SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " benchmark " 260 found 🔀 AI-powered Shuffle
College Jersey Patch Market Finds Its Benchmark
"The market has now been set," one exec told FOS.
Principal Financial Group named World’s Top Disability Inclusive Business
Des Moines-based Principal Financial Group has been recognized as a World’s Top Disability Inclusive Business, based on its performance on the Disability Index, which the company describes as the leading global benchmark for disability inclusion in business. The recognition reflects Principal’s commitment to accessibility, inclusion, and creating opportunities for employees to succeed, and highlights how […]
Why benchmarking clinical LLMs from OpenEvidence, Doximity is complicated
In this edition of AI Prognosis: A conversation about benchmarking leading clinical chatbots, investor view on AI in biopharma, and more.
Clinical chatbots are taking medicine by storm. Should doctors trust them?
As LLMs compete for doctors' attention, some developers say the science of benchmarking their AI tools for safety and accuracy is flawed.
Real Madrid Posts $1.39 Billion Revenue, Record High For Sports Entity
Real Madrid has set a new benchmark in sports finance, reporting a record $1.39 billion in operating revenue for 2025/26, a new high for a sporting entity.
Microsoft is launching its first cybersecurity AI model at half the cost of rivals
MAI-Cyber-1-Flash, paired with GPT-5.4 inside Microsoft's MDASH harness, outperforms models from Anthropic, Google, and OpenAI on a key benchmark
Tech Stocks Tumble on Worries About A.I. Spending and China’s Chip Competition
South Korea temporarily paused trading on Tuesday as its benchmark stock index led losses across the region, closing more than 10 percent lower.
Acxiom Recognized as a World’s Top Disability Inclusive Business
CONWAY, Ark., July 27, 2026--Acxiom®, the connected data and technology foundation for the world’s leading brands, today announced that it has been recognized as a World’s Top Disability Inclusive Business based on its performance on the Disability Index®, the leading global benchmark for disability inclusion in business. Acxiom was recognized in the Advanced category, which reflects strong leadership and the successful integration of disability inclusion practices throughout the organization. P
Geekbench 7 is out with overhauled performance tests for CPU and GPU workloads
Geekbench 7 today, promising benchmarks that more closely reflect how people use their devices across everyday tasks, gaming, and AI workloads.
Geekbench 7 Launches With Redesigned Multi-Core and GPU Benchmarks
Primate Labs today announced the launch of Geekbench 7, the newest version of the popular benchmarking suite for CPU and GPU performance. There's a redesigned multi-core benchmark in Geekbench 7 that's meant to better represent how applications work in the real world. A workload only runs in multi-threaded mode if the task it models runs multi-threaded in real applications.
Top 5 Reasons to Participate in the 2026 Health Benefits Survey This Year
Member News It’s hard to believe but this year is the 28th year that Greater Rochester Chamber has conducted our annual Health Benefits Survey! Like all of our HR benchmarking surveys, this survey is only open to Corporate and Partner members for participation once a year. We know how important it is for members to…
90% of Organizations Report Strong AI and Cloud Foundations, but Business Performance Falls Short, Unisys Finds
/PRNewswire/ -- Unisys (NYSE: UIS) has released a new global report, "Unisys AI & Cloud Insights Report 2026: The benchmarks, the gaps, and what to do next,"...
Gas prices back at $4 a gallon
The average price US drivers pay for gas hit $4 a gallon on Monday, only the second time it has reached that benchmark since the war in Iran disrupted global oil supplies.
A new AI benchmark: Betting on the World Cup.
Can AI predict the future? A World Cup betting contest reveals which models make the smartest wagers.
Assassin's Creed Black Flag Resynced Performance Benchmark Review
Ubisoft's Assassin's Creed Black Flag Resynced breathes new life into the classic with modern graphics, gameplay improvements and full ray tracing support. In our performance review, we examine its graphical fidelity, VRAM usage, upscaling options, and how well it runs across a wide range of modern GPUs.
Python Is So Slow. Can Julia Solve the Two-Language Problem?
By some benchmarks, Julia code can run 10X to 1,000X faster than Python—but there’s a reason it’s not a very popular programming language.
Seahawks’ New Ownership Gets Blunt Devon Witherspoon Take
The Estate of Paul G. Allen announced on July 11 that it has agreed to sell the Seattle Seahawks to the Khosla family. Vinod Khosla, a limited partner with the San Francisco 49ers, is leading the ownership group.Moreover, the transaction will set a new benchmark for NFL franchise sales. ESPN’s Seth Wickersham and Adam Schefter […] The post Seahawks’ New Ownership Gets Blunt Devon Witherspoon Take appeared first on HEAVY.
The McLaren 788HS Is the Ultimate Evolution of the 750S
Limited to just 200 examples, McLaren's "definitive and final" 720S successor inches ever closer to the 800-hp benchmark.
Japan benchmark bond yield hits 30-year high on inflation, fiscal health concerns
The benchmark 10-year Japanese Government Bond (JGB) yield hit a 30-year high on Thursday, as rising oil prices rekindled inflation concerns and investors remained wary of Japan's fiscal health.
Minnesota K-12 health standards’ future unclear after administrative judge objects
Officials expected to phase in the state’s first-ever statewide school benchmarks on topics from nutrition and first aid to sex abuse prevention through 2028. That timeline is in jeopardy now after an administrative law judge ripped the state’s process to implement the standards.
Solid Growth Prospects Assert Sphere Entertainment Co. (SPHR) as a Top Stock to Buy According to Ariel Investments
Sphere Entertainment Co. (NYSE:SPHR) is one of the top stocks to buy according to Ariel Investments. On June 18, Benchmark reiterated a Buy rating on Sphere Entertainment (NYSE:SPHR) and raised the price target to $175 from $155. The price target hike comes amid expectations that the company is well positioned to deliver a solid second […]
How Flutter’s Removal from Key Russell Growth Indices Will Impact Flutter Entertainment (FLUT) Investors
In late June 2026, Flutter Entertainment plc (NYSE: FLUT) was removed from multiple Russell growth benchmarks, including the Russell 1000 Growth, Russell 3000 Growth, Russell Midcap Growth, Russell 3000E Growth, and Russell Small Cap Comp Growth indices. This broad index exclusion matters because it can prompt passive and index-tracking investors to sell shares, potentially amplifying trading volumes and sentiment shifts around Flutter’s stock. Next, we’ll examine how Flutter’s removal from...
Bolsonaro Adviser Says Fiscal Bill to Be Focus
Daniella Marques, an adviser to Brazilian right-wing presidential candidate Flávio Bolsonaro, said in an interview with SBT News that, if victorious in this year’s elections, his eventual government would propose a constitutional amendment focused on controlling spending and the debt trajectory right from the start. The goal would be to restore confidence and facilitate an adjustment in the benchmark interest rate.
