Blog

This season-long crime series starring a future James Bond was canceled after winning best TV show
When the conversation turns to who is the best James Bond is that there is a name that is often forgotten: David Niven. Niven plays the iconic character in 1967 Royal Casinoa non-canon entry that exists as a parody, released the same year as You only live twice. But three years before Niven took on the mantle of 007, he appeared on The Roguesa fun NBC television crime series that won the Golden Globe for Best Television Series in 1964, but it was canceled after its first and only season.

Con artists use their talents for good in ‘The Rogues’

Gig Young, Charles Boyer and David Niven in a promotional card for The Rogues.
Image via NBC

The Rogues was produced by Four Star Televisiona production company made up of Hollywood stars Dick Powell, Ida Lupino, Carlos Boyerand David Niven in 1952. They would be the last two, accompanied by Young showwho appeared in the series as Englishman Alexander “Alec” Fleming (Niven), Frenchman Marcel St. Clair (Boyer) and American Tony Fleming (Young), a trio of con-artist ex-cousins, who use their tricks of the trusty trade for good. For a price, of course.

The series had a “con of the week” format, and with his cousin Timmy (Roberto Coote) and Aunt Margaret (Gladys Cooper) helping, the Rogues faced a series of scoundrels in often witty schemes. In “The Day They Gave Diamonds”, Tony, Marcel and Timmy sell a greedy mine owner who seeks to control the market by making his own diamonds, a worthless diamond-making machine. In “Hugger-Mugger, by the Sea”, Tony buys an entire batch of 22 ships from the US Navy and then tries to unload them on a corrupt shipping magnate, convincing him that one of the ships has $2 million in gold bars hidden. Aunt Margaret poses as Anastasia Romanov in “The Real Russian Caviar”, creating a ruse that tricks the Russian government into believing she knows where the family fortune is hidden, in an effort to collect a fee for handing over the fortune. And in “Mr. White’s Christmas”, the whole family works together to make a modern-day Scrooge see the error of his ways.

‘The Rogues’ wins best television series, but is canceled anyway

David Niven, Charles Boyer and Gig Young in The Rogues
Image via NBC

Typically, episodes of The Rogues are centered around one of its three big stars, with the others appearing occasionally in small roles to complement the protagonist’s story. Much of this was out of necessity, with Boyer and Niven juggling their time on the show with their active film careers (Boyer would be filming the 1965 romantic comedy). A very special favorwhile Niven would be filming comedies Where are the spies and Lady L), leaving most of the thirty episodes in Young’s hands (Niven himself took the lead in only three). This juggling of the lead role became problematic towards the end of the series, taking Larry Hagman to be introduced as another American cousin to replace Young in the final two episodes, according to Hagman’s autobiography.

This is perhaps one of the biggest reasons whydespite having won the Golden Globe for Best Television Series, The Rogues ended after a 30-episode season. Without consistent leadership, viewers never knew who would show up week after week, a problem if you’re trying to get viewers engaged and committed to watching. Another factor was NBC’s questionable decision to pit the show against ABC’s. Sunday night movie and two fer from CBS candid camera and What’s my line on Sunday nights. Fan reviewshowever, they largely attribute the cancellation to one factor: American humor.

The Rogues it was smartly written, with sharp humor placed in the hands of actors with extensive experience on stage and film, at a time when film and TV actors rarely cross-pollinated. As one reviewer perhaps unfairly argues, The Rogues was too good for American tastes, citing Beverly’s Hillbillies as a “more accurate barometer of the American public’s sense of humor.” Given that the series ran for nine seasons based on its fish-out-of-water premise – a family from the Ozarks got rich and moved to “the Beverly Hills” – it’s certainly plausible. Even so, the cast would move forward, with Hagman free to continue. I dream about Jeannie and Niven free to play James Bond in Royal Casinoaccompanied by co-star Bower, who plays Deuxième Bureau executive Le Grand.
Release date

1964 – 1965-00-00

Directors

Lewis Allen, Robert Ellis Miller, Hy Averback, Don Taylor, Ida Lupino, Richard Kinon, John Newland, William A. Graham

Writers

Warren Duff, Ivan Goff, Stephen Kandel, Charles Hoffman, Samuel A. Peeples, Carey Wilber, Roger H. Lewis, William Link, William Bast, Wells Root, Walter Black, Tom Waldman, Stephen Lord, Ron Bishop, Edmund H. North, Robert Buckner, Richard Levinson, Richard De Roy, Marion Hargrove, Lorenzo Semple Jr., Leonard Kantor, Jon Boothe, Frank Waldman, Francis M. Cockrell

Carlos Boyer

Marcel St.

Roberto Coote

Claire
avots
October 12, 2025
The Weeknd is a sinister entity in the psychedelic video for ‘Big Sleep’

I don’t know if The Weeknd’s “Big Sleep” video is specifically inspired Rick and Mortybut given his penchant for guest-starring on animated shows like The Simpsons, Robot Chickenand American fatherIt doesn’t seem that far-fetched.

After all, the concept is eerily similar to the fan-favorite second season Rick and Morty episode, “Get Schwifty.” In this decade-long story (what is time, anyway?), the Cromulons – giant disembodied heads, lovers of reality music shows – appear above Earth, demanding that the people of our puny planet “show me what you’ve got.”

In “Big Sleep,” the gigantic disembodied body parts belong to Abel himself, who appears as both a large floating head and a hand pointing to the sky. Quickly, though, the video gets very psychedelic, rapidly flashing through the RBG spectrum (red, blue, green – it’s what makes your TV work!) like the strobe lights at a rave. Maybe he’s sending subliminal signals, but this is probably the most abstract video ever Hurry up tomorrow however, given that most of them have at least operated on some earthly concern, such as cults, dark fantasy alternate worlds, and the nature of the afterlife.

