“Research published in the World Journal of Men’s Health found evidence that drugs such as Viagra and Cialis may also help with heart disease, stroke risk and diabetes,” reports the Telegraph, “as well as enlarged prostate and urinary problems.”
Researchers found evidence that the same mechanism may benefit other organs, including the heart, brain, lungs and urinary system. The paper reviewed a wide range of published studies [and] identified links between PDE5 inhibitor use and improvements in cardiovascular health. Heart conditions were repeatedly cited as an area where improved blood flow and muscle relaxation may offer benefits. Evidence also linked PDE5 inhibitors with reduced stroke risk, likely to be related to improved circulation and vascular function. Diabetes was another condition where associations with improvement were identified… The review also found evidence of benefit for men with an enlarged prostate, a condition that commonly causes urinary symptoms.
Mobile phones had long been an integral part of our daily lives when April 2000 arrived. People took them everywhere because they were a must-have for younger users. Reporter Lindsey Fallow looked closely at how these phones were on the verge of becoming something major, such as having continual access to email and the internet right in the palm of your hand.
Lindsey starts with checking mobile email. Anyone with a phone that was less than two years old could send and receive text messages. There were services that would forward emails from your regular email account to your phone as text messages, and the greatest part was that registration was free, however each downloaded message cost approximately 6 pence ($.15 today). To respond, you would need to construct a text message, include a specific code at the beginning, and submit it to your service provider. Typing on such tiny keypads took a long time, and the expense quickly mounted up.
Google Pixel 9a is engineered by Google with more than you expect, for less than you think; like Gemini, your built-in AI assistant[1], the incredible…
Take amazing photos and videos with the Pixel Camera, and make them better than you can imagine with Google AI; get great group photos with Add Me and…
Google Pixel’s Adaptive Battery can last over 30 hours[2]; turn on Extreme Battery Saver and it can last up to 100 hours, so your phone has power…
She demonstrates with a short exchange, beginning with an incoming message that reads “Can you meet me for lunch to talk about the report? Can you find a restaurant sushi?” she asks, wondering where to eat. She pulls out a WAP phone, which she refers to as a “mobile with internet built in,” and we can see why: previous attempts to get phones to access the internet failed because the whole web requires a large color screen, and most mobiles at the time only had a couple of inches of screen space.
WAP phones changed all that by rewriting web material specifically for small screen sizes. Pages had to be recoded, so the entire internet remained out of reach. Still, useful sites existed. Fallow navigates to the BBC’s pages and to H2G2—a user-edited guide inspired by The Hitchhiker’s Guide to the Galaxy, full of searchable entries anyone could contribute to. She searches for lunch spots and locates a sushi restaurant right around the corner. The screen shows basic text results, no images or fancy layouts, but the information arrives where she needs it.
These WAP phones were retailing for about £130 ($334 today) with a contract, and more were on their way. Services were also constantly expanding, and Lindsey highlights both progress and problems. When a follow-up email arrives stating that lunch has been canceled and that the report should be sent instead, responding with only text messages is inconvenient and can take hours to complete.
Following that came the early smartphones. Lindsey tries out a prototype with a much bigger screen. It includes a full web browser for WAP material, a calendar, and a note feature, as well as handwriting recognition on a touch-sensitive surface. If the handwriting does not work out, a little keyboard appears that you can use. Navigation is a lot speedier and easier on the eyes. These devices promised to combine the power of the web with organization and communication, all in one convenient package. They were expected to hit the shelves that summer for between £300 to £400 ($770 to $1,029 today) with a contract.
The investment marks a significant moment for the organisation as it prepares to advance its ‘Real World Model’.
Stanhope AI, a London-based deep-tech start-up, has announced the closure of an $8m seed funding round. The round attracted a transatlantic cohort of investors led by Frontline Ventures, with participation from Paladin Capital Group and Auxxo Female Catalyst Fund, as well as follow-on investment from UCL Technology Fund and MMC Ventures.
A 2023 spin-out from University College London and King’s College London, Stanhope AI was founded by Irish computational neuroscientist Prof Rosalyn Moran and theoretical neurobiologist Prof Karl Friston.
