Tokenmaxxing was the hottest trend in Silicon Valley earlier this year, with CEOs encouraging employees to push AI usage as far as it would go. Then the bill came due. Uber reportedly blew through its annual AI budget in a few months, some companies cut Claude licenses for parts of their org, and Meta killed its internal leaderboard. This tension between […]
Whether you're playing poker against a single opponent or find yourself in a bidding war over a home purchase with another prospective buyer, you are operating under conditions of imperfect information. You know what cards you're holding in the poker game, and you also know how much above the home's asking price you can afford, but you don't know your opponent's hand in the card game or how high the other home buyer is willing to go.
What happens when you give AI coding agents a lab full of robotic arms, some compute resources, and a “generous token budget” for teaching the robots various tasks? The agents can apparently figure out a training regimen that teaches the robots to successfully cut zip ties and even insert GPUs into thin sockets on motherboards.
That glimpse into how AI can act in a fully autonomous way to automate robot training was made possible by a new agent harness framework—software that wraps around AI models to enable their use of various tools while also providing capabilities such as memory, context, constraint, and feedback loops. That agentic harness, called ENPIRE, was developed by robotics researchers at the Nvidia GEAR (Generalist Embodied Agent Research) lab alongside collaborators from Carnegie Mellon University in Pittsburgh and the University of California, Berkeley.
“A part of our NVIDIA GEAR lab now self-improves tirelessly overnight,” wrote Jim Fan, director of AI at NVIDIA, in a LinkedIn post. “We just read the reports in the morning.”
French President Macron and Indian PM Modi raised alarms at the G7 summit that the U.S. could cut off access to American AI overnight — a fear the Anthropic blackout just made real.
Late last week, Anthropic took its new Claude Fable 5 and Mythos 5 AI models offline following a United States government export-control directive barring “any foreign national” from using the services. The company has been in talks with the White House since Friday but has yet to secure an agreement that would allow it to reinstate the offerings.
Since Mythos debuted in April, Anthropic has claimed—and warned—that the model has advanced capabilities for not only finding software vulnerabilities to help defenders patch them, but also figuring out ways to exploit them that could be used by bad actors. Anthropic itself noted this double-edged sword in its launch of Mythos 5 and Claude Fable 5. “A great deal of advanced usage of AI models is dual use: the same queries that are beneficial in the hands of cybersecurity professionals and biology researchers could be dangerous if available to malicious actors,” the company wrote in a blog post last week.
With this in mind, the company initially released a version called Mythos Preview to a select consortium as part of a working group known as Project Glasswing. Mythos 5 was also privately released to this group last week, while Claude Fable 5, which is a Mythos-grade model, was released to the general public with specific blocks on its ability to give responses to questions about biology and cybersecurity.
In a groundbreaking advancement at the intersection of oncology and artificial intelligence, researchers have developed a novel preoperative AI model designed to estimate cancer-specific mortality in patients diagnosed with nonmetastatic kidney cancer. This pioneering tool represents a remarkable leap forward in personalized medicine, potentially transforming how clinicians evaluate risks and tailor treatment strategies prior to surgery. The study, led by Larcher, Traverso, Scuri, and collaborators, promises to refine prognosis estimations with unprecedented precision, offering a beacon of hope for patients and healthcare providers alike.
Nonmetastatic kidney cancer presents unique challenges in clinical management, primarily due to its heterogeneous nature and variable outcomes. Traditional prognostic models have often fallen short in capturing the complexity and the subtle nuances influencing cancer-specific mortality. The newly developed AI system leverages vast datasets and sophisticated algorithms to decode patterns that human analysis alone cannot easily discern. By integrating diverse clinical variables and imaging data, the model enhances the predictive accuracy of survival outcomes, facilitating more informed decisions regarding the urgency and extent of surgical intervention.
One of the pivotal innovations underpinning this AI model is its capacity to synthesize multiple dimensions of patient data. Unlike conventional prognostic calculators, which might emphasize isolated factors such as tumor size or grade, this model incorporates a broad spectrum of preoperative indicators. These include patient demographic profiles, detailed tumor characteristics derived from radiologic imaging, and laboratory markers. The model’s architecture employs deep learning techniques to weigh these factors dynamically, adapting to subtle interdependencies and delivering a nuanced mortality risk estimate that is tightly aligned with real-world clinical outcomes.