The clip also touches on The Weeknd’s love for avant-garde cinema; is directed by Gaspar Noé, one of the vanguards of the New French Extremity movement, and features composer Giorgio Moroder, who created soundtracks for films such as Midnight Express, American Gigoloit is clear, Scar.

You can watch the “Big Sleep” video above.

avots

October 12, 2025
50 years of Quirks & Quarks and half a century of science

Peculiarities and Quarks celebrates its 50th anniversary this week, looking back at half a century of science and looking ahead to the next 50 years.

In 1975, when David Suzuki presented the first episode of Peculiarities and QuarksThe excitement of the Moon landings had subsided and North America faced a new challenge – the effects of the 1973 disaster. energy crisis.

Oil supplies from the Middle East have been disrupted. The price of gasoline skyrocketed overnight, gas stations ran out of fuel, and those that had supplies had lines that stretched for blocks.

The keywords at the time were fuel economy and energy efficiency, as scientists and engineers were tasked with stretching limited oil reserves to the maximum, while also looking for alternative energy sources.

Quirks & Quarks hosts over the years. From left, David Suzuki (1975-1979), Jay Ingram (1979-1991) and Bob McDonald (1992-present). (CBC)

Consumers were switching from traditional, large, gas-guzzling North American cars to more efficient models, often from Japan and Europe. The U.S. auto industry followed suit with smaller, lighter, more aerodynamic vehicles powered by small engines capable of covering every possible mile on a gallon of gasoline.

Research dollars have been invested in renewable energy, which has led to the development of alternatives such as clean hydrogen fuel, improved solar panels, wind turbines, geothermal energy and biofuels.

Homeowners were encouraged to insulate their homes, and eventually high-efficiency heat pumps and furnaces became available, which reduced heating bills.

At the same time, industries began to be held more responsible for air and water pollution. Car manufacturers have been forced to install catalytic converters in exhaust pipes to capture emissions, while large industries have been forced to clean up effluents and limit discharges of waste and harmful chemicals into rivers and lakes.

Taken aboard Apollo 8 by Bill Anders, this iconic image shows Earth peering beyond the lunar surface as the first manned spacecraft circumnavigated the Moon, with Anders, Frank Borman and Jim Lovell on board. (NASA)

The 1970s were a time when science began to turn its focus to caring for planet Earth, thanks in part to photos of Earth taken from the Moon by Apollo astronauts and the influence of Rachel Carson’s groundbreaking 1962 book, Silent Spring. We have come to think of our planet as a small oasis of life in a vast and indifferent universe.

In other words, science was increasingly focused on the environment. It highlighted problems such as carbon emissions, but also presented solutions such as green energy.

During this period, a revolution began at home with the arrival of personal computers, which ended up changing the way we communicate, do business and connect with everyone in the world.

In medicine, the Human Genome Project emerged in the 1990s to sequence all human DNA, leading to a deeper understanding of human biology and the evolution of life.

WATCH | Diana Filer tells how she created the Quirks & Quarks program in 1975:

This woman created the name Quirks & Quarks 50 years ago

Diana Filer, the original producer of CBC Radio’s science program Quirks & Quarks, remembers releasing the first episode in 1975.

So here we are, 50 years later, and a lot has changed. In the developed world, we are largely healthier and live longer than ever before. More food is being produced to feed a population that has doubled and we are surrounded by technology that provides information at our fingertips. We can travel almost anywhere on the globe.

Science and technology have turned humans into a superspecies. We explore every domain on the planet and beyond. We discover our place in an unimaginably huge and expanding universe, investigate the mysteries of life down to the molecular level, and discover connections between all living things. Our knowledge has never been greater.

But all this progress came at a cost. Scientists have repeatedly warned us that the enormous environmental stress on our planet is unsustainable. Our technological advances have often brought paradoxical results. For example, vehicles today are cleaner and more efficient. But many of them have grown back to enormous size – so efficiency has fueled consumption, rather than reducing it.

Vehicles today are cleaner and more efficient than ever before, but they’re also larger on average, meaning we’re not seeing as big a drop in emissions as we could. (Patrick Morrell/CBC)

The natural world is under extraordinary pressure. Scientists have calculated that species are disappearing at a rate not seen since the extinction of the dinosaurs. There are a range of causes, from human pollution and pesticides, to habitat loss due to human activity, to climate disruption.

Every summer, more and more hectares of our forests catch fire. Water scarcity and droughts are becoming more common and impactful; stocked through warmer oceans, storms became stronger.

Despite ambitious international agreements, we have not met our commitments to reduce greenhouse gas emissions. Climate concerns often seem to take a backseat to economic interests. And many researchers point to positive environmental policies that are being obstructed by those trying to maintain business as usual.

The human impact on the planet has even been given a new geological name by some scientists: the Anthropoceneperiod defined by the permanent mark that we are making in the geological record.

It may seem like we live in dark times, but there is hope.

Wind energy, as seen here in Pugwash, NS, is just one of the technologies we can turn to to meet our energy needs without burning up our planet, says Bob McDonald. (Cassie Williams/CBC)

Many of the technological innovations that began to be developed in the 1970s to address oil shortages are now mature and realizing their enormous potential. The technologies to produce electricity from the free energy of the sun, the wind, the heat of the Earth or the energy of the atom are there for us.

We know how to keep the lights on, the wheels turning, and the food on the plate without burning the planet. Science and technology, researchers consistently say, are there for us to use.

However, it can be challenging to distinguish between real science and pseudoscience, between evidence-based research and self-appointed experts who claim that climate change is a hoax, that vaccines or painkillers cause autism, that the moon landings were faked, that the Earth is flat, and that alien civilizations live on Mars.

However, this distinction is vital because the issues we face, such as climate change, energy production, food and drinking water supplies, all involve scientific principles. And a scientifically literate society can make intelligent decisions about how to proceed from here.

