
When researchers factor in underemployment (graduates working jobs that don’t require a college degree), then engineers are doing relatively well, coming in below 20 percent, against a 42 percent average across all recent graduates. Many majors reporting lower unemployment are achieving that figure by accepting work entirely unrelated to their field. Scored across unemployment, underemployment, and early-career earnings together, CS and computer engineering still rank among the top fields for overall labor market outcomes.
The degree is not the problem. The hiring pipeline is. Job postings labeled “entry-level software engineer” grew roughly 47 percent between late 2023 and late 2024, while actual hiring into those roles dropped approximately 73 percent in the same window. So-called “ghost jobs,” used to create an illusion of company growth, are everywhere. This makes the front door harder to find, but it exists.
Here Is What To Do About It
Do a broad search of your (real-life) network. Roughly 26 percent of job offers come through referrals. Look at your actual network—classmates, professors, past internship contacts, relatives—and identify people at companies that might be hiring. The goal is a warm introduction to someone who is or knows a decision maker. One introduction carries more weight than a hundred cold applications through a portal.
Find symmetric risk. A junior engineer is a risky hire by definition. A startup carries a matching risk profile, meaning potentially lower compensation, no certainty of longevity, and higher performance expectations. But that shared risk creates mutual interest. The learning curve is steep, the exposure is broad, and the track record transfers directly. For engineers whose longer-term goal is a large organization, a startup is not a detour. It can be how you build the experience those organizations eventually want to see. The first job is for validation and learning. It is not a life sentence.
Advertisement
Manufacture experience rather than waiting for it. Employers want experience but will not hire you to get it. The way through is to create it: a deployed project, an open-source contribution, building something real for a small business or family member. Recruiters are skeptical of toy projects. A deployed application solving a real problem, combined with the ability to talk clearly about the decisions you made and why, still moves the needle.
Gain practical AI engineering skills, not just AI tool fluency. Using Cursor or Copilot is now a baseline expectation. What differentiates candidates is going one level deeper. Most working engineers, including senior ones, have not built a RAG pipeline or designed a multi-agent system. Understanding how to chunk documents, generate embeddings, store and query them from a vector database, and wire it into a production application puts a candidate ahead of a significant portion of the market on a skill in rapidly growing demand. AI and data science roles grew 163 percent in job postings in 2025. The engineers who understand how these systems actually work, not just how to prompt them, are in the shortest supply.
Stop optimizing around conditions you cannot predict. Nobody anticipated the 2021 hiring boom. Nobody predicted this correction. Build durable skills. The demand for engineers who can reason clearly about systems is not going away. Where you start is not where you end.
—Brian
Advertisement
More major workforce reductions are on the horizon at Big Tech companies: Meta announced it will cut 10 percent of its workforce, or about 8,000 employees, and Microsoft plans to offer buyouts for 7 percent of its U.S. employees in a voluntary retirement program. The cuts are understood by many to be linked to AI. But is AI really to blame? For The Conversation, two academics at the University of Sydney give their two cents.
Tom Burick got his start as a roboticist. But when a financial downturn forced him to close his robotics business, he thought of the effect teachers had on his life and decided to pay it forward. Burick now works as a technology instructor at a school for students with autism, where he recently led a project building a full-scale replica of ENIAC, an historic computer celebrating its 80th anniversary this year.
Across several industries, the United States has been moving toward limiting the use of sensitive technology made in China. Now, legislation has been introduced to extend the trend to ground robots, including humanoids, dogs, and crawlers. This could benefit some U.S.-based robotics firms—but many of these companies still rely on Chinese-made components. “The U.S. robotics industry is in a pickle,” writes Spectrum tech policy editor Lucas Laursen.
Legora is already hiring to fill its new offices, with plans to grow its combined EMEA headcount to 700.
Swedish legal AI start-up Legora is expanding its footprint with a new engineering hub in London alongside new offices in Madrid, Milan and Paris. The expansion comes just months after the company raised $550m in a Series D round which valued Legora at $5.5bn.