Meta's Watermelon AI model has caught up to GPT-5.5, Alexandr Wang says
Meta's superintelligence chief says its upcoming Watermelon model now matches GPT-5.5 on key AI benchmarks.
Benchmark Scores Don’t Break. Clinical Reality Does. The Health AI Readiness Illusion.
Nature Medicine's frontier AI evaluation exposes a dangerous gap: models acing medical benchmarks collapse under adversarial pressure, and the FDA has no…
Secret Alchemist, Sarelly and Borntostandout Are Among the Winners of The Business of Beauty Global Awards 2026
From emerging challengers to established disruptors, this year’s Global Awards winners are setting a new benchmark for beauty entrepreneurship.
Flutter Entertainment plc(NYSE: FLUT) dropped from Russell 3000 Growth Benchmark
Flutter Entertainment plc dropped from Russell 3000 Growth Benchmark...
Evaluating the robustness and readiness of large frontier models in health AI applications
Nature Medicine - Adversarial evaluation of leading AI models uncovers gaps between benchmark success and robustness, highlighting limitations in current health AI benchmarks and their ability to...
New T-Mobile 5G Test Shows a Simple Fix for a Common Slowdown
T-Mobile may have found a practical way to make its 5G network work better when several nearby phones are all trying to pull data at the same time. The big takeaway is simple: in a new benchmark study, a newer Nokia-powered setup helped cut interference between phones, which led to faster overall performance. According to Signals Research Group, the testing was done at T-Mobile’s ... [read full article]
Cannes Lions 2026 unveils the Entertainment Lions shortlist, featuring 62 finalists that show how brands are competing for attention through film, content, experiences, talent and culture
Cannes Lions 2026 has released the shortlist for the Entertainment Lions category, one of the Festival’s most relevant benchmarks for
Chinese sodium battery surprised scientists by matching key Tesla benchmarks
Researchers found that a Chinese sodium-ion battery performs far better than expected, with production quality and design features comparable to Tesla’s batteries. If engineers can improve cold-weather charging and energy density, sodium could become a cheaper and more abundant alternative to lithium for EVs and large-scale energy storage.
Does Advertising Business Growth Assert AppLovin Corporation (APP) as a High-Growth Large-Cap Stock to Buy?
AppLovin Corporation (NASDAQ:APP) is one of the high-growth large-cap stocks to invest in now. On June 10, Benchmark reiterated a Buy rating on AppLovin Corporation (NASDAQ:APP)and a $775 price target. The research firm remains bullish on the stock, impressed by the company’s prospects in its consumer advertising business. A recent investor presentation affirmed management’s confidence […]
Multilingual benchmark evaluates how well AI interprets clinical text and health records in nine languages
Researchers at Mass General Brigham recently developed BRIDGE, a multilingual benchmark that evaluates how well large language models (LLMs) understand clinical patient care text, including language used ...
Valve Steam Machine Benchmarks Show Near Twice The Uplift Over Steam Deck & Comparable To Ryzen 5 5600X at 30W
New benchmarks of Valve's Steam Machine show performance comparable to a 6-core AMD Ryzen CPU from 2020, but could be decent if priced right.
Introducing LifeSciBench
Introducing LifeSciBench, an expert-authored, expert-reviewed benchmark for evaluating how AI systems handle real-world life science research tasks and decisions.
Steam Machine Performance Benchmark Leak Signals Approaching Launch
Since the news broke that the Valve Steam Machine would be delayed, there have been countless conversations online about how the Steam Machine would be dead on arrival because of its expected pricing and aging internals—a six-core Zen 4 CPU and a Radeon Navi 33 GPU with 8 GB of VRAM—but a recent set...
Steam Machine one step closer to launch as benchmarks surface online, here’s how it compares to the Steam Deck
We can now have an early look at some Steam Machine benchmarks after results were uploaded to Geekbench, suggesting reviews are underway.
The Steam Machine Is Likely in Testers’ Hands, as 2 Fresh SteamOS Benchmarks Leak
Steam Machine "Fremont" hits Geekbench running SteamOS, hinting at an imminent Summer launch for Valve's new console
Costs, careers and choice: Why Indians are having fewer children
India’s Total Fertility Rate (TFR) has dropped to 1.9 children per woman - lower than the benchmark level of 2.1.
Keck Medicine of USC earns LGBTQ+ Healthcare Equality Leader 2026 designation
LOS ANGELES — Keck Medicine of USC hospitals and USC Student Health, part of Keck Medicine, received the LGBTQ+ Healthcare Equality Leader designation in the Human Rights Campaign Foundation’s 2026 Healthcare Equality Index (HEI).HEI is the industry standard benchmarking tool for LGBTQ+ inclusion and equity practices in the healthcare field,...
BEACON: How Benchmarking Combats the Reproducibility Crisis in Modern Science
A new initiative called BEACON aims to advance scientific reproducibility by uniting siloed benchmarking efforts across biomedical research.