Peculiarities and Quarks has been contributing in small ways to this scientific literacy for half a century, and I am extremely proud to have been a part of it.

avots

October 12, 2025
Between excitement and hope: Seattle biotech leaders weigh in on AI’s real impact on drug development
Seattle Biotechnology and AI panel at a one-day conference on October 8, 2025. From left: Alex Federation, Talus Bioscience; Erik Procko, Cyrus Biotechnology; Marc Lajoie, Outpace Bio; Jamie Lazarovits, Archon Biosciences; and Chris Picardo, Madrona. (Photo by Life Sciences Washington)

Dozens of Seattle biotech companies are using artificial intelligence to design new medical treatments. But at a conference of industry leaders and investors this week, scientists delivered a nuanced message: AI holds enormous promise, but expectations must remain grounded in reality.

The trade association Life Sciences Washington and investment company Madrona hosted the one-day forum in downtown Seattle delving into biotechnology, pharmaceuticals and AI.

“There is some hype about the breadth and depth of the capabilities of some of these AI models,” he said. Jamie LazarovitsCEO and co-founder of Archon Biosciences. Researchers need to “be very cautious” when drawing conclusions from the data generated by the models, he said.

And there’s still a lot to be excited about.

“What was science fiction 15 years ago is now reality,” said Erik Procko, chief scientist at the Cyrus Biotechnology. “So yes, there has been enthusiasm. But there is still enormous potential, and sometimes the progress being made is just dizzying.”

Lazarovits and Procko were part of a panel that included four Seattle startups leveraging AI. Each company is facing different challenges in drug development:
- Archona company that emerged from secrecy more than a year ago with $20 million in funding, is using AI to design proprietary protein structures, known as Antibody Cages or AbCs, that are intended to help antibodies bind to target cells and avoid other cells.
- Ciro is developing medicines that focus on identifying and removing areas that will trigger an immune system response – a challenge called immunogenicity. The 10-year-old company has raised $36.6 million according to PitchBook.
- Outpace Biographya startup founded in 2021 that raised $200 million is designing proteins that aim to boost T-cell therapies for solid tumors, which make up 90% of cancers. Tumors are disappointingly good at deflecting current T-cell treatments, which often stop working within a month.
- Talus Bioscienceslaunched in 2020 and worth almost US$20 million, it targets transcription factors – proteins that are part of the “regulome” that turns genes on and off. The company targets transcription factors that activate genes that cause specific cancers.
In addition to discussing their own work, the panelists identified key principles for how AI should — and shouldn’t — be used in biotech research:

Increasing researchers, not replacing them

Marc Lajoieco-founder and CEO of Outpace, compared the AI tools to the robotic exoskeleton used by Ripley and others in the sci-fi film Aliens to carry heavy loads – and fight the ET Xenomorph Queen.

“It makes the researcher better,” he said. “We are not trying to replace the researcher.”

Models must meet reality

Lazarovits noted that while AI can generate interesting clues and information, it doesn’t mean much until it’s tested in real experiments with cells and organisms.

“Whenever we try to adopt new models, new AI methods, you can get incredibly excited about that silicon validation,” he said. “But the reality is: what do you actually validate in the wet lab?”

The real bottleneck: clinical trials

AI is great for developing new therapies, but the most expensive and labor-intensive part of the drug development process is seeing how they work in patients.

“The most impactful place for AI to truly change the game in drug development would be to conduct smaller, better-powered clinical studies,” Lajoie said. The way to do this, he added, would be to find better drug candidates that perform multiple functions.

Still searching for AI’s breakthrough moment

Procko still hopes that AI will go further to spur advances in biotechnology and pharmacology.

“AI is currently fantastic for predicting, say, a protein structure, but for creating new medicines it hasn’t yet found its killer application,” Procko said. “What does AI allow us to do now to produce new medicines that were previously simply impossible to manufacture? How could this be a game changer?”

The question captures a central tension discussed on the panel and at the conference: While AI has transformed the way Seattle biotech companies approach drug design, the industry is still navigating the gap between computational promise and clinical proof.

avots
October 12, 2025
Australian construction robot Charlotte can 3D print a 2,150-square-foot house in one day using sustainable materials
NEWNow you can listen to Fox News articles!

Construction robots are no longer a distant idea. They are already changing their workplace, facing repetitive, heavy and often dangerous tasks. The latest robot comes from Australia, where a spider-like machine named Charlotte is making headlines.

Charlotte is designed to 3D print an entire 2,150 square foot home in just one day. This is equivalent to the speed of more than 100 bricklayers working simultaneously. This offers a glimpse into how the future of housing can be built.

Sign up for my FREE CyberGuy report
Receive my best tech tips, urgent security alerts, and exclusive offers straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide – for free when you join my CYBERGUY.COM/NEWSLETTER

AUSTRALIA DEBUTS FIRST MULTI-STORY 3D PRINTED HOME – BUILT IN JUST 5 MONTHS

Charlotte the robot 3D prints houses in just 24 hours using eco-friendly materials. (Crest Robotics)

How the Charlotte robot works

Charlotte is a collaboration between Crest Robotics and Earthbuilt Technology. The robot is not limited to stacking bricks or tying rebar. Instead, it uses a giant extrusion system that lays down layers of environmentally friendly material.

This material comes from sand, crushed brick and recycled glass, all locally sourced. The result? A fireproof, flood-proof structure created with a much smaller carbon footprint than traditional construction methods.

3D PRINTED SUSTAINABLE HOME BUILT MAINLY FROM THE GROUND

Why Charlotte the Robot Stands Out

This 3D printing construction robot stands out for its unique combination of speed, strength, versatility and affordable price.
- Speed: Print a house in 24 hours.
- Strength: Uses durable and sustainable materials.
- Versatility: It can stand on spider legs to continue building higher walls.
- Accessibility: Eliminates many of the expensive construction steps.
Although Charlotte is still in the development phase, a scaled-down prototype has already been presented. Researchers believe this could help solve housing shortages where labor is scarce and construction costs are skyrocketing.