Founded as Leya in 2023, Legora is an agentic AI platform supporting legal professionals with research, review and document drafting. It is used by more than 100,000 legal professionals at more than 1,200 law firms and in-house legal teams across more than 50 markets, according to the company.
The Madrid, Milan, and Paris offices will serve as regional hubs for customer success, go to market functions and legal engineering, while the new London engineering hub will be co-located with Legora’s existing presence in the city.
Advertisement
The new offices will open during Q3 this year and represent Legora’s most concentrated Europe, Middle East and Africa (EMEA) investment to date, the company said.
Legora is already hiring for positions to fill the new offices, with plans to grow its combined EMEA headcount to 700 within the next year. According to its website, Legora currently employs more than 400.
“Our customers in these countries have built Legora into the way they work,” said Max Junestrand, the CEO and co-founder of Legora. “Opening offices in Madrid, Milan and Paris means we can be genuinely close to them as we build the future of the platform together.
“Engineers who understand how AI applies in professional contexts are disproportionately concentrated in London,” said Junestrand added. “People here have built things that have to perform under real legal and regulatory constraints. That’s a different problem from building a consumer product, and it’s precisely the problem we’re solving.”
Advertisement
The new offices will bring Legora’s global footprint to 16 cities including Munich, Chicago, Houston, San Francisco, Toronto, Bengaluru and Sydney, as well as the recently announced offices in Singapore and Tokyo, the company said. Legora also has existing engineering hubs in Stockholm and New York.
Legora’s March Series D round was led by Accel, with participation from the likes of Benchmark, general Catalyst, Y Combinator, Menlo Ventures and Salesforce Ventures – taking the legal-tech’s total raise to date to $815m.
The company, at the time, said it planned to use the newly raised funds to further expand across the US, including with new offices in Texas and Illinois, as well as new local hubs. Legora plans to expand its US headcount to more than 300 by the end of 2026.
Engineering teams building agentic coding pipelines now have a concrete open-source alternative to managed models like Claude Fable 5 — one that runs on a single H100. The tradeoff: Cohere’s North Mini Code, which launched Tuesday, generated three times the output tokens of comparable models in independent testing, a verbosity cost that compounds in high-volume production workloads.
The new open-source model is a 30 billion parameter mixture-of-experts (MoE) model with 3 billion parameters active per token, built for agentic software engineering including sub-agent orchestration, architecture mapping, code review and terminal work. The model supports a 256,000 token context window with a 64,000 token maximum generation length, and is available on Hugging Face under an Apache 2.0 license.
What North Mini Code can do
North Mini Code targets the full agentic coding stack. Here is what the model does and what it runs on.
Software engineering. Cohere built North Mini Code specifically for agentic software engineering, not adapted from a general-purpose base. It has integrated tool-use capabilities and supports interleaved thinking, which Cohere says improves performance across multi-step agentic work.
Advertisement
Architecture mapping and code review. North Mini Code can analyze and map systems architecture, surface dependencies and perform code review across large codebases. With a 256,000 token context window, it can hold substantial multi-file projects in a single context pass.
Terminal-based agentic tasks. The model is trained for terminal environments, handling shell interactions, package scripts and command-line tooling. Cohere benchmarked it on Terminal-Bench v2, which tests agents in real terminal environments rather than synthetic code generation tasks.
How it was built
North Mini Code is a sparse mixture-of-experts model with 128 experts, of which 8 activate per token. The compute requirement at inference time is closer to a 3 billion parameter model despite 30 billion total parameters. Nick Frosst, co-founder of Cohere, demoed it running on a Mac Studio via MLX at around 20 gigabytes of RAM, the same machine he uses for his own local coding work.
Cohere trained the model through two stages of supervised fine-tuning followed by reinforcement learning with verifiable rewards across more than 70,000 verifiable tasks spanning approximately 5,000 repositories, deduplicated against SWE-Bench.
Advertisement
Rather than optimizing against a single agent scaffold, Cohere trained across three. SWE-Agent uses a rich CLI with specialized commands. Mini-SWE-Agent uses a single bash tool with raw shell output. OpenCode uses individually typed tools returning structured JSON. Cohere reports a 10 percentage point gain on OpenCode evaluation from the multi-harness approach while maintaining SWE-Agent performance.