We can create the future of science right now
Paul Litvak is the founder and Executive Director of the Robyn Dawes Institute and a Visiting Scholar at UC Berkeley. He has a PhD in Behavioral Decision Research from Carnegie Mellon where he studied emotions and the sunk cost bias. Over 15 years in industry, he solved a wide range of challenging technical problems. At Meta he created models optimizing the ad review process. At Google he ran experiments to measure social influence. At Airbnb, he was a product manager leading machine learning teams optimizing search ranking and price suggestions. He also co-founded and led product at Rhythmic Health, a venture-backed biosensing startup, creating an accurate low cost system using a color changing strip and a smart phone to measure salivary lactate. The bottleneck At this point, it is uncontroversial to say that science needs to stop using the PDF article as the unit of knowledge and currency. The unbearable slowness of scientific publishing, the profit motive and margins: I’m not saying anything new. The PDF also sucks because it’s hard to extract structured information, which makes it hard to do evidence synthesis. As a result, we do much less evidence synthesis than is needed. And evidence synthesis ultimately undergirds most policy and medical decision-making. I can see second-by-second real-time odds for any sporting or newsworthy event, but a school board can’t see the best evidence on whether their 8th graders should be taught algebra. As a society, we don’t treat this as an important problem. Again, not controversial. Not only is the problem well understood, but the solution has already been laid out. What we need is AI-assisted living evidence synthesis - (1) an open knowledge graph of atomic claims, (2) claims linked to evidence, (3) assessment and synthesis of each piece of evidence, and (4) continuous updating with new data. What few realize (yet) is that the technical capacity to build this vision for a significant portion of science already exists. Not only that, scientists and startup teams are already building many of these components. I know this because I’ve been surveying the space and talking to many of the builders. There are some missing pieces: for example, evaluations of how well some of the components work. But at this point, most of what’s missing is a fully end-to-end working integration of all of these parts. In the rest of this essay, I’m going to lay out all the parts of a working living evidence layer for science and who is working on them, and propose concrete next steps for building this system. What’s now possible The diagram above outlines the components of a living evidence synthesis platform, including some of the teams working on each component1. Scientific PDFs are processed into claims with associated evidence. The evidence is subjected to a forensic audit, methodological evaluation and robustness and reproducibility checks. Finally it’s given a weight in a continuously updating synthesis. What follows is a description and status of each component and a few of the teams working on them. Document understanding The first thing you need to be able to do is turn an article into structured data. Mostly that means parsing PDFs. There are often multicolumn layouts that confuse non-specialized PDF to text processing libraries. For scientific papers, there is the added complexity of parsing formulas and tables and figures. This is a really hot area - there are startups offering APIs, and it seems like a new open source package gets posted to Github every few weeks. What follows isn’t exhaustive. A package called GROBID was the state of the art for a while, they didn’t update their package for nearly two years until very recently. In the meantime reducto.ai released an AI powered PDF extraction API, PaddleOCR became popular, IBM released a model called Docling, and both Mistral and Gemini created models and libraries. I also know of at least one other well-funded psychology research group working on a paper parser. By contrast, there are few open evals in this space, with no extensive evals for complex table comprehension in particular. Nonetheless, I’m confident this will be a solved problem soon, given the combination of LLM advances and developer interest. Hypothesis level extraction There has been increasing interest in comprehending the extracted text of papers and linking information to evidence for each hypothesis. A lot of work has already been done. Trialstreamer (Marshall et al. 2020) and RobotReviewer LIVE (Marshall et al. 2023) demonstrated automated extraction of trial population, intervention, and outcome at scale on clinical RCTs. PaperQA2 (Skarlinski et al. 2024) and Ai2 ScholarQA (2024) extended this to retrieval-augmented question answering with citation grounding. Elicit, Consensus, and SciSpace operationalized claim-level extraction for end users. OpenEval (Booeshaghi et al. 2026) is the most recent and most ambitious: 1.96 million atomic claims extracted from 16,087 eLife manuscripts using Claude Sonnet 4.5, grouped into ~299,000 results, with LLM evaluations showing 81% agreement with human peer review on a 2,487-paper subset. None of these solutions link claims to test-statistics, as you would need to evaluate randomized controlled trials. This is why I built the evidence.guide API - to extract hypotheses and associated test statistics from behavioral science papers. The best public eval of this kind of extraction I’m aware of comes from the recent SCORE project - they had humans code thousands of psychology papers to extract their claims by hand. It would be extremely helpful to the world if all scientific PDFs were available as structured open data. I’ve been working to make this happen, both directly at Berkeley and through coordination with large entities I can’t yet speak of; as hard as it is to do, I think it’s possible2. Forensic audit A lot of work has been done on forensic audit, but some gaps remain. Of course, for biology papers that rely on images for evidence, there are a variety of tools (notably Proofig and ImageTwin) to spot anomalies. These are still well short of what sleuths like Elizabeth Bik can do on her own, but these tools are constant companions among fraud analysts. There’s someone working on auditing Excel files for anomalies, and a number of teams are automating numerical checks like GRIM and SPRITE, including Scrutiny project, the INSPECT-SR team as well as statcheck. The regcheck team is building a way to use AI to compare preregistrations to analyses in papers, to ensure there aren’t significant deviations. Nonetheless, there are many other kinds of anomalies to screen for, both public and less publicly known. And there are no formal evals for anomaly detection that I’m aware of. Still, there’s a lot to draw from in this space and I’m pretty certain we will be able to scan papers for most kinds of obvious anomalies in the near future. Methodological review This area has been white hot, though I fear for many of the startups in this space, because this capability may become commoditized. There are at least six different AI peer review companies, including Refine.ink, Reviewer3, ReviewerZero.ai, Q.E.D. Science, Paper Wizard, and Isitcredible. Coarse (a pun on refine) was also recently created as an open source alternative. These systems provide qualitative feedback on the content of papers, spotting methodological weaknesses and mathematical errors. They seem to work pretty well, and many academics report bitterly that they exceed the average quality of typical peer reviewers. But there are few evals here either. What evals exist so far involve using LLM-as-judge (circularity problems abound) or comparing against human reviews of questionable quality. What you’d ideally want is an eval that measures capturing known errors in papers3. Reproducibility and robustness Another active area has been using AI agents to automate computational reproducibility4 and robustness5 checks in papers that report numerical results. For more recent papers where data and code are available, AI agents can see whether they can re-run the analyses and produce the numbers reported in the published paper. In addition to a handful of individual academics who have been experimenting using Claude Code for this, the Institute for Replication is a leading group working on building an end to end system. The evals related to this problem are the most mature, with CORE-Bench (Siegel, Kapoor, Narayanan 2024) and PaperBench (OpenAI 2025) available to benchmark agents on this task. There is also work on getting AI agents to test alternative ways of analyzing the data to ensure the results are robust to small analytic design choices. Synthesis This is the most underdeveloped area where significant investment is required. Although some automated evidence synthesis systems exist — for example, otto-sr is building an AI agent to write systematic reviews — none of these incorporate the full range of paper level signals to weight evidence appropriately. Nor is there anything like an eval or a gold standard for a good systematic review. Arguably Cochrane reviews are the closest we have to gold standard human systematic reviews, though I’ve heard academics in the know complain about their uneven quality. A key question for a synthesis platform is how to weight anomalies and methodological issues in assessing the quality of a piece of evidence. This is an unsolved problem and one I’m very keen to work on. Continuous updating There are many pieces of basic infrastructure available for monitoring for new research and initiating updates. OpenAlex is the current open citation graph. Retraction Watch integrated into Crossref in October 2023. Scite tracks how citations support, contrast, or mention prior claims. The Living Evidence Network demonstrated continuous-update workflows in clinical