Its spider-like legs allow it to climb and build higher, reducing costs and saving time. (Crest Robotics)

The Future of 3D Printed Moon Bases Outside Earth

Charlotte’s creators are also keeping an eye on the stars. They envision future versions of the robot building lunar bases for research and exploration. With its compact design and autonomous operation, Charlotte could adapt to the extreme environments of space as well as the challenges on Earth.

THE NEW ROBOT THAT COULD MAKE TASKS A THING OF THE PAST

What does this mean for you

If Charlotte lives up to its promise, it could reshape the way homes are built around the world. Faster construction means faster housing availability. Lower costs and sustainable materials mean more affordable homes with a lower environmental impact. For anyone facing rising housing prices or construction delays, technology like Charlotte can bring a ray of hope.

Take my test: How safe is your online security?

Do you think your devices and data are really protected? Take this quick test to see where your digital habits stand. From passwords to Wi-Fi settings, you’ll get a personalized analysis of what you’re doing right and what needs improvement. Take my test here: Cyberguy.com/Quiz

AMERICA’S LESSONS FROM THE WORLD’S LARGEST 3D PRINTED SCHOOLS

Future versions could even build lunar bases for research and exploration. (Crest Robotics)

Kurt’s Key Takeaways

Charlotte may be years away from building its first large-scale home, but its prototype already points to a future where robots take on critical roles in construction. From solving housing crises on Earth to building shelters on the Moon, Charlotte shows how robotics and 3D printing can work together to solve real problems.

CLICK HERE TO GET THE FOX NEWS APP

Would you live in a house 3D printed by a robot like Charlotte, or even one built on the moon? Let us know by writing to us at Cyberguy.com/Contato

Sign up for my FREE CyberGuy report
Receive my best tech tips, urgent security alerts, and exclusive offers straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide – for free when you join my CYBERGUY.COM/NEWSLETTER

Copyright 2025 CyberGuy.com. All rights reserved.

Kurt “CyberGuy” Knutsson is an award-winning technology journalist who has a deep love for technology, gear and gadgets that make life better with his contributions to Fox News and FOX Business starting mornings on “FOX & Friends.” Have a technical question? Get Kurt’s free CyberGuy newsletter, share your voice, a story idea or comment on CyberGuy. with.
avots
October 12, 2025
iPhone 18 Pro Rumored to Have These 6 New Features

While the iPhone 18 Pro and iPhone 18 Pro Max are still almost a year away, there are already rumors of some new features and changes for the devices.

Below, we recap some of the early iPhone 18 Pro rumors so far.

Minor Dynamic Island

The standard iPhone 18, iPhone 18 Pro, and iPhone 18 Pro Max will be equipped with a slightly smaller Dynamic Island, but the devices will not feature Face ID under the screen, according to the Instant Digital Weibo Account.

There have been conflicting rumors about whether the iPhone 17 Pro models would have a smaller Dynamic Island, but its size has not changed. Now, the rumor is back on the table for the iPhone 18 series, and there’s a good chance it’s true this time, as it would be a stepping stone towards the rumored all-glass 20th Anniversary iPhone.

Under-screen Face ID, however, is no longer expected until iPhone 19 Pro models or later.

Instant Digital has about 1.5 million followers on Weibo, a popular Chinese social media platform. The account has accurately leaked information about future Apple products in the past, such as the iPhone 17 Pro models with a vapor chamber cooling system, the yellow finish for the iPhone 14 and iPhone 14 Plus, and the Apple Watch Ultra 2’s Titanium Milanese Loop. However, the account doesn’t have a perfect track record and some of the other iPhone 17 rumors it has shared in recent months have not panned out. achieved.

‘Translucent’ MagSafe area

Overall, the iPhone 18 Pro models will feature a similar design to the iPhone 17 Pro models, according to digital chat stationa previously accurate leaker with more than three million followers on the Chinese social media platform Weibo.

The leaker said the devices will have the same rear camera system design as the iPhone 17 Pro models, with a “plateau” housing three lenses in a triangular arrangement. They also expect the iPhone 18 Pro and iPhone 18 Pro Max to have the same 6.3-inch and 6.9-inch screen sizes used since the iPhone 16 Pro and iPhone 16 Pro Max.

Notably, the leaker claimed that the Ceramic Shield area on the back of the iPhone 18 Pro models will feature a “slightly transparent design,” without giving further details.

Perhaps this just means that the Ceramic Shield cutout on the back of the iPhone 18 Pro models will have a more matte appearance than on the iPhone 17 Pro models, but we’ll have to wait for additional rumors to surface for clarification.

Variable Opening

The main 48-megapixel Fusion camera on both iPhone 18 Pro models will offer variable aperture, according to supply chain analyst Ming-Chi Kuo.

With variable aperture, users would be able to control the amount of light that passes through the camera lens and reaches the sensor. The main cameras on all models from iPhone 14 Pro to iPhone 17 Pro have a fixed aperture of ƒ/1.78, and the lens is always wide open and shooting at this wider aperture. With the iPhone 18 Pro models, users would be able to manually change the aperture, according to this rumor.

A variable aperture on the iPhone 18 Pro models should give users greater control over depth of field, which refers to the sharpness of an object in the foreground compared to the background. However, given that iPhones have smaller image sensors due to size constraints, it’s unclear exactly how significant this improvement would be.

None of the iPhone 17 Pro cameras offer variable aperture.

Other rumors

Read our iPhone 18 roundup for more information.

avots

October 12, 2025
Together AI’s ATLAS adaptive speculator delivers 400% inference acceleration when learning from real-time workloads

Companies that scale AI deployments are hitting an invisible performance wall. The culprit? Static speculators who can’t keep up with changing workloads.