Where it fits
North Mini Code enters a market that now includes Mistral Devstral Small 2, GitHub Copilot, Cursor, and Claude Fable 5 — each with distinct cost and deployment tradeoffs.
Cohere’s primary benchmark comparison is against Mistral Devstral Small 2, a 24 billion parameter dense model. In vendor-reported internal tests, Cohere claims 2.8x higher output throughput and a 30% inter-token latency advantage over Devstral Small 2 in internal tests under identical hardware configurations. Cohere also claims, in its Hugging Face technical post, that North Mini Code outperforms open-source models up to four times its parameter count on its reported benchmarks, including models at 120 billion parameters.
Artificial Analysis independently ranks it eighth of 127 comparable open-weight models on output speed at 210 tokens per second, with a time to first token of 0.25 second against a class median of 1.95 seconds. It places 18th of 127 on the Artificial Analysis Intelligence Index. One flag from the same data: the model generated 75 million output tokens to complete the Intelligence Index against a class median of 25 million. In high-volume agentic pipelines, that verbosity compounds into inference cost and latency.
Advertisement
“Suddenly people are thinking like hey, am I getting enough economic value out of the tokens from a model?” Frosst said during the launch video. “Local deployment is one way of empowering people and making AI really something that works for them.”
GitHub Copilot, Cursor and Claude Code operate on per-usage or subscription pricing with no on-premises option. Anthropic’s Claude Fable 5, now the most capable publicly available managed coding model, runs at $50 per million output tokens. For Frosst, the model is the polar opposite of Fable.
“Its small, cost effective, apache 2.0, and locally deployable. This is the way LLMs should go. small, open source, transparent and sovereign, vs large, expensive, proprietary and hegemonic,” Frosst wrote in a post on X.
What this means for enterprises
For teams building production agentic coding pipelines, North Mini Code’s release clarifies a set of decisions that have been forming for months.
Advertisement
Purpose-built agentic training is now a baseline to evaluate against. The distinction between models fine-tuned for code and models trained specifically for agentic workflows, with verified tool calls and multi-harness robustness, is now a material factor in pipeline decisions. Any model vendor claiming agentic coding capability should be able to answer whether its training used verifiable agentic tasks or was adapted from a general-purpose base.
Verbosity is a hidden pipeline cost that benchmarks do not surface. Artificial Analysis measured North Mini Code generating three times the output tokens of comparable models. That verbosity compounds across inference cost and latency in high-volume pipelines. Throughput testing against actual workload volume is the evaluation step the benchmark rankings skip.
The frontier pricing split is now a real architectural decision. Fable 5 at $50 per million output tokens and North Mini Code on a single H100 represent a genuine tradeoff between cost control and data residency on one side, and managed infrastructure overhead on the other. Teams running high-volume agentic coding pipelines should model both cost paths against their actual workload before committing to either.
AI is rapidly expanding across finance, but most agentic offerings have yet to reach core production systems. Only 10% of enterprises are using AI tools in a meaningful, production-grade way. Not because of a lack of interest, but because connecting AI to core systems to trade capture, risk, and surveillance is still a work in progress.
These systems offer the greatest opportunity for AI to simplify finance operations through efficient workflows and live trading queries. Yet, legacy systems force this technology to operate in isolation. The volume of architecture connected to traditional platforms often creates this constraint.
Andrew George
Managing Director and Solutions Architect at 3forge.
The financial services industry has forced firms to adapt core architecture rather than replace it, preserving operations, but limiting AI compatibility. Now, the challenge is incorporating AI into these existing systems without forcing an infrastructure replacement that would cause platforms to pause or fail.
Latest Videos From
To bridge the gap between existing systems and modern demands, firms need an architectural layer to help bridge legacy access, implement a governed AI gateway, and introduce AI-native workflows within trusted guardrails. With the right foundation, firms can extend these capabilities directly into production systems and utilize the full value of AI.