The team at Stanhope AI has been building a new AI model for autonomous systems that allows machines to “mimic the human brain”, drawing from Friston’s ‘Free Energy Principle’ – a framework developed to explain how intelligent systems minimise uncertainty through continuous perception and action.
Advertisement
According to the start-up, this “brain-inspired paradigm”, known as active inference, enables machines to learn and adapt on the move, which Stanhope AI believes is a crucial capability missing from large language model-based systems that rely on large static datasets.
Stanhope AI’s technology is currently being tested in autonomous drone and robotics applications with international partners, with the goal of teaching machines to behave more intelligently in unpredictable, real-world environments.
According to the organisation, the investment marks a significant milestone as Stanhope AI advances its ‘Real World Model’, which it described a next-generation framework for adaptive intelligence, “designed to function in dynamic, physical environments beyond the limitations of large language models”.
“We’re moving from language-based AI to intelligence that possesses the ability to act to understand its world, a system with a fundamental agency,” said Moran, who is also the company’s CEO. “Our approach doesn’t just process words, it understands context, uncertainty and physical reality.”
Advertisement
In a post on LinkedIn, she explained that the investment is about more than just fresh capital, stating it is a “clear point of technology maturity”.
“Over the past two years in London, we’ve progressed from foundational research and early prototypes to production-grade systems operating in real customer environments, engineered for explainability and scalability,” she said. “The round is also a validation of that journey and evidence that our technology performs beyond the lab.
“We’re proud to be building from London, a deep-tech ecosystem increasingly global in its reach, and equally proud to be backed by investors spanning the UK, US and Europe. That transatlantic support reflects both the ambition of the technology and the scale of the opportunity ahead.”
She added that the funding will accelerate deployments, expand the team and advance the “next phase of applied AI via active inference”.
Advertisement
In other AI start-up funding news, on Tuesday (10 February), Dublin-based property management AI start-up Marc raised $1m from angel investors in a pre-seed funding round. The platform uses AI to analyse fragmented sources of vendor contract and invoice data related to property units and consolidates the information for use by owners and managers to help identify discrepancies leading to overpayments.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
The compilation continues Konami’s recent strategy of remastering the franchise’s most celebrated entries for today’s hardware while retaining their original design and character. Read Entire Article Source link
Starting in 2009, the U.S. government have given car manufacturers towards reducing greenhouse gas emissions if they included “start-stop” systems in cars with internal combustion engines. (These systems automatically shut off idling engines to reduce pollution and fuel consumption.)
But this week the new head of America’s Environmental Protection Agency eliminated the credits, reports Car and Driver:
[America’s] Environmental Protection Agency previously supported the system’s effectiveness, noting that it could improve fuel economy by as much as 5 percent. That said, the use of these systems has never actually been mandated for automakers here in the States. Companies have instead opted to install the systems on all of their vehicles to receive off-cycle credits from the feds. Virtually every new vehicle on sale in the country today also allows drivers to turn the feature off via a hard button as well. Still, that apparently isn’t keeping the EPA from making a move against the system.
“I absolutely hate Start-Stop systems,” writes long-time Slashdot reader sinij (who says they “specifically shopped for a car without one.”) Any other Slashdot readers want to share their opinions?
Advertisement
Post your own thoughts and experiences in the comments. Start-Stop systems — fuel-saving innovation, or a modern-day auto annoyance”
This week Apple patched iOS and macOS against what it called “an extremely sophisticated attack against specific targeted individuals.”
Security Week reports that the bugs “could be exploited for information exposure, denial-of-service (DoS), arbitrary file write, privilege escalation, network traffic interception, sandbox escape, and code execution.”
Tracked as CVE-2026-20700, the zero-day flaw is described as a memory corruption issue that could be exploited for arbitrary code execution… The tech giant also noted that the flaw’s exploitation is linked to attacks involving CVE-2025-14174 and CVE-2025-43529, two zero-days patched in WebKit in December 2025…
The three zero-day bugs were identified by Apple’s security team and Google’s Threat Analysis Group and their descriptions suggest that they might have been exploited by commercial spyware vendors… Additional information is available on Apple’s security updates page.