Speculators are smaller AI models that work alongside large language models during inference. They prepare several tokens in advance, which the main model checks in parallel. This technique (called speculative decoding) has become essential for companies trying to reduce inference costs and latency. Instead of generating one token at a time, the system can accept multiple tokens at once, drastically improving throughput.

Together AI today announced research and a new system called ATLAS (AdapTive-LeArning Speculator System) that aims to help companies overcome the challenge of static speculators. The technique provides a self-learning inference optimization capability that can help deliver inference performance up to 400% faster than a base level of performance available in existing inference technologies such as vLLM.

The company that had its beginning in 2023, has focused on optimizing inference on your enterprise AI platform. At the beginning of this year the company raised US$305 million as adoption and customer demand grew.

“The companies we work with generally, as they grow, see changes in workloads and then don’t see as much acceleration in speculative execution as they once did,” Tri Dao, chief scientist at Together AI, told VentureBeat in an exclusive interview. “These speculators often don’t perform well when their workload domain starts to change.”

The workload diversion problem no one talks about

Most speculators in production today are “static” models. They are trained once on a fixed dataset representing expected workloads and then deployed without any adaptability. Companies like Meta and Mistral ship pre-trained speculators alongside their core models. Inference platforms like vLLM use these static speculators to increase throughput without changing the quality of the output.

But there is a problem. When a company’s use of AI evolves, the static speculator’s accuracy plummets.

“If you are a company that produces coding agents and most of your developers write in Python, suddenly some of them switch to writing Rust or C, then you see the speed starts to slow down,” Dao explained. “The speculator has a mismatch between what he was trained in and what the actual workload is.”

This shift in workload represents a hidden tax on AI scaling. Companies either accept performance degradation or invest in recycling custom speculators. This process only captures a snapshot in time and quickly becomes outdated.

How Adaptive Speculators Work: A Dual Model Approach

ATLAS uses a dual speculator architecture that combines stability with adaptation:

The static speculator – A heavy model trained on large data provides consistent baseline performance. Serves as a “speed floor.”

The adaptive speculator – A lightweight model continuously learns from live traffic. It specializes in emerging domains and usage patterns.

The conscious controller of trust – An orchestration layer dynamically chooses which speculator to use. Adjusts speculation “look ahead” based on trust scores.

“Before the adaptive speculator learns anything, we still have the static speculator to help provide the speedup at the beginning,” Ben Athiwaratkun, AI scientist at Together AI, explained to VentureBeat. “When the adaptive speculator becomes more confident, the speed increases over time.”

The technical innovation lies in balancing acceptance rate (how often the target model agrees on drafted tokens) and draft latency. As the adaptive model learns from traffic patterns, the controller relies more on the light speculator and extends the look-ahead. This increases performance gains.

Users do not need to adjust any parameters. “On the user side, users do not need to turn any knobs,” Dao said. “For our part, we rotate these knobs for users to adjust to a setting that achieves good acceleration.”

Performance that rivals custom silicon

Together AI tests show that ATLAS achieves 500 tokens per second on DeepSeek-V3.1 when fully adapted. Most impressively, these numbers on Nvidia B200 GPUs match or exceed specialized inference chips like Groq’s custom hardware.

“Software and algorithmic improvement are able to bridge the gap with truly specialized hardware,” Dao said. “We were seeing 500 tokens per second on these huge models that are even faster than some custom chips.”

The 400% speedup the company claims for inference represents the cumulative effect of Together’s Turbo optimization suite. FP4 quantization provides 80% speedup over the FP8 baseline. Static Turbo Speculator adds another 80-100% gain. The adaptive system is at the top. Each optimization combines the benefits of the others.

Compared to standard inference engines like vLLM or Nvidia’s TensorRT-LLM, the improvement is substantial. Together, the AI compares the strongest baseline between the two for each workload before applying speculative optimizations.

The memory-computation tradeoff explained

Performance gains come from exploiting a fundamental inefficiency in modern inference: wasted computational power.

Dao explained that normally during inference, much of the computational power is not fully utilized.

“During inference, which is actually the dominant workload nowadays, you mainly use the memory subsystem,” he said.

Speculative decoding trades idle computation for reduced memory access. When a model generates one token at a time, it becomes memory bound. The GPU is idle while waiting for memory. But when the speculator proposes five tokens and the target model checks them simultaneously, computation utilization increases while memory access remains approximately constant.

“The total amount of computation to generate five tokens is the same, but you only needed to access the memory once instead of five times.” Dao said.

Think of it as a smart cache for AI

For infrastructure teams familiar with traditional database optimization, adaptive speculators work like an intelligent caching layer, but with a crucial difference.

Traditional caching systems like Redis or memcached require exact matches. You store exactly the same query result and retrieve it when the specific query is executed again. Adaptive speculators work differently.

“You can see this as a clever way of caching, not by caching exactly, but by discovering some patterns you see,” Dao explained. “Broadly speaking, we’re looking at you working with similar code, or working with similar code, you know, controlling computation in a similar way. We can then predict what the big model will say. We just get better and better at predicting it.”

Instead of storing exact answers, the system learns patterns from how the model generates tokens. It recognizes that if you are editing Python files in a specific codebase, certain token sequences become more likely. The speculator adapts to these patterns, improving his predictions over time without needing identical data.

Use cases: RL training and evolving workloads

Two business scenarios particularly benefit from adaptive speculators:

Reinforcement Learning Training: Static speculators quickly fall out of alignment as policy evolves during training. ATLAS continuously adapts to changes in policy distribution.

Evolving workloads: As companies discover new AI use cases, the composition of the workload changes. “Maybe they started using AI for chatbots, but then they realized, hey, it can write code, so they started switching to code,” Dao said. “Or they realize that these AIs can actually call up tools and control computers and do accounting and things like that.”