guidelines. Engineering this is a relatively straightforward task. When you look over this technical architecture and all the progress being made, it’s hard not to be optimistic that a living guide to scientific evidence will be built. The stakeholders are ready The social infrastructure for this is starting to coalesce — it’s not just a pie in the sky academic exercise to imagine this coming into existence. Institutions like the Center for Open Science, the Institute for Replication, the INSPECT-SR, the Living Evidence Network and more are all working on scaling work to improve research quality. Funders are also aligned. The Sloan Foundation has funded living evidence work through COS. Coefficient Giving supports the Institute for Replication and COS. The Astera Institute and the Institute for Progress have shown interest in this space. NIH has established an Office for Replication and Reproducibility. Although there are (very unfortunately) serious headwinds in science funding generally, there is an active group of funders interested in metascience. A brief word about what I’ve been doing at RDI. First, as a Visiting Scholar at Berkeley I’ve been actively figuring out how a non-profit and a public university can conduct and make public the results of large-scale academic article data mining. With some of the money I raised from donors, I commissioned a legal analysis of recent case law and publisher text data mining (TDM) agreements in order to understand whether a massive open data mining of academic articles is possible (with caveats, it is). I’ve also been working to bring together stakeholders in this space, and identify gaps. I’ve also been doing some software development in this space, with more to come. A pilot proposal The assumption undergirding all of this is that an AI, given all this information, would make the right judgment about a scientific claim with lots of conflicting evidence, weighing all the factors appropriately. That’s the hypothesis we need to test. Randomized control trial research is the best place to focus on first. RCTs are used to make many of the important decisions in society - from medical trials to public policy changes. And they use a relatively uniform set of inferential statistics with lots of known and available diagnostics. Behavioral science experiments, within the broader realm of RCTs, should be first used as a testbed whose results can be generalized. Because behavioral science is at the vanguard of open science practices, replications abound (there are thousands of them) to serve as ground truth training data. Key Hypothesis Therefore the pilot would test, in behavioral RCTs that have been replicated, whether the quality of evidence for a claim can be used to accurately predict whether that claim will replicate. Secondary Hypotheses Compared to claims that replicated, non-replicated claims demonstrate a greater share of forensic anomalies in their source literatures. Hypothesis level claim and statistic extraction is accurate enough to scale living evidence without onerous human review costs. Replication prediction is more accurate than prediction markets6 or journal prestige. If all the different quality signals we gather do accurately predict which studies will replicate, then we can use that model to score evidence to power the living evidence layer. Why this is informative regardless of outcome If the pilot succeeds, the architecture extends to medical RCTs (where Living Evidence already operates and integration is mostly about claim representation), then to slices of basic biology with stable replication structure. If it fails, the field learns which quality signals are load-bearing and which ones metascience has oversold. Either result is a contribution to knowing what the literature supports. Implications for funders Because this burgeoning ecosystem of builders already exists, a well-informed philanthropic or government funder could play a crucial catalyzing role in bringing this future about. They could play at least three roles: creating open structured datasets, publishing open benchmarks and incentivizing the solving of key technical challenges. First, open archives of papers that the government maintains, like PubMed, could be turned into structured data amenable to large scale metascience and claim aggregation. I know the US government already has an interest in doing this, though some key open questions remain unanswered. How do you determine the best models and systems for accurately extracting information from papers? How can you establish a robust way to allow researchers to flag errors and correct them? And finally, how do you create a legal regime in working with publishers to maximize the scope of available papers? For the latter, a university or a private philanthropy may be better positioned to make structured data publicly available under journal subscription terms or fair use. Philanthropists or government funders could also coordinate to create or commission benchmarks that evaluate whether important problems have been solved. For example, an open benchmark for claim extraction from a range of different scientific article types would be extremely helpful. Ensuring the underlying data are accurate is vital for creating these evaluations. I’ve discussed opportunistically using various human-created datasets for this purpose, but a consistent problem is that errors in human data make it difficult for them to serve as a gold standard. Finally, with structured data and benchmarks available, the government or private philanthropy could use them to incentivize groups to develop machine learning systems that meet these benchmarks. Prizes are one potentially valuable tool for this. For example, you could establish a prize for a system that accurately updates a living meta-analysis for a small set of claims. Prizes are particularly useful as signals of problem importance, and can help create vibrant ecosystems of public and private research—see, for example the role the government played in kickstarting the current work on nuclear fusion. Conclusion The drawbacks of the current scientific publishing system are known. Scientists agree, metascientists agree, philanthropists agree: the published PDF plus citation graph isn’t the right substrate for maintaining a representation of the evidence base in science. The pieces needed to build the alternative either already exist or are rapidly taking shape. The community is forming around exactly this problem, with concrete partnerships and shared infrastructure. A pilot should start on behavioral science RCTs because that’s the slice of empirical science most amenable to legibility, where replication ground truth is richest, and where the failure modes are best documented. What’s been missing is the galvanizing mission to assemble these pieces into something that works. That’s what I’m proposing to build. *** 1 I have a broader field map that I’ll release publicly soon. This is me, building in public! 2 If this is something that you are excited about, please reach out and talk to me. 3 More on this very soon too! 4 This tests whether, given the code and the data, you can get the same statistics as reported in the published paper. 5 This tests whether the results are the same as a paper’s given alternative analytical decisions in conducting the analysis (like outlier omission). Closely related is the idea of a “multiverse” where you come up with many different ways of answering the same underlying research question with the same data, and test whether the results hold in all those alternative methods. There’s been work on the latter as well. 6 Some of the replications, e.g. those from the SCORE project, had paired the experiments with forecasts from prediction markets. So we get to look at this for free.
Steelers gift Joey Porter Jr. all the leverage
Joey Porter Jr.’s price may have climbed after Pittsburgh set a costly internal benchmark with Nick Herbig’s extension.
Chrome Sets Browser Speed Records on M5 MacBook Pro
Google's Chrome browser hit new records on browser benchmarking tools Speedometer 3.1 and JetStream 3, Google said today. Chrome earned a score of 61 on Speedometer, a five percent improvement since last year. It earned a 469 on JetStream 3, a 10 percent improvement since the beginning of 2026. Tests were done on an M5 MacBook Pro running macOS 26.0.1. Google says it holds a dual record across all browsers, beating every other Mac browser, including Safari.
Building Structured Generative AI Evaluation at Enterprise Scale
How Oracle built Veritas, a scalable GenAI evaluation framework for structured benchmarking, reusable pipelines, reproducible testing, and enterprise AI evaluation workflows.
Here’s What Put Seaport Entertainment Group (SEG) on a Strong Footing For Success
US-based investment company, McIntyre Partnerships, returned -19% gross and -20% net in the first quarter of 2026 compared to the Russell 2000 Value Index’s 5% return. A copy of the letter can be downloaded here. Since inception, the fund has returned ~14% gross and ~10% net per annum, surpassing the benchmark’s return of ~7% per […]
“Sets an entirely new benchmark for portable, affordable audio interfaces”: SSL just released its cheapest audio interface yet
The SSL 1 is a 2-in/2-out device that comes at a very tempting price
College sports isn’t broken — it’s ungoverned. And the fix is being ignored
Want to save college football? Tie earning power, player movement to academic benchmarks.
UK gilt yields retreat from multi-decade highs as political drama mellows, rate hike expectations ease
As bond markets returned from a bank holiday on Tuesday morning, the yield on the benchmark 10-year bond stood at 4.85%.
Aaron Rodgers’ 2026 Steelers Contract Incentives Revealed
Aaron Rodgers could earned $2.5 million in incentives next season, but he has his work cut out for him to hit those benchmarks.
