L’allineamento planetario di giugno 2026 continua a regalare uno spettacolo affascinante agli appassionati di astronomia. Anche se il momento culminante del fenomeno è ormai trascorso, Mercurio, Venere e Giove restano ancora visibili vicini tra loro nel cielo della sera, offrendo un’occasione unica per osservare tre pianeti allineati a occhio nudo. Le condizioni di osservazione rimangono […]
O Trump Phone não é o que parece e uma nova desmontagem da iFixit confirma o que muitos suspeitavam. Afinal, o smartphone é quase idêntico ao HTC U24 Pro, lançado em 2024.
I turned off Siri on the Mac years ago and never looked back. Similarly, I found Apple Intelligence so fruitless I never engage with it. But the new Siri AI coming to macOS 27 Golden Gate has at least got me slightly rethinking things.
I'm still early in testing Siri AI, as I've only had access to it in the macOS 27 developer beta for little more than 24 hours. It's also in an early preview state on the dev beta, so there should be lots of runway for improvements before it releases later this year. I don't even know if it's done indexing my files and folders on our review unit M5 MacBook Air and M5 Max MacBook Pro. Unlike on the iOS 27 dev …
iPhone owners are getting real, native AI photo editing for the first time.
The most popular camera in the world just got its first set of serious AI photo editing features, and I don't think any of us are ready.
As far as AI photo editing goes, the new features in iOS 27 are pretty tame compared to what you can do on, say, Google's Pixel phones. But for the iPhone, they represent a tipping point in what the native photos app allows you to do to your photos. I mean memories. I mean, I don't know anymore.
These new features are part of the iOS 27 developer beta right now, so bear in mind that Apple may continue making tweaks to them before they're released to the general public. There are three, or maybe two and a …
Dark matter remains one of the biggest mysteries in science. Scientists cannot see it directly, yet they believe it makes up most of the matter in the universe. Now, a new study suggests that dark matter may gather around supermassive black holes, forming large invisible clouds that influence their surroundings. Researchers from Virginia Tech have […]
Un gruppo di ricercatori guidato da Richard Levenson (UC Davis) ha dimostrato che i piccioni comuni (Columba livia) possiedono capacità di riconoscimento visivo e astrazione tali da poter identificare tumori maligni al seno con precisione millimetrica. Addestrati tramite condizionamento operante, i volatili hanno analizzato biopsie e microcalcificazioni mammografiche, raggiungendo un'accuratezza dell'85% sui singoli casi e del 99% quando le risposte venivano aggregate in gruppo ("flock-sourcing"). I limiti emergono solo di fronte alle masse radiologiche sfumate e prive di contorni netti, dove sia i piccioni che gli algoritmi di intelligenza artificiale addestrati sul loro modello falliscono. La scoperta apre alla possibilità di usare questi animali per validare l'efficacia dei nuovi sistemi di imaging medico, riducendo il carico di lavoro dei radiologi umani.
80 millions : c'est le nombre de requins tués chaque année par la pêche mondiale, selon une étude publiée dans la revue Science. Une pression colossale qui ne reste pas sans réponse de la part de ces prédateurs. Dans le golfe du Mexique, les requins s'adaptent, et pas forcément dans le bon sens.
La Coupe du monde de foot 2026 s’annonce chaude et électrique sur le continent américain. Les prévisions météo des prochains matchs aux États-Unis, au Canada et au Mexique, risquent de compliquer la vie aux joueurs comme aux spectateurs.
Does the fact that we see light from very distant galaxies prove that the one-way speed of light is finite? Dr. Jason Lisle responds to this question. The conventionality thesis plays a critical role in understanding the issue, suggesting that measurements usually assume an isotropic light speed.
Andrew Yang made a list of everything Americans overpay for — housing, food, wireless — and thinks the next startup gold rush is giving that money back.
Astronomers using NASA’s powerful James Webb Space Telescope may have solved one of the biggest mysteries of the early universe. A strange object known as a “little red dot” appears to be a rapidly growing supermassive black hole hidden inside a thick cloud of gas. Little red dots were first discovered by the James Webb […]
Extremely powerful large language models (LLMs) still operate as though they’re typing on a keyboard, processing workloads in a simple left-to-right fashion. But in locally-run, single-user scenarios, this sequential processing can leave graphics processing units (GPUs) and tensor processing units (TPUs) underutilized.