Advertisement
Taming the legacy stack without rewiring it
Years of regulations, acquisitions, asset-class specialization, and incremental development without a shared core have created an extensive stack of internal software required to keep operations running – a stack that was never designed to support responsive, AI-driven interaction.
Rather than rebuilding these systems, financial institutions are introducing an architectural layer that unifies access across fragmented infrastructure. This virtualized approach eliminates the need for costly rewiring while allowing organizations to consolidate access to both static and streaming data.
Instead of adding complexity, it creates a simpler path to deploying AI within existing environments.
Advertisement
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
IT teams can start this process by establishing a single abstraction layer across fragmented systems, allowing technology integration while applying entitlements at the data layer. In practice, this would allow:
Natural-language interrogation: Organization-specific data through chatbots and AI assistants.
Virtualization of systems: Abstraction of all systems behind a permission-aware access point.
Safe interaction: AI accessible touchpoints within operational infrastructures.
When organizations effectively apply abstraction layers to existing legacy architecture, AI can improve functions while interacting with internal systems through a controlled, permission-aware layer.
A controlled gateway for AI interaction
Abstraction layers are most effective when financial institutions apply them with gateways for AI access. When organizations apply these models together, this infrastructure creates a controlled AI interaction layer that provides a deliberate medium for producing deterministic, repeatable outputs.
Advertisement
Agents can then access data exclusively through the created pathway. This architecture creates transparency and provides for the application of a consistent set of data and functional access controls.
Ultimately, it allows stakeholders to gain confidence and trust, allowing agentic solutions to migrate from an assistive layer to an operational one capable of coordinating workflows, executing logic, and interacting with live systems.
Through this channel, agents can operate within defined policies and fully log all outputs, verifying repeatability and providing compliance teams with unified oversight of operations. A single control plane can grant permissions, log events, and instantly kill defective outputs, assuaging regulatory concerns.
Advertisement
These capabilities allow AI to expand financial institution growth in production-ready technological environments.
Accelerating development inside trusted boundaries
Once these foundations are in place, AI development can accelerate inside trusted boundaries. By doing so, organizations can reduce code surface area and shorten audit cycles.
Within these types of environments:
Advertisement
AI is equipped with proper boundaries for successful development.
Agents can generate layouts, workflows, and full applications.
AI can operate inside transparent and fully auditable runtimes.
Advanced coding can often power this controlled scale, offering development workflows that promote multimodal interaction, including voice, visual, and text. These capabilities further facilitate AI to fully operationalize efficient workflows across financial organizations.
However, when implementing AI adoption pathways, many organizations are now working through how to scale these capabilities consistently across systems. Financial firms facing this dilemma should follow the example of other industries.
The shift from rebuilding to building on top
Other industries have already solved a similar challenge of rebuilding their technology stacks much earlier in the development process. When this issue arose, they standardized their foundation across their industry, focusing on differentiated delivery rather than excessive rebuilding.
This often meant establishing application engines, a feature now used in gaming (Unity/Unreal), E-Commerce (Shopify), and general CRM (Salesforce).
Advertisement
If IT teams adopted these systems, purpose-built for finance, financial firms could focus primarily on delivery.
An engine could lay the foundation for virtualized legacy access, AI-governed gateways, and AI-native development within trusted guardrails, avoiding a full infrastructure replacement and establishing a safe way to integrate technology that reduces manual reconciliation.
A new foundation in financial systems
As AI moves deeper into core financial systems, the opportunity is not just in deploying models but rethinking how software is built and operated. Application engines provide a path forward by allowing firms to integrate AI into live systems, scale workflows, and generate new functionality from human intent, all within a governed environment.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit
NASA has named Randy Bresnik, Luca Parmitano, Frank Rubio, and Andre Douglas as the crew for Artemis III, which has been reworked from a moon-landing mission into a roughly two-week Earth-orbit test of lunar landers being built by SpaceX and Blue Origin. NBC News reports: Randy Bresnik, Luca Parmitano, Frank Rubio and Andre Douglas are expected to launch into Earth orbit next year, with the goal of testing two commercially developed lunar landers that are slated to carry astronauts to the surface of the moon during the Artemis IV mission in 2028. Bresnik will be the mission’s commander, with Parmitano, an Italian astronaut with the European Space Agency, serving as the pilot. Douglas and Rubio will be mission specialists, and Bob Hines will train with the crew as a backup member. “This test flight will enable us to prove we can carry out highly choreographed operations with our partners across hardware interfaces, software propulsion systems and life support elements with crew in the high-stakes space environment,” Jeremy Parsons, NASA’s Artemis program manager, said during NASA’s announcement on Tuesday.