New MatterChat Model Helps AI to ‘See’ the Language of Science
From writing emails to generating computer code, much of the artificial intelligence prevalent in our daily lives has succeeded by mastering one domain: text. However, this leaves a major blind spot in the physical sciences, where models depend on the high-resolution, three-dimensional data of the physical world, like the intricate lattice of atoms in a crystal. Delivering on the promise of using AI for science requires teaching these data-driven text models to seamlessly “talk to” physics-based models. Now, a new AI framework from Lawrence Berkeley National Laboratory (Berkeley Lab), called MatterChat, solves this problem by creating a specialized “bridge.” It connects the conversational power of a Large Language Model (LLM) with a physics-based AI that models “interatomic potentials”: the complex physical forces between atoms. The resulting system already significantly outperforms general-purpose AI tools like GPT-4 at predicting material properties, and the team hopes it can accelerate scientific discovery by serving as a robust research partner that provides grounded insights and generates step-by-step instructions for synthesizing novel materials. A paper describing this work was recently published in Nature Machine Intelligence. “Traditional simulations can provide the physical rigor required for materials science, yet their computational cost remains prohibitive for high-throughput screening. Conversely, while LLMs excel at rapid knowledge synthesis, they inherently lack the ‘structural vision’ to interpret materials directly from their underlying atomic coordinates,” said Yingheng Tang, a postdoctoral researcher in Berkeley Lab’s Applied Math and Computational Research Division (AMCR) and lead author on the paper. “MatterChat was built to solve this dilemma, empowering LLMs with a structural ‘vision’ that allows researchers to leverage their full potential for solving complex, real-world materials challenges.” Empowering language models to solve complex challenges in materials science To build MatterChat, the Berkeley Lab team drew inspiration from technologies like Vision Question Answering (VQA) and Text-to-Image (T2I) generation. In these tasks, AI must translate high-level text concepts into visual images, or vice versa. Doing so requires developers to build tools that “bridge” two fundamentally different forms of data. The researchers adapted this concept to the physical sciences. With MatterChat, they created a “bridge model” that successfully connects an LLM’s general knowledge with the deep understanding of the atomic-scale world encoded in scientific interatomic potentials. Until now, researchers using LLMs to solve materials problems usually had to feed them raw data files as if they were just strings of text. It’s like asking an AI to understand a complex 3D engine based only on a parts list: the LLM can read the names, but it can’t “see” how the atoms fit together in space. MatterChat solves this by training a specialized AI bridge model, pre-trained on millions of crystal structures and an LLM, to align the LLM’s representation of the world with the interatomic potential’s representation of the world. The bridge model can translate physical insights into a format the LLM can actually understand. By giving the LLM these “scientific eyes”— a scientific “inductive bias” in the terminology of AI — the Berkeley Lab team has transformed it into a robust research tool capable of providing grounded scientific insights into complex materials challenges, such as predicting thermal stability or analyzing electronic band gaps. “We think of atoms as living in a physical space, but from a machine learning perspective, they are just vectors living in some very non-trivially structured manifold in a high-dimensional Euclidean space; and, of course, the same is true for the sentences and paragraphs that we use to express our ideas about those atoms,” said co-author Michael Mahoney, Berkeley Lab’s Scientific Data Division (SDD) AI Initiative Research Lead. “The bridge model basically gets those two structures to ‘talk with’ each other.” As a proof-of-concept of this general approach, the team trained their bridge model on a dataset curated by pairing nearly 143,000 stable atomic structures from the Materials Project with their corresponding physical properties. This training data was automatically assembled using the Materials Project’s API and deliberately enriched with properties fundamental to microelectronics design — like formation energy and bandgap — allowing MatterChat to learn the complex patterns connecting a material’s atomic blueprint to its functional performance. To validate their model, the researchers benchmarked MatterChat against a suite of other AI systems, from general-purpose LLMs to other specialized scientific AI methods. The results show that MatterChat consistently outperformed its competitors across a range of tasks. The model was more accurate in classifying material types and demonstrated superior precision in predicting numerical properties. For example, it excelled at predicting a material’s bandgap, a property critical for designing new electronics from high-capacity energy storage to next-generation computer chips. “Our design is significantly more efficient because we don’t have to build a massive AI model from the ground up,” said co-author Zhi (Jackie) Yao, a research scientist in Berkeley Lab’s AMCR. “Instead, we take two powerful, pre-trained models — a structural encoder for materials physics and an open-source LLM — and use them off-the-shelf. The only component we actually train is the lightweight ‘bridge model’ that translates between them. It’s the difference between building an entire car factory and simply designing a smart adapter that connects a world-class engine to a world-class navigation system. This approach is not only computationally efficient, but also makes the system modular, so we can easily upgrade components or adapt the bridge for other scientific domains in the future.” Crucially, this modular design highlights exactly how institutions like Berkeley Lab and the Department of Energy are carving out a highly valuable niche in the booming AI landscape. Rather than competing with Silicon Valley tech giants to build ever-larger language models from scratch, the lab is focusing on the specialized connective tissue that makes commercial AI useful for hardcore science. Because the bridge model approach underlying MatterChat is forward-compatible, it is perfectly positioned to leverage these parallel tracks of innovation. As Mahoney pointed out, “We expect that industry will continue to develop improved LLMs, and we expect domain scientists and facilities will continue to generate new data. An important part of scientific machine learning is not simply to solve problems on today’s data, but instead to develop general methods that will be forward-compatible with orders of magnitude more data, whether from scientific domains or from LLMs.” According to Yao, the MatterChat project, which was initially developed and enhanced with funding from a Berkeley Lab Laboratory Directed Research & Development (LDRD) Program, will now expand its capabilities. In a collaboration with Fermilab, MatterChat is already contributing to a U.S. Department of Energy Genesis Mission project — called Accelerating eXtreme Environment Specs-to-Silicon (AXESS) — that aims to speed up the development of next-generation, high-speed, radiation-hardened detectors for challenging particle physics experiments by using advanced 3D integrated circuits (chiplets) and AI-driven data analysis. In addition to the LDRD support, the team also credits supercomputing resources at the National Energy Research Scientific Computing Center (NERSC), located at Berkeley Lab, with MatterChat’s success. “We are incredibly grateful to NERSC; this research simply would not have happened without access to the Perlmutter supercomputer through their AI for Science program,” said Tang. Wenbin Xu, a NERSC postdoctoral fellow at the time, was also a major co-author of the work, as was Benjamin Erichson, a research scientist in Berkeley Lab’s SDD, highlighting the benefits of AMCR-SDD collaboration on AI for science. ### Lawrence Berkeley National Laboratory (Berkeley Lab) is committed to groundbreaking research focused on discovery science and solutions for abundant and reliable energy supplies. The lab’s expertise spans materials, chemistry, physics, biology, earth and environmental science, mathematics, and computing. Researchers from around the world rely on the lab’s world-class scientific facilities for their own pioneering research. Founded in 1931 on the belief that the biggest problems are best addressed by teams, Berkeley Lab and its scientists have been recognized with 17 Nobel Prizes. Berkeley Lab is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy’s Office of Science. DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.