In a flutter session, the adaptive system can specialize for the specific codebase being edited. These are files not seen during training. This further increases acceptance rates and decoding speed.

What this means for businesses and the inference ecosystem

ATLAS is now available on Together AI’s dedicated endpoints as part of the platform at no additional cost. The company’s more than 800,000 developers (compared to 450,000 in February) have access to optimization.

But the broader implications go beyond a supplier’s product. The shift from static to adaptive optimization represents a fundamental rethinking of how inference platforms should work. As companies deploy AI across domains, the industry will need to move beyond once-trained models toward systems that continually learn and improve.

Together AI has historically released some of its research techniques as open source and collaborated on projects like vLLM. Although the fully integrated ATLAS system is proprietary, some of the underlying techniques may eventually influence the broader inference ecosystem.

For companies looking to lead in AI, the message is clear: adaptive algorithms on commodity hardware can match custom silicon at a fraction of the cost. As this approach matures across the industry, software optimization increasingly outperforms specialized hardware.

avots

October 12, 2025
Scientists just took a giant step towards expanding nuclear fusion: ‘What we’ve done here is the beginning of what is still a long journey’

A team of MIT researchers thinks they may have lowered one of the main barriers to achieving large-scale nuclear fusion—putting us one step closer to making an abundant form of energy a reality.

By harnessing the same processes that power stars, we would have access to a clean, safe and virtually unlimited source of energy. Scientists have built reactors to try to tame fusion, one of the most explored being the tokamak. Essentially a donut-shaped tube that uses strong magnets to confine the plasma needed to power fusion reactions, the tokamak has shown great potential. But to fully realize this, scientists must first navigate the potential pitfalls this energy brings with it, including how to slow down a fusion reaction once it’s underway.

That’s where the new search enter: Using a combination of physics and machine learning, researchers predicted how the plasma inside a tokamak reactor would behave given a set of initial conditions—something researchers have long puzzled over (after all, it’s difficult to look inside a fusion reactor mid-run). The article was published Monday in Nature Communications.

“For fusion to be a useful energy source, it will have to be reliable,” said Allen Wang, lead author of the study and graduate student at MIT. MIT News. “To be reliable, we need to be good at managing our plasmas.”

With great power comes great risk

When a tokamak reactor is fully functioning, the plasma stream inside it can circulate at speeds of up to about 100 kilometers per second and at temperatures of 180 million degrees Fahrenheit (100 million degrees Celsius). That’s hotter than the Sun’s core.

If the reactor has to be shut down for any reason, operators begin a process to “reduce” the plasma current, slowly de-energizing it. But this process is complicated and the plasma can cause “scratches and scars inside the tokamak – minor damage that still requires considerable time and resources to repair,” the researchers explained.

“Uncontrolled plasma terminations, even during deceleration, can generate intense heat fluxes damaging the internal walls,” explained Wang. “Often, especially with high-performance plasmas, slowdowns can actually drive the plasma closer to some instability limits. So it’s a delicate balance.

In fact, any misstep in the operation of fusion reactors can be costly. In an ideal world, researchers would be able to run tests on working tokamaks, but because fusion is not yet efficient, running one of these reactors is incredibly expensive, and most facilities will only use them a few times a year.

Looking into the wisdom of physics

For their model, the team found a delightfully clever method to overcome limitations in data collection – they simply went back to the fundamental rules of physics. They paired their model’s neural network with another model that describes plasma dynamics, and then trained the model with data from the TCV, a small experimental fusion device in Switzerland. The dataset included information about variations in the initial temperature and energy levels of the plasma, as well as during and at the end of each experimental run.

From there, the team used an algorithm to generate “trajectories” that showed reactor operators how the plasma would likely behave as the reaction progressed. When they applied the algorithm to real TCV runs, they found that following the model’s “trajectory” instructions was perfectly capable of guiding operators to safely slow down the device.

“We’ve done this several times,” Wang said. “And we did things much better across the board. So we had statistical confidence that we improved things.”

“We are trying to address the scientific questions to make fusion routinely useful,” he added. “What we’ve done here is the beginning of a journey that’s still long. But I think we’ve made some good progress.”

avots

October 12, 2025
The Vince Staples Show’s Frenetic Season 2 Trailer Promises More Surreal Shenanigans

Get ready for more surreal antics from Long Beach rapper Vince Staples in the highly anticipated second season of his Netflix dramedy The Vince Staples Show arrives on November 6th. The streamer shared the season 2 trailer today, which promises violent action, awkward comedy, and an anguished Vince struggling to overcome his traumatic past.

Calling Vince’s 2025 “quiet” would be an understatement. After their intense activity over the past two years, including albums like last year’s Dark times and 2022 Ramona Park broke my heartacting in roles White men can’t jump and the first season of The Vince Staples Showand a walk, Black in AmericaLast year, Vince took a break from the spotlight while filming his show.

But now that the show is returning, perhaps it’s a sign that he intends to return to public life – after all, he’s had plenty of time to record new music and was never able to host the Limbo Beach festival he wanted, so there must be a few things on his agenda. And for those who are missing Vince’s music, there were even guest verses on Ab-Soul and JID projects this year as well.

You can watch the trailer for the second season of The Vince Staples Show above.

avots

October 12, 2025
Taylor Swift, Lady Gaga and a host of other megastars started out as support acts and went onto become WAY more successful than their headliners – but do you know who they opened for?

While it may seem like a long-lost memory for devoted Swifties, the modern world will never forget the Swift-mania caused by the almighty Eras Tour.

With a total of 149 shows across five continents from its beginning in March 2023, to its end in December 2024, the three hour long epic show saw Taylor performing to sell out stadiums night after night.

Similarly Lady Gaga has just departed the UK capital after four sell out nights at London’s O2 arena for her spectacular Mayhem Ball Tour.