Bresnik has been to the International Space Station twice, most recently as commander of an expedition in 2017. A retired U.S. Marine colonel, he was selected as a NASA astronaut in 2004. Bresnik has helped oversee development and testing of spacecraft for the Artemis program as an assistant to the chief of the Astronaut Office, which manages astronaut training and operations. Parmitano has also done two stints on the ISS and served as commander of an expedition in 2019. He has completed a total of six spacewalks and also performed the first live DJ set in orbit. Before becoming an astronaut, Parmitano was a test pilot for the Italian air force.
For Rubio, a physician with 28 years of service in the Army, Artemis III will be his second trip to space. From 2022 to 2023, he spent 371 days on the space station, breaking the record for longest-duration spaceflight by an American, according to NASA. Douglas is the only crew member making his spaceflight debut. An engineer who previously worked on space exploration and robotics at Johns Hopkins University Applied Physics Lab, he became a NASA astronaut in 2022. Douglas was the backup crew member for the Artemis II mission around the moon earlier this year. He told NBC News in an interview after Tuesday’s announcement that the role had at times been a challenge. “It was hard to figure out how do you balance getting ready to go, not go, all that stuff,” he said. “But to go now is just fantastic.”
Back in January of this year, RFK Jr. clearly strong armed the CDC into changing the childhood vaccination schedules in America to mimic those of Denmark. The public messaging was crafted to sound as reasonable as possible and amounted to a claim that America was going to revise vaccination schedules to match those of another successful, industrialized, peer country. There were a couple of problems with the move.
For starters, Kennedy did his usual move of trying to make this change completely outside of the normal process for such things. There was no indication that any of this was done at the behest of his reformed ACIP panel. It didn’t go through the normal scientific checks and balances. And even if it had, the courts later put a stay on all such changes, because Kennedy didn’t follow the American Procedure Act in either those revised schedules or even the formation of ACIP itself. The Trump administration has appealed that decision.
The other main issue with the change was the obvious one: America is not Denmark. Calling Denmark a peer nation to America is laughable for many reasons. As one Danish official pointed out at the time: Denmark has a homogeneous population, universal free healthcare, lower serious outcomes from infectious diseases that they don’t vaccinate for, and a population that actually largely trusts government institutions. America doesn’t have any of that, in large part because the party of Trump doesn’t want us to have it.
Donald Trump doesn’t know how to take an “L”, though, so of course he simply picked up a pen recently and is attempting to executive order his way to trying to change those same vaccination schedules.
Advertisement
While the federal government is appealing that injunction, the new executive order on Friday reaffirms Kennedy’s plans to adopt Denmark’s strategy, calling for “realigning” US vaccine policy with “best practices from peer, developed countries.”It states that the scientific assessment written by Høeg and Kulldorff is a “guiding resource for the Federal Government” and that the CDC shall ” take any appropriate steps to update the United States childhood and adolescent vaccine schedule.”
As before, the AMA is strongly against the unilateral change made without backing from scientific evidence.
“Altering [the vaccine schedule] without clear, evidence-based justification risks continued confusion for parents and patients, undermining trust in vaccines, and ultimately lowering vaccination rates,” Mukkamala said. “That would put more children and communities at risk of preventable illness.”
The American Medical Association (AMA) wasn’t the only one to come out against this top-down edict. The American College of Physicians (ACP) likewise pushed back on the EO publicly, stating unequivocally that it must not be implemented or there would be severe negative health outcomes for American children.
As did, hilariously, scientists in Denmark itself.