Only 3% of students meet standards in science: National assessment data revealed
Partial data from Israel’s 2025 Meitzav national assessment exams has been revealed, showing that only 3% of ninth-grade students met the benchmark in the science exam, while 54% failed.
Sustainable science mapping: benchmarking green AI against transformers for cross-disciplinary abstract classification using arXiv
Scientific Reports - Sustainable science mapping: benchmarking green AI against transformers for cross-disciplinary abstract classification using arXiv
Microsoft’s multi-agent AI system tops Anthropic’s Mythos on cybersecurity benchmark
Microsoft's new vulnerability-scanning system, codenamed MDASH, scored 88.45% on the CyberGym benchmark, surpassing single-model systems from Anthropic and OpenAI by using more than 100 specialized AI agents across multiple models.
AI chatbots struggle with subtle mental health cues
AI models miss the cues clinicians catch, a new benchmark finds.
TribeVibe Entertainment sweeps WOW Awards 2026 with five wins, cementing its position as a defining force in India's live entertainment landscape
TribeVibe Entertainment, one of India's fastest-growing live entertainment companies, was honoured with five prestigious accolades at the Shows of India Conclave 2026 - among the industry's most respected benchmarks for excellence in live experiences and entertainment - notching up as many as five WOW Awards across categories.
Preparing for The Quantum Era: AT&T Business Debuts Post-Quantum Cryptography Secure SD-WAN, Powered by Cisco
AT&T sets a new benchmark for networking as the first major North American service provider to deliver Post‑Quantum Cryptography-enabled SD‑WAN
Delaware Health Care Spending Grows to $11.3 Billion, Significantly Exceeding Target - State of Delaware News
Per capita health care spending in Delaware increased to $11.3 billion dollars in calendar year 2024, an increase of about 8.7% compared to 2023.
Sony Xperia 1 VIII: European model leak showcases Snapdragon 8 Elite Gen 5 performance but only 12 GB RAM
The Sony Xperia 1 VIII has now surfaced on a popular benchmarking website. Featuring Qualcomm's Snapdragon 8 Elite Gen 5 chipset, the new Xperia flagship is once again saddled with 12 GB of RAM unlike many of its peers.
COA and Flatiron Health Study Finds Patients May Experience Longer Survival in Community Oncology Settings
Patients diagnosed with metastatic breast cancer and non–small cell lung cancer had longer survival compared to national benchmarks when treated in independent community oncology practices, according ...
Iterative Health Reports Expanded Trial Performance Data at Digestive Disease Week 2026, Demonstrating Consistent Outperformance of Industry Benchmarks at Double the Scale
CAMBRIDGE, Mass., May 04, 2026--Iterative Health announced the presentation of updated scientific research at Digestive Disease Week (DDW) 2026.
The science behind the Adidas shoes that helped two marathoners break the two-hour mark
A sub-two hour marathon has long been seen as a tantalizing benchmark for elite runners—and shoemakers have been in a race to design footwear that can help them get there
Indiana hospital systems meet benchmark on direct-to-employer health care deals
The report comes as Indiana hospitals attempt to address widespread criticism — including from state leaders — over high prices.
Scene One Entertainment opens Market Arcade 8
Scene One Entertainment, alongside Benchmark Management Group, opened the Market Arcade 8 on Wednesday in downtown Buffalo - the latest for the Schenectady-based entertainment company, and its first foray in Western New York.
Indiana hospital systems meet benchmark on direct-to-employer health care deals
Nonprofit hospital systems are required under Indiana law to offer direct-to-employer arrangements at or below 260% of Medicaid pricing.
AMD Ryzen 9 9950X3D2 Benchmarks: The Best Desktop Performance For Linux Developers, Creators Review
Today we can finally share performance benchmarks of the long-rumored AMD Ryzen 9 9950X3D2 Dual Edition processor.
How MLS is changing to compete with world football’s elite
Major League Soccer is preparing for one of the most ambitious transformations in its recent history. The goal is to position the league as a global benchmark capable of competing with Europe’s bigg...
Harvard Business Review Recognizes BBVA as a Benchmark for Corporate AI Adoption
Harvard Business Review has identified BBVA as an example of how expanding access to generative AI tools can channel internal demand
New Debian Project Leader Elected For 2026
Sruthi Chandran has been elected the new Debian Project Leader 'DPL' after running unopposed in this year's elections.
Dimensity 9600 Pro Beats A19 Pro And Snapdragon 8 Elite Gen 5 By 25% In Benchmarks, But Extreme Heat May Erase That Gap
A tipster shared the first single-core and multi-core scores of the Dimensity 9600 Pro, revealing it’s much faster than the A19 Pro and Snapdragon 8 Elite Gen 5
Chargé Arnold Speech - Implementation Plan of Five-Year Health MOU - U.S. Embassy in Madagascar
Chargée d’Affaires Stephanie Arnold's Speech Discussion on the implementation plan for the five-year health Memorandum of Understanding (…) Colleagues, and honored guests, Good morning, and welcome to the United States Embassy. It is a pleasure to host you today for the second day of this important workshop to discuss the implementation of our bilateral health Memorandum of Understanding on U.S. Government health activities in Madagascar, made possible through the Trump Administration’s America First Global Health Strategy. Today’s focus— discussing a shared vision and developing a preliminary draft for the health MOU Implementation Plan and the terms of reference for the two committees that will steer its implementation—is a decisive step in moving from commitments on paper to concrete action over the next five years. On December 22, 2025, the Governments of Madagascar and the United States signed a five-year bilateral health MOU, with a financial intention of approximately 554 billion Ariary from the United States and This jointly decided MOU intends to strengthen Madagascar’s health system, support lifesaving services, and improve disease prevention and response that matter to both our countries’ health security. For the United States, this means detecting and containing infectious disease outbreaks including malaria, polio, and emerging health threats before they can reach American shores through international travel and trade, protecting American families while reducing the need for costly domestic emergency responses Yesterday, technical experts from the Ministry of Public Health and the U.S. Government worked in thematic groups to draft a preliminary Implementation Plan. Today, your task is to review the shared vision and preliminary draft MOU implementation plan, and the proposed terms of reference for the Joint Steering Committee and the Operational Committee that will guide its implementation. These committees are not just administrative structures. They are the mechanisms that will ensure clear roles and responsibilities, regular dialogue between our two governments, consistent monitoring of progress, and real accountability for the results that will directly benefit the lives of Malagasy citizens. They will also track performance against the benchmarks we have jointly established, which will guide the continuation of U.S. support over the five-year period . Continued U.S. support will be contingent on meeting these performance benchmarks. This ensures accountability to both American and Malagasy taxpayers and guarantees that our partnership delivers measurable results. America First Foreign Assistance means that U.S. support must be strategic, results-driven, and accountable to our taxpayers. It also means we aim to support greater self-reliance and reduced long-term dependence on foreign assistance, with systems that are increasingly financed and led by Madagascar itself. This approach benefits both our nations: Madagascar gains sustainable health capacity, while the United States advances our health security interests and creates opportunities for American health innovations and companies to support Madagascar’s development. In that spirit, I want to recognize the presence of the Ministry’s leadership and the technical directors and program heads responsible for surveillance, basic health services, family health, immunization, countering malaria, laboratories, medicines, and health information systems. Your leadership and technical expertise are essential if this bilateral health MOU is to translate into real and lasting improvements for people in need across the country. Let me emphasize how important collaboration is to the United States in this effort. We can bring American resources, technical expertise, and tools. But this bilateral health MOU can only succeed if it reflects Madagascar’s priorities and if Malagasy institutions lead and maintain its implementation. We see this not as a U.S. project, but as a joint framework under Malagasy leadership, with U.S. support. That is why we are genuinely pleased to host this event here at the Embassy. Opening our doors to you today symbolizes our commitment to work side by side and to remain transparent about how we use U.S. resources in support of Madagascar’s health system. In closing, thank you for your time and your partnership. We are very happy to welcome you here and proud to accompany Madagascar in this next step of implementation planning for our bilateral health MOU. I wish you a productive and focused session. The work you do today will shape how our two governments cooperate on health in the years ahead; it will influence the quality and reliability of health services that Malagasy communities receive; and, most importantly, it will help protect both our nations from infectious disease threats while building Madagascar’s capacity for long-term health system sustainability. Thank you.