Google is betting that DiffusionGemma can get around this bottleneck. The new experimental open model generates text “exceptionally fast,” creating entire blocks of text simultaneously through diffusion techniques rather than through token-by-token processing. The company says this technique results in 4x faster inference compared to auto-regressive models that rely on sequential processing.
It can also save users money. Technology analyst Carmi Levy noted that existing pay-per-token monetization models “penalize the use of less than optimally efficient AI solutions.”
But DiffusionGemma “could herald a new generation of task-defined, efficient solutions that can enable expanded compute capacity without draining the operations budget,” he said.
A contrast to left-to-right processing
Built on Google’s Gemma 4 family and its Gemini Diffusion research, DiffusionGemma is a 26B mixture-of-experts (MoE) model designed to maximize text output generation.
It essentially shifts how models use hardware, giving processors a larger hunk of work each cycle so it can draft full 256-token paragraphs in sequence. This allows the model to generate text up to 4x faster on GPUs, Google claims. It activates only 3.8B parameters during inference, and, when quantized, can fit within 18GB VRAM on high-end consumer GPUs like Nvidia RTX 5090.
“It upgrades your model inference from a single, sequential typewriter to a massive printing press that stamps the entire block of text simultaneously,” Google research scientists Brendan O’Donoghue and Sebastian Flennerhag wrote in a blog post.
AI image generators begin with pure, random ‘visual noise’ and iteratively refine that into a finalized picture (what’s known as ‘diffusion’); DiffusionGemma applies this same process to text. It does not generate tokens in order, but begins with a “canvas of random placeholder tokens” that it processes in multiple passes, identifying the context tokens it feels are most relevant and using those to refine the rest.
The model has the ability to self-correct, using confidence scoring to re-evaluate tokens in the next pass. “The model iteratively refines its own output, allowing it to evaluate the entire text block at once to fix mistakes in real-time,” O’Donoghue and Flennerhag explained.
DiffusionGemma also has bidirectional attention, they wrote. “Generating 256 tokens in parallel with each forward pass allows every token to attend to all others.” This can be particularly helpful in domains that are non-linear in nature, such as mathematical graphs, code infilling, and in-line editing, they said.
DiffusionGemma is optimized across Nvidia’s hardware stack, making it compatible with consumer setups as well as with high-performance enterprise systems like Hopper and Blackwell.
Because it is released under the Apache 2.0 license, developers can freely use, modify, distribute, and commercialize the software using their preferred tools. It can be run on GPUs or in the cloud through Google Cloud Model Garden or Nvidia NIM, and is available on Hugging Face, GitHub, and vLLM, with support for the open-source library llama.cpp coming soon.
Key use cases
The model is particularly useful in local workflows that are “speed critical,” such as generation of non-linear text structures, and unlocks what Google calls “new patterns of model behavior” like multimodal understanding and generating and rendering code in near real-time.
Levy explained, “DiffusionGemma is particularly well suited for interactive coding and editing where its efficiency allows rapid processing and iterations,” noting that its ability to fit within 18GB of VRAM and its deployability on commonly available local GPUs can potentially benefit customer service-related workloads that lean heavily on real-time interaction and local processing.
“DiffusionGemma also incorporates a thinking mode that is especially adept at problem solving,” he said. For instance, the model was fine-tuned to play Sudoku, a typically challenging task for autoregressive models because each token depends on future tokens. This “rather handily” illustrates the model’s capability to solve more complex problems, Levy noted.
Limitations
Google freely admits that DiffusionGemma is geared to specific workflows, and there are “key trade-offs.”
The model is engineered for small batch size inferencing and low-latency, high-speed generation low-to-medium batch sizes on a “single capable accelerator.”
In high-QPS cloud serving environments, (where infrastructure is designed to handle tens or hundreds of thousands of requests per second with ultra-low latency), DiffusionGemma’s parallel coding “offers diminishing returns,” and can even result in higher serving costs, Google conceded. In addition, its overall output quality is lower than that of standard Gemma 4, which is built for apps demanding maximum quality.
However, Levy noted that while DiffusionGemma “can be less precise than other models in certain workloads,” subsequent refinement cycles could overcome this limitation.