Advertisement
Anders Hviid, who leads research on vaccine safety and effectiveness at the Statens Serum Institut, Denmark’s equivalent of the CDC, told The New York Times in December that it did not make sense to compare the US to Denmark. “It’s not at all fair to say look at Denmark unless you can match the other characteristics of Denmark,” he said.
Hviid also told the Times that the US public health policies under Kennedy “get crazier and crazier” by the month. “It is surreal, and it is difficult, from a Danish perspective, to understand what’s going on.”
Trust me, dear Anders, it’s difficult to understand from within the American borders, too.
Now, neither Trump nor Kennedy give a flying damn about Denmark, of course. That much is obvious to anyone with a working frontal cortex. The country’s vaccination schedules are merely being used as a prop to reduce the vaccination schedules for American children because that’s all Kennedy really wants. Over the objections, it turns out, of Danish scientists themselves.
I’m sure the AMA, ACP, or the American Academy of Pediatricians (AAP) will be filing lawsuits over this Executive Order. And I see no reason why the courts shouldn’t put a hold on its implementation, as it did to Kennedy.
Advertisement
But the real mystery is why the do-nothings in Congress just can’t be bothered to push back directly on all of this.
MSP says it is ‘absolutely devastated’ as woman arrested on suspicion of murder
Neil Muller, newly appointed Group CEO of managed service provider Node4, has died after an alleged stabbing at his home. He was 54.
Muller, a well-respected and long-serving figure in Britain’s tech supply chain, was found with chest wounds at his residence in Claverdon, Warwickshire, in the early hours of June 7.
Advertisement
Warwickshire Police said in a statement: “We received a report from ambulance services at 6.15am about a man in his 50s who required emergency medical care following a stab wound in his chest. Sadly, he was declared deceased at the scene at 6.37am.”
A 55-year-old woman from Birmingham was arrested on suspicion of murder at 7.33am and has since been released on bail. Police confirmed an investigation is underway and said there is “no wider risk to the public.”
Muller had only taken on the Group CEO role at Node4 this month, tasked with refining its strategy and expanding its AI-augmented managed services platform. The MSP said it was “absolutely devastated” by his death, adding: “Although Neil only recently joined Node4, he made a meaningful impact in a short space of time. Our thoughts are with Neil’s family at this very difficult time.”
Before Node4, Muller led Digital Space for seven years, and prior to that he was chief exec at telecoms biz Daisy Group, whose B2B ops merged with Virgin Media O2 last year. Muller started his career at Computacenter – one of Europe’s largest services-based resellers – rising through sales and operations to become UK and Ireland managing director during a 21-year tenure.
Advertisement
Mike Norris, Group CEO at Computacenter and a close friend of Muller, told The Register that he was “deeply saddened from a personal point of view.”
Norris was not alone: many in Britain’s tech business community expressed shock. Charles Bligh, former TalkTalk chief operating officer, wrote on LinkedIn: “Just so shocked to hear this terrible news. Neil was a class act and he filled the room with his energy and leadership. My condolences to his family and his children should know their father was a respected, liked and thought leader in the business community. I know this is cold comfort. Neil you will be missed terribly and RIP.”
Muller is survived by his wife and two children. ®
Even before the recent nationwide rise in gas prices in response to geopolitical circumstances in the Middle East, fuel economy has always been one of the biggest factors that buyers consider when choosing a new vehicle. Whenever gas prices go up, the differences in fuel efficiency ratings between different models matter more to people.
Yet even with elevated fuel prices, the popularity of pickup trucks in America seems unlikely to fade. And while these vehicles may not be as efficient as your average sedan or crossover, recent fuel economy gains have been impressive. So which trucks are the most efficient? If you’re looking for maximum pickup truck efficiency overall, Rivian is the truck brand that gets the win, but its all-electric models are not directly comparable to internal combustion models.
Advertisement
When it comes to traditional pickup trucks with internal combustion engines, a few models come out on top in the fuel economy rankings, though their efficiency is heavily dependent on which engine option you choose. Thanks in particular to the growth of the hybrid truck market, many of these pickups have far better MPG figures than you might assume — we’ve highlighted the standouts below.