The implications of this model extend beyond mere risk stratification. By providing a more precise assessment of cancer-specific mortality likelihood before surgical intervention, clinicians can better balance the benefits of aggressive treatment against potential complications and quality-of-life considerations. This is especially vital in cases where surgical morbidity may be significant, or where alternative treatments and surveillance might offer comparable prognostic advantages. Enhanced preoperative prognosis not only informs surgical planning but also supports better patient counseling, setting realistic expectations grounded in individualized risk profiles.
Technically, the model was trained and validated using extensive multicentric datasets, drawn from diverse patient populations to ensure widespread applicability and robustness. The algorithm underwent rigorous cross-validation procedures, testing its predictions against known survival outcomes to eliminate biases and confirm reliability. Importantly, the model’s interpretability was also prioritized, enabling clinicians to understand which variables contributed most significantly to mortality risk predictions, thereby fostering trust and facilitating integration into clinical workflows.
The methodology incorporated advanced imaging analytics alongside conventional clinical parameters. Radiomic features extracted from preoperative CT scans were a major component, providing detailed textural and morphological insights that correlate with tumor behavior. These imaging biomarkers, when coupled with biochemical data such as serum creatinine and hemoglobin levels, painted a comprehensive portrait of both tumor aggressiveness and patient physiological status. This multimodal data fusion became a cornerstone of the AI’s ability to deliver personalized prognostic assessments.
In terms of deployment, the AI model is designed to be user-friendly and seamlessly integratable into existing hospital information systems. Its interface allows surgeons, oncologists, and multidisciplinary teams to input clinical data and receive risk assessments in real-time. Such accessibility enables rapid decision-making, potentially influencing decisions about the timing of surgery, the scope of resection, and adjuvant therapy planning. Moreover, the model’s adaptability means it can incorporate new data as clinical understanding and treatment modalities evolve.
A critical aspect of this research lies in the ethical deployment of AI in clinical settings. The authors address concerns regarding algorithmic transparency, patient privacy, and the risk of over-reliance on machine predictions. They emphasize that the AI model is meant to augment, not replace, clinical judgment, serving as an advanced decision-support tool. Continuous monitoring, feedback mechanisms, and physician input are integral components of the model’s operational framework, ensuring that clinical expertise remains central.
The timing of this innovation is particularly significant given the increasing incidence of kidney cancer worldwide and the ongoing quest for precision oncology solutions. Current staging systems and prognostic indices, while valuable, often underestimate individual variability and outcomes seen in everyday clinical practice. By harnessing the predictive power of AI, this model addresses a critical unmet need, offering a reliable compass to navigate the complex risk landscape of nonmetastatic kidney cancer.
Looking forward, the researchers envision expanding the scope of their AI system. Integration with genomic and molecular profiling data could further enhance prognostic accuracy, capturing the genetic underpinnings and heterogeneity of kidney tumors at a granularity impossible with imaging and clinical data alone. Such multi-omic AI-driven platforms could revolutionize not only prognosis but also treatment selection, ushering in an era of truly personalized cancer care.
Operability in diverse clinical settings also forms a core consideration in the model’s development. The use of widely available imaging modalities and routinely collected clinical data ensures that the technology can be disseminated globally, including in resource-constrained environments. This raises the tantalizing possibility of democratizing high-quality kidney cancer risk assessments, bridging gaps in healthcare disparities that currently limit access to advanced oncologic prognostication.
Another dimension includes the potential of the AI model to guide clinical trials and research. By precisely stratifying patients based on their predicted cancer-specific mortality risk, clinical investigators can design trials with more homogenous cohorts, enhancing statistical power and facilitating the discovery of targeted therapies. Moreover, the model could identify patients who might benefit most from novel interventions, optimizing resource allocation and accelerating therapeutic developments.
In essence, the AI-driven preoperative model is a testament to the transformative power of machine learning in medicine. It encapsulates the promise of personalized, data-driven healthcare that adapts to individual patient profiles rather than relying solely on broad population averages. As this technology moves from development to clinical adoption, it heralds a future where kidney cancer prognosis is not a matter of chance but an informed, precise science guiding bespoke treatment pathways.