While Google isn’t sharing runtime costs, it’s clear that this is an efficiency play, he added. “When deployed across the kinds of workloads that would optimally benefit from its architecture, DiffusionGemma seems to have the potential to reduce processing overhead and related costs,” he said.
ShinyHunters claims it exploited a critical Oracle PeopleSoft zero-day to compromise more than 100 organizations, including the University of Nottingham, where it says it stole 40GB of student and billing data. "ShinyHunters posted the UK university on its data leak site on Tuesday before publishing the stolen files later that same day, presumably because the school refused to pay the extortion demand," reports The Register. From the report: "University of Nottingham on our leak site is one of the first publicly confirmed incidents," a ShinyHunters spokesperson told us. "We have only just started outreach to affected orgs and are actively looking to reach an agreement with affected orgs." They didn't say when they planned to post the other 100 or so claimed victims.
A Google threat intelligence report published Thursday afternoon corroborated ShinyHunters' claims to have compromised more than 100 organizations. Google said it spotted malicious activity, "consistent with the exploitation of CVE-2026-35273," between May 27 and June 9, and notified more than 100 global orgs "whose IP addresses correlated with potentially vulnerable endpoints." Most of these, we're told, are based in the US and 68 percent are in the higher-education sector. Oracle has released a "patch availability document," but it's unclear whether a patch is currently available.
NASA’s X-59 quiet supersonic research aircraft reached its target speed and altitude for future community overflights for the first time during a flight on Friday, June 12, 2026. The milestone marked the first time the aircraft flew at Mach 1.4 and 55,000 feet — conditions it will fly when gathering community response data to its quiet sonic thump.
NASA/Lori Losey
NASA’s X-59 experimental aircraft reached a major milestone Friday, June 12, flying Mach 1.4 (about 924 mph) and an altitude of 55,000 feet, the conditions required for the aircraft to make future flights critical to its mission.
The X-59 still has months of performance testing ahead, but after those are complete, NASA’s Quesst mission will fly the aircraft over several U.S. communities to collect data on public perception of the quiet sonic thump it will make at supersonic speeds. Those community overflights will include flights at Mach 1.4 and 55,000 feet.
The milestone comes just days after the X-59’s first supersonic flight. That flight showed the aircraft performed as expected at Mach 1.1, but Friday’s mission conditions flight was an even more critical step for NASA.
NASA’s X‑59 eXternal Vision System (XVS) shows Mach 1.4 at 55,030 feet on Friday, June 12, 2026 during the aircraft’s first mission conditions flight, the approximate conditions it will fly when gathering community response data to its quiet “sonic thump.” The milestone marks the first time the aircraft flew at its designed speed and altitude.
NASA/Lori Losey
The X-59 was designed to fly supersonic without causing a loud sonic boom. However, for these early supersonic flights it has been accompanied by a NASA F-15 research aircraft, a traditional supersonic jet that causes booms obscuring any noise the X-59 makes. During upcoming flights, a shock-sensing probe mounted to the F-15 will gather measurements of the X-59’s shock wave signature, an early measure of its supersonic performance.
After the team conducts more tests at a variety of altitudes and conditions to complete envelope expansion, the X-59 will enter the acoustic validation phase of Quesst. During this phase, researchers will thoroughly measure the aircraft’s supersonic acoustic signature — the quiet thump it’s designed to make — to confirm it is performing as intended.
Each flight brings NASA one step closer to flying the X-59 over communities and gathering feedback that could help shape the future of commercial supersonic flight over land.
One of the world’s most active ransomware groups exploited a critical vulnerability in Oracle’s PeopleSoft software suite and used it to target about 100 customers and extort at least one of them to pay up in exchange for not leaking stolen data, researchers said.
The group, tracked as ShinyHunters, had been exploiting the PeopleSoft vulnerability for more than two weeks before Oracle flagged it. CVE-2026-35273, as the vulnerability is tracked, carries a severity rating of 9.8 out of 10, making the former zero-day one of the year’s most critical vulnerabilities to be exploited.
Google’s Mandiant security team said it’s an SSRF (server-side request forgery), a vulnerability that allows attackers to send requests from a susceptible server to systems used by the targeted organization. Oracle said the SSRF is remotely exploitable, and the company has issued a stopgap mitigation but has yet to fully patch the flaw. Google has confirmed that victims are receiving extortion demands.
Login