United States and Madagascar Discuss Five-Year Health MOU Implementation Plan - U.S. Embassy in Madagascar
United States and Madagascar Discuss Five-Year Health MOU Implementation Plan The United States and Madagascar’s Ministry of Public Health are in the process of developing a plan for implementing the U.S.-Madagascar five-year Memorandum of Understanding (MOU) on health cooperation, signed under the Trump Administration’s America First Global Health Strategy. The two-day workshop, at the U.S. Embassy in Antananarivo, brought together technical experts and senior leadership from the Malagasy and U.S. governments to discuss a shared vision through a preliminary draft MOU Implementation Plan and to establish the joint governance structures that will guide U.S. foreign assistance in health over the next five years. The jointly decided bilateral health MOU, signed on December 22, 2025, builds on decades of U.S. global health assistance to Madagascar. The health MOU involves the U.S.’s plans, working with Congress, to provide approximately $134 million (over 554 billion Ariary) and a Malagasy commitment of $41 million (over 169 billion Ariary). It aims to strengthen Madagascar’s health system, improve disease prevention and response, and support services that address health security concerns for both countries, including malaria, maternal and child health, polio, and global health security issues that could reach U.S. shores if not contained at their source. During the recent workshop, technical staff from the Ministry of Public Health and the U.S. Government health team worked in thematic groups to draft detailed implementation plans for the five-year MOU period. On the second day, U.S. Chargé d’Affaires Stephanie Arnold and Ministry of Public Health and Economic and Finance Ministry leadership joined the technical teams to discuss plans and proposed terms of reference for two committees: a Joint Steering Committee and an Operational Committee. These structures will ensure clear roles and responsibilities, regular dialogue between the two governments, consistent monitoring of progress, and accountability for results. The framework includes performance benchmarks that will guide continued U.S. support and track Madagascar’s progress toward self-reliance. “We see this not as a U.S. project, but as a joint framework under Malagasy leadership, with U.S. support,” said U.S. Chargé d’Affaires Stephanie Arnold during the workshop’s opening session. The approach reflects U.S. foreign assistance principles that emphasize strategic, results-driven support that advances American interests while helping recipient countries build sustainable health systems. The framework is designed to build systems that Madagascar will increasingly finance and lead itself, reducing long-term dependence on foreign assistance.
Outpatient & Specialty Care Stocks Q4 Recap: Benchmarking agilon health (NYSE:AGL)
The end of the earnings season is always a good time to take a step back and see who shined (and who not so much). Let’s take a look at how outpatient & specialty care stocks fared in Q4, starting with agilon health (NYSE:AGL).
Industry Insight: See How Your Business Compares With Pay and Benefits for Key Roles
The Scoop’s annual salary survey offers benchmarking data to aid with recruiting and retaining ag retail employees.
U.S. wholesale prices surged 4% last month during the Iran war
The surge in prices complicates the work of the inflation fighters at the Federal Reserve, who have faced intense pressure from President Donald Trump to lower their benchmark interest rate. But some Fed policymakers are inclined to raise rates instead, as higher energy costs increase the inflation threat.
“AI polls” are fake polls
A few weeks after Donald Trump’s second presidential win, I took the train up from London (where I was living at the time) to Oxford to attend a conference on polls and forecasts of the 2024 election. Most of the attendees were pollsters or academics, but I also watched presentations from Aaru and Electric Twin, two companies that do what is interchangeably called synthetic sampling, silicon sampling, or synthetic audiences. Stripped of startup jargon, that means they use large language models (LLMs) to simulate responses to public opinion polls by having AI agents take on the role of survey respondents. I had already heard of Aaru thanks to some articles with eye-catching headlines like “No people, no problem: AI chatbots predict elections better than humans” in the months leading up to Election Day. The founders were making some big, some might even say far-fetched claims, such as: “within two years, we will simulate the entire globe — from the way crops are grown in Ukraine to how that impacts production of oil in Iraq, trade through the strait of Malacca, and elections for the mayor of Baltimore.” When Semafor asked Aaru’s cofounders — Cameron Fink and Ned Koh — about my boss, they said “we respect all those who came before us.” Nate (as he so often does) shared his thoughts on Twitter: Fink and Koh were relatively good-natured about this back-and-forth when we spoke at Oxford. They even offered to mail me one of the t-shirts featuring Nate’s quote they apparently had made. I never took them up on the offer, which I now somewhat regret. These synthetic sampling companies fell off my radar for a while, but they do still exist. In fact, Aaru recently received a $1 billion valuation. Is what they’re doing anywhere close to the most important frontier in AI development? Not by a longshot, especially when Anthropic just developed a model so adept at exploiting software vulnerabilities that it’s only being released to 40 companies. Still, silicon sampling is increasingly finding its way into public polling. Axios reported in March that “a majority of people trust their own doctors and nurses” based on findings from Aaru — without mentioning that the “people” in that sentence were actually LLMs. Around the same time, the Public Sentiment Institute “boosted” their online sample of 373 real survey respondents with 114 AI agents.1 (Spoiler alert: even the co-founder of Electric Twin doesn’t think that’s a particularly defensible approach.) Polling companies like Qualtrics and Ipsos are also developing synthetic data panels. So, what should we make of these … “polls”? Let’s get one thing out of the way: whatever they are, they’re not polls in the way that term is usually defined. Subscribe You can’t replace polls with AI On one hand, using LLMs to essentially make up fake survey respondents sort of sounds like the dumbest idea ever, one that will at best imperfectly replicate real polls while introducing all sorts of biases. On the other hand, with LLMs improving at a remarkable, perhaps even alarming rate, maybe that means I’m a dinosaur at the ripe old age of 24 because I still want to rely on polls that talk to actual people. I’m not going to argue that synthetic samples are completely useless. In fact, as I’ll return to later, there is evidence that some techniques can replicate topline survey results quickly and cheaply. But the marketing from certain companies can be slightly optimistic. “No traditional poll will exist by the time the next general election occurs,” said Fink in 2024. We’re just 206 days away from the midterms, and based on the fact that I still have to collect a bunch of polls every day, I’d say he should have run that prediction by a sample of AI agents before the interview.2 To see