Advertisement
Ford Maverick
Sure, you could put an asterisk next to the Ford Maverick’s inclusion in this list because it’s not a “real” body-on-frame pickup and instead uses a more car-like unibody platform with front-wheel drive being the standard drivetrain layout. Buyers don’t seem to mind, though, and the Maverick has been an absolute hit as far as sales figures go.
When it comes to gasoline use, the Maverick is the most fuel-efficient pickup choice out there, by a significant margin. In its front-wheel-drive hybrid form, the Maverick delivers an incredible 38 MPG combined rating from the EPA, with the all-wheel-drive hybrid version coming in just behind that. Maverick buyers can also opt for non-hybrid models with the more powerful 2.0-liter turbocharged engine, and though the fuel efficiency ratings for the 2.0 Maverick aren’t as high, they are still quite good by pickup truck standards.
As our testing has found, though, there’s more to the Ford Maverick than just great MPG. The truck has also won over buyers who like its compact size, and the Maverick’s affordable price tag is another huge draw at a time when many feel that pickup truck prices are ballooning out of control. When looking for a sensible truck, the potential to save serious money at the gas pump is just one part of the Maverick’s compelling formula.
Advertisement
Chevy Silverado 1500 and GMC Sierra 1500
One might assume that full-size half-ton pickup trucks would be pretty far down a list like this, well behind the smaller mid-size trucks, but that’s not actually the case. Per the EPA’s fuel economy ratings, the truck that comes in second place to the Ford Maverick for fuel efficiency is the Chevrolet Silverado 1500, along with its GMC twin, the Sierra 1500.
More specifically, it’s the Silverado and Sierra powered by the stellar 3.0-liter Duramax inline-six turbodiesel engine. When equipped with this engine, the two-wheel drive Silverado and Sierra 1500 both deliver a combined EPA rating of 25 MPG. The engine’s stout horsepower and torque numbers aside, this is a number that was formerly unheard of when it comes to full-size pickups.
Advertisement
However, when talking about fuel efficiency and saving money on gas, it’s important to note that diesel prices are often higher than gasoline, and even more so under current conditions. You can, of course, get Silverados and Sierras with gasoline engines like the 2.7-liter turbocharged four-cylinder. Despite its lower fuel economy rating, the gas 2.7 could end up being cheaper to fuel than the more efficient turbodiesel models, and even the EPA’s official estimates show the expected annual fuel costs of the two engines are nearly identical.
Advertisement
Ford F-150
The Ford F-150 pickup isn’t just a perennial bestseller, depending on powertrain options, it can also be one of the most fuel-efficient trucks on the road. When equipped with its available hybridized 3.5-liter V6 engine, the F-150 PowerBoost is actually the most fuel-efficient non-diesel full-size truck you can buy right now.
The F-150 PowerBoost has an EPA combined rating of 23 MPG, which includes an especially high rating of 22 MPG in city driving — and better yet, that’s the rating for a four-wheel drive model. The Chevy and GMC Duramax diesel half-ton trucks might have the edge in pure fuel efficiency, but per the EPA, the gasoline-powered F-150 Hybrid should be significantly cheaper to fuel up than a comparable Duramax truck.
In our review of the 2025 F-150 Hybrid, we found the added efficiency of the F-150’s hybrid option to be well worth its extra cost, which also buys you a nice bump in horsepower and a fairly substantial boost in torque over the non-hybrid 3.5 EcoBoost F-150. Ford of course, offers other less efficient powertrains in the F-150, including the Raptor R’s gas-guzzling supercharged V8 — but when it comes to balancing performance and fuel economy, the PowerBoost version is hard to beat.
Advertisement
Toyota Tacoma
If you are in the market for a mid-sized pickup truck and place a high value on fuel economy, the Toyota Tacoma is going to be difficult to top. Toyota currently offers the Tacoma with two different engines — a standard 2.4-liter turbocharged four-cylinder or a hybrid i-Force Max version of that same engine, which is available on the truck’s upper trim models.