The broader oncology community has greeted this development with enthusiasm, recognizing its potential to reshape risk assessment paradigms. The study stimulates a conversation about the role of AI in preoperative oncology, highlighting both opportunities and challenges inherent in integrating advanced computational tools into clinical decision-making. It serves as a clarion call to clinicians, researchers, and technology developers to collaborate and refine these early innovations for maximal patient benefit.
In conclusion, the work by Larcher, Traverso, Scuri, and their colleagues marks a decisive step toward integrating artificial intelligence into the clinical management of nonmetastatic kidney cancer. Their AI model exemplifies how cutting-edge technology, grounded in robust data and sophisticated analytics, can empower clinicians with actionable insights that elevate patient care. As machine learning continues to evolve, such tools are poised to become indispensable allies in the global fight against cancer.
Subject of Research:
Artificial Intelligence application for prognostic prediction of cancer-specific mortality in nonmetastatic kidney cancer patients.
Article Title:
A preoperative Artificial Intelligence model to estimate cancer-specific mortality in nonmetastatic kidney cancer patients.
Article References:
Larcher, A., Traverso, A., Scuri, P. et al. A preoperative Artificial Intelligence model to estimate cancer-specific mortality in nonmetastatic kidney cancer patients. Nat Commun (2026). https://doi.org/10.1038/s41467-026-74419-9
With Firefly Graph, users can connect multiple tools in visual workflow, with each “node” performing a specific task before passing its output to the next node. This gives creative professionals more control over generated outputs, according to Adobe, and makes it easier to try out ideas by swapping, adjusting or adding components.
For example, a user could start with a text prompt box that connects to a node that generates an image using an AI model from Adobe or third-parties such as Google and OpenAI. Further along the chain, the user could add nodes to remove a background or upscale an image, for instance, before producing an image, video or other asset ready for use.
Changing one aspect, such as adding a reference image or adapting the text prompt, would change the final output.
With so many AI tools available to creative professionals, workflows can get complex and hard to replicate, said Elliot Sedegah, director for strategy and product marketing at Adobe. Firefly Graph provides access to more than 300 different node types, including images, video editing and AI generation tools across Adobe’s portfolio and third-party tools.
“Whether you’re working at a mom-and-pop shop or a larger enterprise, you’re looking for consistency and then bringing that into a workflow so that you’re not hopping in and out of different tools,” he said. “Putting all that together takes massive amount of time, and sometimes it’s very difficult to even know what you did.”
Once created, workflows can be shared across an organization as repeatable processes for other individuals or teams to use. “Think of that rock star creative that you have and the recipes they create: those are now canonized as workflows, as assets, that the rest of the organization can take and reuse over and over again,” said Sedegah.
In addition, while creative professionals are needed to created high quality assets, reusable workflows can be put into the hands of broader teams to create content for large audiences, said Sedegah.
Firefly Graph addresses a challenge that most large creative organizations face, said Lisa Gately, principal analyst at Forrester — namely that their best creative workflows “live inside the heads of a few experts.
“Teams can generate images and video with AI, but reproducing the exact sequences of creative decisions, model selections, edits, and refinements that lead to a high-quality result is difficult and inconsistent. Firefly Graph turns those workflows into reusable assets,” she said.
While other node-based workflows aim to address similar problems, Adobe’s pitch is that Firefly Graph provides customers with the benefit of integration into its product suite.
“Firefly is a full, broader AI creative studio, not just a node-based tool, so [Firefly Graph] is a part of a bigger picture,” said Sedegah. “The strength is having everything in one place with the tools that people know.”
“Where Adobe differentiates is in enterprise integration,” said Gately, with Adobe connecting Firefly Graph to a range of other Adobe tools. Those include Creative Cloud applications; Firefly Boards for ideation; and Firefly Creative Production.
“The workflow becomes part of a broader content supply chain instead of a standalone creation tool,” she said. ”Organizations committed to other tools are unlikely to migrate for a node-based canvas — making a change is about the broader content supply chain.”
Firefly Graph is available now to Adobe Creative Cloud for Enterprise subscribers. Creative Cloud for Enterprise is licensed by seat, requiring custom, volume-based enterprise agreements. Creative Cloud for Enterprise customers receive credits for Firefly Graph. Creative Cloud for teams customers can sign up for a public beta online. Creative Cloud pricing starts at $99.99 per license each month.