Advertisement
Brian Milbier, deputy CISO at Huntress, tells the Register that the dyld/WebKit patch “closes a door that has been unlocked for over a decade.”
Thanks to Slashdot reader wiredmikey for sharing the article.
I spend a lot of time talking with teams that are trying to expand their AI efforts, and I’ve noticed a consistent pattern: AI pilots are multiplying across the board, but a majority of them fail to see the light of production.
Data tells a similar story: Only 26% of leaders report more than half of their pilots scaling to production. Meanwhile, 69% of practitioners (the front-line teams embedding AI into workflows) say most of their pilots are never scaled.
And while leaders remain confident in their AI pilot timelines, 75% of practitioners believe leadership underestimates how hard AI execution really is.
Advertisement
Brandon Sammut
Chief People & AI Transformation Officer at Zapier.
Advertisement
That gap between ambition and execution is where momentum leaks out of AI programs. But when leaders spend time closing that gap, they can turn AI pilots into meaningful progress.
Where AI momentum starts to slip
As AI pilots move closer to production, the work changes shape.
Early progress often looks clean. Teams define a use case, test a model, and show early results. As pilots expand, they start touching real systems, shared data, security reviews, and downstream workflows. That’s where timelines stretch and attention fragments.
Advertisement
One of the most consistent signals in the data is where teams get stuck: leaders rank integration complexity and system sprawl as the biggest barriers to AI execution. Practitioners confirm the same reality, pointing to integration backlogs and policy delays as top blockers. What’s missing here isn’t effort or intelligence. It’s orchestration.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
AI pilots are often designed as standalone wins. They prove that a model can work, but not that it can survive inside a web of existing tools, data sources, approvals, and workflows. When those connections aren’t planned early, teams end up rebuilding work that already “worked,” just not at enterprise scale.
This is where momentum quietly drains away. Each delay feels reasonable on its own. Taken together, they stretch timelines, consume trust, and make scaling feel heavier than starting.
Advertisement
Confidence fades when visibility lags
Another pattern shows up in how leaders and practitioners experience progress.
Eighty-one percent of leaders say they’re confident in their visibility into AI execution challenges. At the same time, 57% of practitioners believe leadership doesn’t fully see what’s happening day to day.
Advertisement
That disconnect matters because feedback usually arrives late. Leaders learn about failures after the fact, most often through escalations or informal conversations. By then, projects have already lost momentum or require rework.
What’s left is a cycle of fixing execution issues instead of moving work forward. That reactive rhythm makes AI feel unpredictable, even when the underlying goals are clear.
Leaders who close the gap don’t rely on confidence alone. They create shared visibility into execution, so friction shows up while there’s still time to address it.
Advertisement
What actually helps AI pilots turn into progress
You don’t have to rewrite strategy to close the gap between ambition and execution. Instead, spend more time where friction actually shows up. These focus areas consistently make a difference.
Start with clear ownership
AI pilots move faster when someone is accountable beyond the initial delivery. This is where internal AI champions come in. They’re the ones responsible for production outcomes and can be the clear decision-maker when tradeoffs arise.
That ownership helps teams resolve integration questions, prioritize follow-up work, and keep pilots from drifting into the ether of failed experiments.
Advertisement
Invest early in integration planning
Integration complexity is a common execution challenge across leaders and practitioners. When integration work is deferred until after a pilot proves technical value, teams often revisit assumptions under time pressure.
This is where it’s crucial to ask questions early and often about systems, data flows, and workflow dependencies to help teams design pilots with scale in mind. That early clarity reduces rework and shortens the path to production.
Advertisement
AI execution accelerates when teams share both a common foundation and a way to learn from each other. Standardized tools and approaches reduce the overhead of every new pilot by giving teams familiar patterns to build on.
Pairing that consistency with peer learning and internal upskilling helps knowledge travel across the organization. When teams share lessons from what worked and what didn’t, each pilot builds on the last. Over time, experimentation turns into repeatable capability instead of isolated wins.