Toyota has been greatly increasing hybrid options across its vehicle lineup in recent years, but the hybrid Tacoma is less about raw fuel efficiency and more about the added horsepower and torque compared to the non-hybrid version. In our review of the 2024 Toyota Tacoma TRD Pro, we found the engine’s impressive 326 hp and 465 lb-ft of torque to be a significant improvement over the base, non-hybrid versions.
Advertisement
However, if you’re simply looking for the most fuel-efficient Tacoma, there actually isn’t a massive difference between the Tacoma Hybrid 4WD’s 23 MPG combined EPA rating, and the non-hybrid 4WD model’s 21 MPG combined rating. If you factor in the added cost of the hybrid powertrain, the money-saving math probably won’t work out in the hybrid’s favor, but that’s not surprising. As mentioned a moment ago, the i-Force Max version of the Tacoma is more about driving performance than saving money at the pump.
Advertisement
Ford Ranger
Not to be overlooked between the compact Ford Maverick and the larger F-150 is the mid-sized Ford Ranger, which also offers strong fuel-efficiency numbers for its class. Unlike some of the other trucks on this list, the Ranger gets these numbers not from a hybrid or upmarket turbodiesel powertrain, but with its base 2.3-liter turbocharged EcoBoost four-cylinder.
In its two-wheel-drive spec, the 2.3-liter Ranger earns an EPA rating of 23 MPG combined, which, for the money, is about as good as you’ll find in any body-on-frame pickup truck on sale right now. Our review of the current generation Ranger XLT found that the truck’s real-world economy backs those numbers up, which we found very impressive for a non-hybrid, base engine. Those looking for more power from their Ranger can opt for the larger 2.7-liter EcoBoost V6 engine, but as you’d expect, that drops overall fuel economy numbers by a few miles per gallon.
Beyond that, there’s the much more powerful Ford Ranger Raptor, which, not surprisingly, drops that combined EPA rating down even further to 17 MPG. Given recent electrification trends, perhaps at some point Ford will add a hybrid Ranger option similar to the F-150, which would likely take the Ranger’s already-strong fuel economy numbers to a new level.
Advertisement
Methodology
When putting this list together, we used official EPA fuel ratings as our primary source to pick five of the best-performing, current pickup trucks with internal combustion engines. We also used our first-hand experience on most of these models to back up the selection with real-world fuel economy and performance observations. We all allowed current truck models with either gasoline, hybrid, or turbodiesel engines.
Marshall’s prolific 2026 continues with the launch of the Stockwell II wireless speaker, which boasts some pretty big specs.
The Stockwell II was one of our favourite Bluetooth speakers when it launched, so we’re looking forward to seeing how Marshall can improve, and on paper, there are some big improvements.
But before that, let’s start with sustainability. The Stockwell III introduces replaceable and modular components, a list which includes the battery, carry strap, silicone sleeve as well as the front and rear grilles to ensure that the Stockwell III can last as long as possible without the need for a full replacement if it gets damaged.
To help its longevity, battery life has been expanded from the Stockwell II’s 20+ to 40+, practically double the battery life than before. The speaker can also act as a powerbank for other devices (such as your mobile device).
Advertisement
Image Credit (Marshall)
Advertisement
Marshall’s True Stereophonic 360 sound has made its way to the Stockwell III, offering consistent audio from whichever angle the speaker is placed to ensure there’s no ‘sweet spots’. Dynamic Loudness also features, taking care of managing bass, mid and treble at any volume and keeping them in balance.
The design remains practically the same, with its vertical silhouette and guitar-inspired PU leather strap and velvet lining. Controls have been updated to make it easier to access presets with the M-button or skip tracks with the media jog. An IP55 rating means that it’s not fully waterproof, but can survive a dip into water for a small amount of time. That’s still a jump up from the Stockwell II’s IPX4 rating.
Pricing is within reach of the older model in some markets, but overall it is more expensive. The Marshall Stockwell III has a price of £199.99 / $249.99 / €229.99 with availability in August (you can register interest in the speaker now). Colours are a choice of Black and Brass, and Cream.
You must be logged in to post a comment Login