This article was updated on June 19 with pricing information from Adobe.
Anthropic has spent much of this week fighting to get its newest AI models back online after the Trump administration abruptly ordered the company to cut access for all foreign nationals, including users inside the US and its own employees, forcing Anthropic to block access to Fable 5 and Mythos 5 for everyone.
"To my knowledge, this is the first time US export controls have been used to control access to an AI model in this way."
The Trump administration has not publicly explained the legal basis for the order, but in a statement on its website, Anthropic said the government cited "national security authorities" to justify "an export con …
The escalating threat posed by rising sea levels has become a defining environmental challenge of the 21st century, with profound implications for densely populated coastal regions worldwide. Recent research sheds new light on this crisis by examining the spatial and temporal evolution of sea level rise under the ongoing forces of climate change. Particularly vulnerable to these changes are Shanghai and other parts of China’s coastal waters, where natural and anthropogenic factors synergize to amplify the risks.
This pioneering study deploys advanced climate modeling combined with detailed regional analysis, revealing that sea level rise in China’s coastal zones exceeds average global projections. The unique convergence of local land subsidence, riverine water discharge variability, and dynamic marine currents exacerbates the vulnerability of these regions. Such localized processes interact with global sea level trends, creating hotspots where the risk is acutely intensified.
Land subsidence, the sinking or settling of the ground surface, has emerged as a critical factor accelerating local sea level rise in coastal China. Driven largely by groundwater extraction and urban development, subsidence lowers the land elevation relative to the rising ocean, effectively increasing flood risk and saltwater intrusion. This phenomenon, when combined with the gradual thermal expansion of seawater and melting ice sheets, underlines the compound nature of sea-level hazards facing mega-cities like Shanghai.
Moreover, the study brings to the forefront the role of river water discharge, an often overlooked element in sea level dynamics. Changes in the volume and timing of freshwater flow from major river systems alter sediment supply and coastal circulation patterns, which have direct impacts on shoreline stability and ecological health. In the Yangtze River estuary, fluctuations in discharge affect sediment deposition that could either reinforce or undermine coastal defenses against the rising tides.
Marine currents along China’s coastline further complicate the picture by influencing regional sea level changes. Currents can redistribute water masses, leading to regional variations in sea level rise that diverge significantly from global averages. Variability in monsoon strength and ocean-atmosphere interactions introduce additional temporal complexity, affecting sea level trends on seasonal to decadal scales.
The implications of these findings are far-reaching. Beyond the increased risk of flooding and coastal erosion, rising seas threaten the socio-economic fabric of communities reliant on coastal infrastructure and ecosystems. Shanghai, a global financial hub with millions residing in low-lying areas, confronts immense challenges in urban planning, disaster preparedness, and climate resilience efforts.
This comprehensive research integrates satellite data, ground observations, and climate projections to afford a high-resolution understanding of these localized sea level dynamics. By unraveling the intricate interplay between climatic and human-induced factors, the analysis offers crucial insights for policymakers and urban planners striving to mitigate and adapt to future sea level scenarios.
Furthermore, the study underscores the necessity of cross-disciplinary collaboration, merging geoscience, hydrology, oceanography, and urban studies. It challenges simplistic global sea level rise metrics by demonstrating that effective mitigation and adaptation require granular regional assessments that account for local geophysical and hydrodynamic conditions.
In the context of climate change mitigation strategies, this research elevates the urgency of addressing greenhouse gas emissions while concurrently managing human activities that exacerbate coastal vulnerability, such as groundwater over-extraction. It suggests that adaptation measures, including coastal defenses, managed retreat, and ecological restoration, must be tailored to localized conditions revealed through detailed spatial-temporal analysis.
Moreover, the insight gained concerning the temporal evolution of sea level rise offers valuable foresight into critical tipping points and thresholds. Understanding when and where acceleration of sea level increase may occur enables more precise targeting of resources and emergency preparedness efforts.
In conclusion, this comprehensive examination of sea level rise in China’s coastal waters highlights a complex mosaic of contributing factors that extend beyond traditional global climate models. The findings serve as a clarion call for integrative and localized climate action to safeguard vulnerable coastal megacities against the rising tides of the future.