Build governance into delivery
Governance pressure tends to increase as AI pilots move closer to production. When governance enters late, teams pause work while policies are interpreted, approvals are routed, and risks are reassessed.
Advertisement
Embedding governance directly into delivery workflows helps teams understand constraints earlier and move with confidence. Clear guardrails, applied consistently, reduce late-stage surprises and keep execution moving as pilots scale.
Create shared visibility into execution
Execution slows when leaders and practitioners see different versions of progress. Leaders often track milestones and timelines, while practitioners experience day-to-day friction through integration work, reviews, and rework.
Shared visibility bridges that gap. Live signals, clear escalation paths, and agreed-upon success criteria surface issues while they’re still manageable. That alignment reduces firefighting and helps teams stay focused as scope expands.
Advertisement
Budget for scale
Many AI pilots stall once early funding runs out. Integration, governance, and long-term delivery require sustained investment beyond the initial build.
Leaders who plan budgets with scale in mind give teams room to carry successful pilots forward. Dedicated funding signals that execution matters as much as experimentation, and it creates the conditions for pilots to mature into systems that last.
Advertisement
Turning ambition into results
The next phase of enterprise AI won’t be defined by who launches the most pilots. It’ll be defined by who learns how to see execution clearly enough to keep those pilots moving.
As AI becomes part of everyday operations, the advantage shifts toward leaders who stay close to the work as it scales, who notice friction early, and who treat execution signals as strategic input rather than noise. That kind of attention compounds. Teams spend less time recovering.
Decisions get made faster. Confidence grows where progress is visible.
Over time, AI stops feeling fragile. It becomes dependable. And when that happens, ambition no longer outruns impact, it sets the pace for it.
This article was produced as part of TechRadarPro’s Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro
From top left, clockwise; Gatefolded founder Jasen Samford; StackIQ founder Jana Schuster; SageOx co-founder Ajit Banerjee; Vivu founder Shawn Neal; HYV Social co-founder Jason Lee; and PrimeOrbit founder Mahadev Alladi.
We’re back with our latest spotlight on early stage Seattle-area startups. This edition features founders building software for video editing, releasing music, AI chats, SaaS sprawl, coding with AI agents, and making in-person connection.
Read on for brief descriptions of each company — along with pitch assessments from “Mean VC,” a GPT-powered critic offering a mix of encouragement and constructive feedback.
Check out past Startup Radar posts here, and email me at taylor@geekwire.com to flag other companies and startup news.
The business: A music tech platform that helps artists securely share unreleased tracks while also building direct relationships with fans. Since launching in January, the bootstrapped startup has signed up dozens of artists and begun converting early trial users to paid plans at $49 per year.
Advertisement
Leadership: Founder and CEO Jasen Samford spent a decade at DistroKid, a music tech company that helps musicians get their work onto streaming and video platforms.
Mean VC: “You’re addressing a clear need around pre-release security and direct fan engagement, and early paid conversions suggest some initial product-market resonance. I’d focus on demonstrating consistent artist retention, measurable fan engagement metrics, and a scalable acquisition strategy that shows this can grow beyond early adopters without relying on high-touch onboarding.”
The business: A mobile app designed to help remote and busy professionals turn spontaneous interest in going out into real-world connection. The bootstrapped startup, which launched a beta in Seattle at the end of last year, uses geo-location and consent-based signals to show who nearby is open to meeting in the moment, aiming to reduce social hesitation and awkwardness for busy professionals.
Advertisement
Leadership: Co-founder Jason Lee is a longtime security leader who spent nearly 14 years at Microsoft and was CISO at both Zoom and Splunk. Co-founder Brandon Sene also worked on security at Microsoft, and co-founder Cody Cronberger was a software engineer at Amazon.
Mean VC: “There’s something compelling about turning fleeting ‘I should go out’ moments into action, especially for time-constrained professionals. But this only works if you can create critical mass and a clear reason to open the app repeatedly — so I’d focus obsessively on retention, safety, and proving strong engagement in a single neighborhood before expanding.”