Meanwhile last summer saw both Sabrina Carpenter and Noah Kahan smash their headline slots at BST in Hyde Park and Ed Sheeran has released his highly-anticipated eighth studio album Play.

But these artists all had to start somewhere. There was a time when headlining their own show or hitting the top of the charts was but a dream.

Supporting upcoming talent is of the utmost importance in the music industry, but what happens when the supporting act goes on to surpass their headliner.

Would you believe that Rihanna actually started out supporting Gwen Stefani? Or that Sabrina once opened for The Vamps?

Some of the world’s now most famous headlining artists began on much smaller stages as the warm-up acts.

But which opening acts have gone on to dominate the music industry today, and become more successful than the artists they started out supporting?

Your browser does not support iframes.</p> <p> </p> <p>

Taylor Swift

Before the steps of Wembley Stadium were painted with her face, Miss Swift started out supporting American country music band Rascal Flatts in 2006.

From Nashville Tennessee the band recorded six studios albums which were certified platinum or higher but despite their success they have been completely trumped by Swift.

According to statistics from Ticket Source, the popstar ranks first among support acts turned superstars, boasting nine more top 10 albums, 57 more top 10 singles and 688% more Spotify listeners.

She also has nearly 3,000% more award nominations winning 14 Grammy Awards, including four Album Of The Year wins, the most by an artist.

Last week the singer released her twelfth record, The Life Of A Showgirl, after shocking fans with the surprise news of an album release back in August.

Taylor has also found happiness in her personal life after announcing her engagement to her fiancé Travis Kelce on Instagram.

Before the steps of Wembley Stadium were painted with her face, Miss Swift started out supporting American country music band Rascal Flatts in 2006 (Taylor pictured in 2024)

From Nashville Tennessee the band recorded six studios albums which were certified platinum or higher but despite their success they have been completely trumped by Swift (pictured in 2006)

Lady Gaga

The Mayham Ball has just left London after an incredible four sold out dates showcasing her array of hits from over the years.

However, Lady Gaga’s first supporting act was for New Kids on the Block back in 2008, when she was coincidentally the new kid on the block herself.

The American boy band from Massachusetts had success in the late 80s and early 90s, selling over 80 million records worldwide.

Yet the support of Gaga’s little monsters, as her fans are known, has taken the Poker Face star to dizzy levels of fame selling an estimated 124 million records and being ranked among the greatest artists in history by Billboard and Rolling Stone.

Her search interest before and after her support slot with the band has increased by 2408% while there is a 4248% difference in award nominations for the two artists.

Last year her track Die With A Smile with Bruno Mars was Spotify’s most-shared song while her new song The Dead Dance features in the Netflix series Wednesday.

Aside from singing, Lady Gaga has also carved out a successful acting career with starring roles in A Star Is Born, House Of Gucci and Joker: Folie a Deux.

Her hit song Shallow from A Star Is Born also won an Oscar for Best Original Song in 2019.

Lady Gaga’s first supporting act was for New Kids on the Block back in 2008, when she was coincidentally the new kid on the block herself (pictured performing in London last month)

The American boy band from Massachusetts had success in the late 80s and early 90s, selling over 80 million records worldwide (pictured 2008)

Ed Sheeran

Ed Sheeran is one of just 39 artists to have achieved Diamond status from the Recording Industry Association of America (RIAA) on multiple songs.

But before his rise to chart stardom the songwriter was the support for Snow Patrol on their 2012 tour.

The Northern Irish-Scottish band are most notably known for their hit single Chasing Cars, yet since touring with the band Ed has gone on to claim eight more top 10 singles and five more top 10 albums than the headline act.

He has also achieved 416 more award nominations and 86 million more monthly Spotify listeners.

He released his eighth studio album Play this month and is heading on an arena tour later this year.

However while Snow Patrol helped catapult Ed to fame, the Shape Of You singer repaid the favour after match making band member Johnny McDaid with his long-term partner Courteney Cox.

The pair met at one of the Friends star’s parties back in 2013 where Ed brought Johnny as his plus one. The pair have been together ever since.

But before Ed Sheeran’s rise to chart stardom the songwriter was the support for Snow Patrol on their 2012 tour (pictured in 2022)

The Northern Irish-Scottish band are most notably known for their hit single Chasing Cars, yet since touring with the band Ed has gone on to claim eight more top 10 singles and five more top 10 albums than the headline act (pictured in 2012)

Katy Perry

Before the success of California Girls, Katy Perry supported No Doubt on their 2009 tour.

The American rock band who launched Gwen Stefani had mass success with their singles Don’t Speak and Just A Girl.

And watching them perform clearly proved useful for Katy, who has gone on to become one of the best-selling artists of all time.

She has sold over 151 million records with massive hits including I Kissed A Girl, Hot n Cold and Firework.

She has won five American Music Awards, 14 People’s Choice Awards and a Brit.

After spending some time out of the spotlight following the birth of her daughter Daisy, four, who she shares with her ex Orlando Bloom, she is currently on the road for The Lifetimes Tour.

Before the success of California Girls, Katy Perry supported No Doubt on their 2009 tour (pictured in July 2025)

The American rock band who launched Gwen Stefani had mass success with their singles Don’t Speak and Just A Girl (pictured in 2000)

The Weeknd

Known for his hits Blinding Lights and Save Your Tears, The Wknd began as a support act for Florence + The Machine when he joined on her 2012 tour.

Florence achieved a number one hit with her 2012 banger Spectrum (Say My Name) with Calvin Harris and was the first British female to headline Glastonbury in the 21st century.

And she helped propel The Weeknd, real name Abel Tesfaye, to fame with the singer-songwriter since achieving four Grammy Awards and 20 Billboard Music Awards.

He has sold over 75 million records putting him up there with the best-selling artists of all time.

His record After Hours went on to become the most streamed R&B album in history and his After Hours til Dawn Tour set the record for the highest-grossing R&B tour.