Subject of Research: Spatial and Temporal Evolution of Sea Level Rise Under Climate Change
Article Title: Rising tides: Unveiling the spatial and temporal evolution of sea level rise under climate change
A groundbreaking advancement in photonic technology promises to redefine the landscape of medical artificial intelligence (AI) diagnostics, as a research consortium led by Professor Han Zhang at Shenzhen University has unveiled a novel all-fiber photonic AI platform utilizing black phosphorus (BP)-based tunable modulators. This innovative approach, which hinges on photon-based computing, surmounts the intrinsic constraints posed by traditional electronic processors, delivering unprecedented energy efficiency and ultrafast performance crucial for real-time clinical decision-making.
The prevailing paradigm in medical AI, deeply reliant on electronic architectures such as GPUs, faces severe limitations related to energy consumption, heat dissipation, and processing latency. These factors collectively impede the integration of AI into time-sensitive diagnostic workflows and exacerbate systemic carbon footprints. Photonic computing, with its reliance on photons rather than electrons, offers dramatically accelerated data transmission speeds and parallelism through wavelength multiplexing, alongside near-zero heat generation. Until now, however, the practical implementation of photonic AI had been hampered by the bulkiness, inefficiency, and fabrication challenges synonymous with conventional optical modulators.
Professor Zhang’s team has ingeniously engineered an atomically thin van der Waals heterostructure by stacking black phosphorus and molybdenum disulfide (MoS2), leveraging the extraordinary optoelectronic properties of these 2D materials. This heterostructure is meticulously integrated onto a microfiber knot resonator (MKR), an ultra-fine looped optical fiber with a diameter narrower than a human hair. This configuration significantly intensifies the interaction between light and material, enabling modulation of the refractive index with minimal applied voltage, which shifts the light’s wavelength to encode computational information with remarkable precision and energy economy.
The fabrication process commences with mechanical exfoliation to obtain pristine BP/MoS2 bilayers, followed by a dry transfer technique to position these heterostructures onto the MKR. The MKR’s torsional geometry enhances evanescent field coupling, ensuring that the optical mode overlaps maximally with the active layered materials. To broaden the modulator’s linear response and suppress nonlinear distortions, the researchers incorporated a Ring-Assisted Mach–Zehnder Interferometer (RAMZI) design. This integration extends the operational voltage range, ensuring fidelity and robustness essential for medical-grade AI computations.
The team transcended single-device innovation by architecting a full all-fiber photonic neural network (PNN). By deploying two RAMZI modulators in tandem with a highly sensitive photoreceiver within a sophisticated time-division multiplexing schema, they realized a closed-loop, coherent computing unit capable of executing neural network operations entirely in the optical domain. This milestone substantiates the transition of photonic neural networks from theoretical constructs to viable, deployable platforms for clinical diagnostics.
Clinical performance validation was conducted in collaboration with esteemed hospitals. The PNN demonstrated expert-level diagnostic accuracy in two stringent medical tasks: detecting retinal detachment from B-scan ultrasound imagery and diagnosing hepatocellular carcinoma (HCC) using multiphase liver computed tomography (CT) scans. Diagnostic accuracy and specificity metrics paralleled those of seasoned radiologists, underscoring the system’s potential for integration into frontline healthcare.
Beyond accuracy, the photonic system showcased transformative efficiency advantages. While a conventional NVIDIA A100 GPU requires approximately 85 milliseconds to process a liver CT study, the photonic platform accomplished the same in a mere 0.8 milliseconds. Additionally, the energy expenditure per operation plummeted to an astonishing 0.608 femtojoules, a staggering 246-fold decrease compared to electronic counterparts consuming 150 femtojoules. This synergy of speed and efficiency paves the way for deploying AI diagnostics in resource-constrained or mobile medical environments like rural clinics and ambulances, dramatically expanding access to critical health services.
The system’s environmental implications cannot be overstated. Conventional AI’s sizable energy footprint contributes appreciably to healthcare’s carbon emissions. The BP/MoS2-based photonic platform fundamentally disrupts this model by combining ultrafast processing with near-negligible thermal dissipation, forging a sustainable pathway for green AI implementations in medicine. This breakthrough aligns with expanding global commitments to environmentally responsible technologies without sacrificing computational power.