The business: An operating layer for AI conversations focused on turning chat-based interactions into completed actions and workflows across channels. The bootstrapped company aims to help AI-driven products increase growth and engagement by closing the loop after a conversation ends.
Advertisement
Leadership: Founder and CEO Mahadev Alladi spent 17 years at Microsoft, where he helped lead teams working on advertising tech.
Mean VC: “This tackles a real problem — AI chats rarely translate into completed actions — and closing that loop could drive meaningful lift for AI products. The priority should be narrowing to one high-value workflow and proving measurable impact, since broad infrastructure positioning will struggle in a crowded market.”
The business: Tools for AI-native teams where humans and coding agents work side by side. The company describes its product as an “agentic hivemind” designed to capture shared context and keep human developers and AI agents aligned as software increasingly ships with minimal human intervention.
Advertisement
Leadership: CEO Ajit Banerjee previously founded three startups and most recently was at Hugging Face. His co-founders include Milkana Brace, who previously founded Jargon (acquired by Remitly), and Ryan Snodgrass, who spent 15 years at Amazon.
Mean VC: “The vision is timely — AI-native teams need better coordination between humans and agents — and shared context could become critical as autonomous coding scales. The risk is abstraction: focus on a concrete workflow where misalignment is painful today and prove clear productivity gains, or ‘agentic hivemind’ will sound more conceptual than indispensable.”
The business: A decision intelligence platform to help enterprises figure out which SaaS and AI tools they actually need — and which are redundant. StackIQ is working with early customers and design partners, and raised a friends-and-family round.
Advertisement
Leadership: Founder and CEO Jana Schuster held leadership roles at Groupon, Sears, Farmer’s Fridge, Visibly, Amazon, The Honest Company, and most recently Deputy.
Mean VC: “You’re going after a real and growing pain point — SaaS and AI sprawl is expensive and chaotic — and if you can consistently surface redundant spend, your value to enterprises is clear and budget-aligned. To make this investable, you need to prove hard ROI with specific numbers and show how you’ll become embedded in procurement or IT workflows so you’re not just another analytics dashboard that gets replaced or absorbed.”
The business: The bootstrapped startup is working with early pilot customers on an “agentic video workspace” for marketing and growth teams that already have footage but need help turning it into a steady stream of videos. Teams upload real campaign assets, and Vivu drafts multiple editable variants — including hooks, cutdowns, captions, and formats — to speed up production without relying on fully synthetic AI content.
Advertisement
Leadership: Founder Shawn Neal was a manager at Google and Microsoft, and more recently led product at a video AI startup.
Mean VC: “This is a pragmatic wedge — marketing teams sitting on unused footage care about increasing output without going fully synthetic, and editable variants fit how teams actually work. The key will be proving you can deliver materially faster production cycles or higher-performing creatives than internal teams and existing AI tools, or you risk blending into a crowded video tooling market.”
The programme offers third level students practical work experience at Ireland’s national marine research and development agency.
Third level students aspiring to be among the next generation of marine scientists and experts can now apply to the Marine Institute’s 2026 Bursary Programme. The initiative, which has run for 30 years, offers students practical work experience and the opportunity to develop essential skills.
The programme is aimed at undergraduate students enrolled in national or international universities and institutes for higher education. To qualify for participation, students must have completed two years of study in a relevant discipline by June of this year.
Participants will have the opportunity to network with fellow students from third level colleges as well as with experts in their fields. The aim is to enable students to form future connections within the marine research sector.
Advertisement
Successful candidates will work with full time Marine Institute staff on critical work programmes in areas including marine and freshwater fisheries, oceanography, machine learning, AI, marine chemistry, molecular biology, marine spatial planning, remote sensing, web development, socio-economics and corporate services.
The bursaries are based at Marine Institute facilities in Oranmore, Co Galway and Newport, Co Mayo.
Glenn Nolan, the institute’s Bursary Programme lead, said: “For more than 30 years, the Marine Institute Bursary Programme has enabled undergraduate students to develop their skills and strengthen their knowledge of the marine sector.
“Participating students emerge equipped to make informed decisions early in their studies about the marine and maritime careers they would like to pursue.”