Known for his hits Blinding Lights and Save Your Tears, The Wknd began as a support act for Florence + The Machine when he joined on her 2012 tour (pictured last month)

Florence achieved a number one hit with Spectrum (Say My Name) with Calvin Harris in 2012 and was the first British female to headline Glastonbury in the 21st century (pictured in 2024)

Billie Eilish

Billie was also a protegee of Florence’s after supporting her on her High As Hope Tour in late 2018 until January 2019.

But it wasn’t long before the young star’s career soared after the release of her debut album When We All Fall Asleep, Where Do We Go?

The record debuted top of the US Billboard 200 and UK Albums Chart and was one of the year’s best-selling albums as well as contributing to her to win Apple Music’s Artist of the Year in 2019.

She also headlined Glastonbury in 2022.

The star secured the Apple Music accolade once again two years later due to her ‘extraordinary impact throughout 2024’ which saw her bag an Academy Award and two Grammys for her contribution to the Barbie soundtrack.

Billie penned the emotional song What Was I Made For Her for the award winning film by Greta Gerwig before releasing her third album Hit Me Hard and Soft.

She also collaborated with Charli XCX for guess, the song of the summer from the star’s Brat album and is currently on a world tour.

On top of all of that, the Ocean Eyes singer has been won two Oscars, nine Grammy awards, two Golden Globes and has a 12400% Instagram follower difference to Florence.

Billie was also a protegee of Florence’s after supporting her on her High As Hope Tour in late 2018 until January 2019 (pictured at the Grammy’s in 2025)

Rihanna

Ri-Ri had strong female footsteps to follow in after learning from the best when she supported Gwen Stefani back in October 2005.

The Sweet Escape hitmaker had a young Rihanna join her on her first solo tour following her departure from No Doubt.

Soon after Rihanna released A Girl Like Me which featured the singer’s first number one hit SOS.

Since then the star has released eight studio albums, earning nine Grammy Awards, twelve Billboard Music Awards and 13 American Music Awards.

Her last album Anti came out in 2016 and since Rihanna has prioritised her family life, welcoming three children with her partner ASAP Rocky.

She created a viral moment after announcing her pregnancy with her second child during the Super Bowl half time show in 2023.

Meanwhile Rihanna and Gwen joined forces once again after the Umbrella singer served as a key advisor for the coaches of The Voice in 2015.

Ri-Ri had strong female footsteps to follow in after learning from the best when she supported Gwen Stefani back in October 2005 (pictured at the Superbowl in 2023)

The Sweet Escape hitmaker had a young Rihanna join her on her first solo tour following her departure from No Doubt (pictured in 2024)

Noah Kahan

From growing up on a tree farm in Vermont, Noah first opened for American singer-songwriter Ben Folds back in 2017.

Ben found fame as the frontman and pianist of the alternative rock trio Ben Folds Five from 1993 until 2000 when he went solo.

And despite his success in the US, he has been far surpassed by Noah who blew up when his hit Stick Season went viral on TikTok and triggered his mainstream commercial breakthrough.

His album of the same name led to his nomination for Best New Artist at the Grammy’s in 2023 and also earnt him a Billboard Music Award.

Following a double run of The Stick Season Tour, he also played Glastonbury this year and headlined BST in Hyde Park.

Addressing his biggest crowd ever he said: ‘I remember my mum driving me home from shows when there would be 20 people in the room and now there are 60,000 of you.

From growing up on a tree farm in Vermont, Noah first opened for American singer-songwriter Ben Folds back in 2017 (pictured in July)

Ben found fame as the frontman and pianist of the alternative rock trio Ben Folds Five from 1993 until 2000 when he went solo (pictured in 2008)

Sabrina Carpenter

She gained a whole new audience after supporting Taylor Swift on her Eras Tour in 2023 and 2024.

However prior to the Taste hitmakers tour with Taylor, she started out supporting The Vamps in 2017.

The British boyband, who are currently on a hiatus to pursue solo projects, had their first number one on the UK Albums Chart with Night and Day in 2017 and are the first band to headline London’s O2 arena five years in a row.

However the four piece haven’t come close to Sabrina’s success after achieving multi-platinum singles and topping the US Billboard 200 with her sixth studio album Short N’ Sweet.

At this year’s Grammy’s she picked up two awards for Best Pop Vocal Album and Best Pop Solo Performance for Espresso.

This summer she headlined London’s Hyde Park for two nights at BST and her most recent record Man’s Best Friend has also soared to the top of the charts, earning her another number one single with Manchild.

Prior to the Taste hitmakers tour with Taylor Swift, she started out supporting The Vamps in 2017 (pictured in March)

The British boyband, who are currently on a hiatus to pursue solo projects, had their first number one on the UK Albums Chart with Night and Day in 2017 (pictured in 2014)

Chappell Roan

From Pink Pony Club to HOT TO GO! Chappell is has taken the world by storm with her belting vocals and extravagant outfits.

However prior to her mainstream success she joined Australian singer Vance Joy on his 2017 tour.

If you are unfamiliar with the name Vance Joy you will most definitely be familiar with his 2013 hit Riptide which entered the top ten of the charts in the UK, US and Australia.

However Chappell has reached new levels of fame after her single Good Luck, Babe! became a global top-five single last year.

Since then several tracks from her 2023 album The Rise And Fall of a Midwest Princess have gained popularity and at the Grammy’s this year she took home the accolade for Best New Artist.

She also headlined Reading and Leeds Festival and Primavera Sound during the summer.

Prior to her mainstream success Chappell Roan joined Australian singer Vance Joy on his 2017 tour (pictured in February)

If you are unfamiliar with the name Vance Joy you will most definitely be familiar with his 2013 hit Riptide which entered the top ten of the charts in the UK, US and Australia (pictured in 2022)

avots

October 12, 2025