Acknowledging the infancy of current implementations, the authors outline avenues for scaling. Presently limited to a single processing layer realized through two modulators, the system’s complexity and image resolution capacity necessitate expansion. A promising strategy involves wavelength-division multiplexing (WDM), exploiting the device’s ∼30 nm bandwidth resonances to channel up to 40 distinct wavelengths. This parallelism would amplify computational density approximately 40-fold while maintaining the existing clock rate, facilitating thousands of multiplications per network layer—an enhancement critical to tackling high-dimensional medical image data.
Device longevity and stability pose additional challenges, given the sensitivity of BP to environmental degradation. While the MoS2 layer affords interim encapsulation, the team plans industrial collaboration to pioneer large-scale protective coatings via atomic layer deposition of Al2O3 and scalable chemical vapor deposition processes. These advancements will bolster device durability, uniformity, and manufacturability, accelerating clinical adoption and commercial viability.
The study also signifies the maturation of fiber-based photonic neural networks into practical diagnostic platforms rather than experimental novelties. Optical fibers’ inherently low transmission loss (<0.2 dB/km), synergized with the modulator’s efficient 0.25 V·cm electro-optic response and expanded linear range, addresses critical bottlenecks impeding prior photonic computing endeavors. Consequently, the technology promises to empower future applications in drug discovery, genomics, and real-time clinical imaging with unmatched energy footprints.
The pioneering work led by Professor Han Zhang not only foregrounds interdisciplinary collaboration across physics, materials science, and biomedicine but also sets a precedent for industry-academia partnerships catalyzing the commercialization of ultrafast photonic devices poised to transform healthcare AI. As research progresses towards integrated multi-layer architectures and industrial encapsulation, the healthcare sector stands on the cusp of embracing a new generation of diagnostic tools that are faster, greener, and broadly accessible.
This heralds a paradigm shift where medical AI transcends the limitations inherent to electronic systems, embracing photon-driven computing as a sustainable, high-performance alternative that aligns with the twin imperatives of clinical excellence and environmental stewardship.
Subject of Research: People
Article Title: Tunable phosphorene modulators – accelerating medical diagnosis with ultra-efficient photonic platforms
News Publication Date: 28-May-2026
References:
Zhang, H., Liu, Y., Wang, H., Zhu, H., et al. (2026). Tunable phosphorene modulators – accelerating medical diagnosis with ultra-efficient photonic platforms. Opto-Electronic Advances. DOI: 10.29026/oea.2026.250332
Image Credits: Professor Han Zhang of Shenzhen University
Keywords: photonic computing, black phosphorus, molybdenum disulfide, van der Waals heterostructure, microfiber knot resonator, electro-optic modulator, photonic neural network, medical imaging AI, energy efficiency, wavelength-division multiplexing, ultrafast diagnostics, sustainable healthcare
Social media feeds are becoming more customizable as platforms like Threads, Instagram, and TikTok introduce tools that let users directly influence the algorithms powering their recommendations.
An anonymous reader quotes a report from Reuters: Artificial Intelligence will lead to labour shortages, not the replacement of humans, Amazon founder Jeff Bezos predicted in a highly optimistic appearance at the VivaTech technology conference in Paris on Wednesday. Bezos put forward a rosy vision of how technology will help humanity, speaking about projects including his space venture Blue Origin and his new AI startup Prometheus, which is aimed at speeding up physical manufacturing. "I know there's a lot of concern that many people have, including many smart people, that AI is going to make humans redundant and so on," Bezos said. "I totally disagree with this point of view. And I think, in fact, AI is going to create a labor shortage."
Half of Americans fear the rise of AI could put them or someone in their household out of work, a Reuters/Ipsos poll found this month. Bezos, the world's fourth-richest person with a net worth around $250 billion, argued that people have "endless" things to do, and are currently limited by barriers that he said AI would lower. One goal of space exploration is to move polluting industries off Earth, said Bezos, whose Blue Origin aims to compete with trillionaire Elon Musk's SpaceX in rockets. "If space travel gets reliable enough and inexpensive enough, and we can get materials from asteroids and near-Earth objects and the moon, then this garden planet can be returned to its pre-Industrial Revolution state," Bezos said.
Login