Advertisement
To apply for the 2026 Bursary Programme, potential applicants can go to the Marine Institute website.
In October 2025, the Marine Institute announced a five-year project designed to restore native flat oyster reefs and boost the resilience of Ireland’s coasts, with €1.5m in funding from the Marine Institute’s Marine Research Programme.
The BRICONS project is being led by Dr Paul Brooks from the School of Biology and Environmental Science at University College Dublin and includes partners at Atlantic Technological University, Queen’s University Belfast and Trinity College Dublin.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
The match was organized by Rek, a San Francisco-based company, and drew hundreds of spectators who had paid about $60-$80 for a ticket to watch modified G1 robots go at each other. Made by Unitree, the dominant Chinese robot maker, they weighed in at around 80 pounds and stood 4.5 feet tall, with human-like hands and dozens of joint motors for flexibility. The match had all the bells and whistles of a regular boxing bout: pulsing music, cameras capturing all the angles, hyped-up introductions, a human referee, and even two commentators. The evening featured two bouts made up of five rounds, each lasting 60 seconds. The robots pranced around the cage, throwing jabs and punches, drawing ohs and ahs from the crowd. They fell sometimes, and needed human intervention to get them back on their feet.
The robots were controlled by humans using VR interfaces, which led to some odd moments with robots hitting into the air, throwing multiple punches that failed to even connect with their opponents. One robot controller was a former UFC fighter, the article points out, but “The crowd cheered as a 13-year-old VR pilot named Dash beat his older competitor….”
The company behind this event plans more boxing matches with their VR-controlled robots, and even wants to develop “a league of robot boxers, including full-height robots that weigh about 200 pounds and are nearly 6 feet tall.”
Mori3, a modular robot developed by the bright minds at EPFL in Lausanne, Switzerland, is made up of four triangular modules stacked in an origami-inspired pattern that is actually pretty ingenious. Each module communicates with its neighbors, establishing a little team that can change shape, move about, and be useful in a variety of ways.
Mori3’s approach to reliability is what truly sets it apart. Traditional modular robots begin to come apart, literally, as the number of units increases, because all of the extra connections offer a slew of potential weak points. The EPFL team performed the inverse. Power, communication signals, and sensor data are shared directly between modules. This results in hyper-redundancy: the entire system has access to a pool of shared resources, rather than each module doing its own thing. And when additional modules are added, it becomes much more reliable.
LEGO SET FOR ADULTS – The WALL-E and EVE (43279) building set offers adults 18 years old and up an immersive construction challenge featuring…
4 DISNEY PIXAR CHARACTERS – Builders can create iconic robots WALL-E, EVE, M-O and Hal from the hit movie—each with authentic functionality like…
MINDFUL BUILDING EXPERIENCE – This detailed construction set lets builders practice advanced construction techniques for an immersive and relaxing…
Engineers tested the concept by actively attempting to destroy the central module, cutting its power, blocking its wireless communications, and turning off its sensors. That would be disastrous in any normal system; the entire thing would collapse. But not in Mori 3. The remaining three modules just stepped in and took over. They provided power, conveyed data, and transmitted sensor readings to the disabled device. The entire robot continued to function as if nothing had happened.
In a demonstration, the robot walks across tough, uneven terrain with simulated damage and just kept going. When it encountered a low obstacle, the modules altered their layout and managed to slip beneath it, and it just continued marching on the other side. The purportedly “dead” central module was fully functioning and contributed to the effort the entire time; no separate backup hardware was required, only the shared resources of the other modules.
The modules’ triangular form makes them extremely adaptable, allowing them to be moved in a variety of ways. They can walk, flatten out to fit through narrow spaces, and the current configuration of just four units is only the beginning. The approach scales up, so you could picture a group of many modules sharing resources over several connections, making the entire system more resilient as additional units are added.
This technology has a wide range of possible uses, such as a swarm of autonomous robots that can dock together quickly to exchange energy and data. If one unit is damaged, the others can take up the burden, and the entire swarm can continue on. The research demonstrates that it is totally possible to create machines that will not quit up, even when severely injured. [Source]