why synthetic samples can’t replace polls, here’s a quick primer on how they work. The simplest version of these models involves taking a LLM (like ChatGPT or Claude), giving it a demographic profile (e.g., a white, college-educated woman who lives in Utah and makes $70k a year), and then asking it to respond to a survey question. You repeat that process a few thousand times using different demographic profiles and end up with a sample of synthetic survey responses. The actual models used by private companies are more sophisticated than this, usually because they incorporate more hypothesized demographic characteristics for each agent and provide them with extra information. Aaru, for example, feeds agents a diet of news and information they’d be likely to consume, while Electric Twin incorporates their customers’ proprietary data about the audience they’re trying to replicate. The way Ben Warner, the co-founder of Electric Twin, explained it to me was “we have a large amount of data on […] for instance, 5,000 people. Can we make an accurate prediction of how they would respond to another question?” Still, it should be obvious why synthetic samples can’t replace polls. Polling is fundamentally a data collection process. We might use surveys to make predictions by feeding them into election forecasts, but the main purpose of a poll isn’t prediction, it’s gathering new data about what people think and how they feel. Silicon sampling, on the other hand, produces no new data. It’s simply a model: you input LLM training data, demographic prompts, and a bunch of other information, and it spits out a prediction for what a poll would say. We love models here, but models aren’t polls. That difference is an important philosophical sticking point for most pollsters I talk to. “I think politics should stay away from [synthetic sampling], because we’re trying to […] represent the voice of the people,” said Natalie Jackson, a vice president at GQR Insights. Democratic pollster John Hagner told me: “I think I’m just incredibly skeptical of this idea. I don’t think it’s research. At that point, you’re asking the machine to tell you what you already believe.” Hagner has seen some presentations of early synthetic sampling experiments, but so far, “if it’s being used in a campaign, people are keeping it incredibly quiet.”3 But Eli, I hear you saying, aren’t polls themselves increasingly governed by modeling decisions? Indeed they are: pollsters’ choices on which sampling method to use, how to define their likely voter models, and how to weight their samples can and do lead to dramatic differences in the results they publish. Aaru even referenced these limitations in the methodology statement included with that maternal mortality “poll” — although I’m using the term “methodology statement” loosely here, because it doesn’t really explain how the model works at all. We can ignore the (frankly preposterous) implication that synthetic sampling isn’t subject to a separate set of biases. The important point is that there’s still a meaningful difference between using weighting and other statistical techniques on actual polling data and using a model to predict what a poll would say. The latter is far closer to election forecasts or techniques like MRP — potentially useful models, but not a replacement for polls.4 To be fair, other synthetic sampling companies are perfectly happy with the distinction between polls and models. Warner compared polling and synthetic sampling to different tools in a toolbox. “The mistake I think we make is we think that these new tools should either work in exactly the same way or somehow replace these old tools,” he said. “Rather than thinking of it as, okay, so we’ve always had the hammer, we’ve always had the screwdriver, now we’ve got a saw. But don’t use a saw to try [to] do the job of a hammer.” A quick comment from Nate Eli didn’t ask me for a comment — rather rude of him, don’t you think? But since I’m editing this story, I figured I’d add a few quick thoughts rather than putting words in his mouth. Beyond the frequently misleading marketing, what bothers me about the AI “poll” hype is that as AI tools make statistical inference cheaper and/or better (note that these are not synonyms) that actually increases the comparative value of collecting original data. You might be able to train a model to make a reasonable estimate of what some hard-to-reach poll respondent would say — say, a young Black man who voted for Trump. (Such a person checks a number of boxes for a voter who is usually hard to reach in surveys.) Indeed, this is closely related to what models like the Silver Bulletin forecast already do. They essentially smooth out the kinks in noisy survey data by making inferences based on past voting patterns or national polls or surveys of other states. But you don’t actually know what these voters think unless you’re reaching them directly. If there’s a shift in opinion among this subgroup, you’re not going to detect it. So if I were running a campaign, I’d invest more in going the extra mile to find a representative sample of those voters. And then I’d hire some smart quants — or Claude? — to figure out the implications for campaign strategy based on proprietary data that my competitors didn’t have access to. -Nate Silver Are these models any good? If synthetic surveys are just a new type of model, the next obvious question is whether the models are at least accurate. The answer very much depends on who you ask. On one end of the spectrum, you have the maximalist argument that synthetic sampling is more accurate than actual polls. “It’s an incredibly challenging problem to go to someone and say ‘hey, we’re going to be more accurate at predicting human behavior than you, even when you talk to your customers directly’,” Koh recently told CNBC. In his view, synthetic sampling isn’t a saw to polling’s hammer, it’s “magic.” There’s certainly evidence that synthetic samples can replicate certain survey toplines. But if Aaru does have any examples of their approach outperforming the polls, they’re keeping those to themselves.5 Aaru’s 2024 election model, for example, had Kamala Harris leading in Michigan, Nevada, Pennsylvania, and Wisconsin on November 4th. And although they’ve since taken down their forecast page, they gave Harris a 50.5 percent chance of winning the race on November 2nd.6 After the election, Fink told Semafor he was happy enough with those results because they were “within margin of error,” a term that is completely meaningless when applied to a “sample” of AI agents. And of course, Aaru says their models have improved since 2024, so supposedly now they’d be more accurate than the polls? Still, their stronger argument is on cost: “We are significantly faster and cheaper than traditional polling, and still more accurate,” said Fink. The first two claims are undeniably true, but the third brings us to the opposite end of the spectrum. Both Jackson and Hagner are skeptical that these models are reliable for anything beyond replicating common survey toplines. “I just […] don’t think the machines are what we want when we’re looking for nuanced views. My example on this is people in Arizona and Nevada in 2024 who voted for Trump and voted for expanding abortion in their states on ballot initiatives,” said Jackson. Hagner identified another issue. Maybe the synthetic respondents, like sycophantic LLMs, are inhuman in one important way: they’re too nice. “The reports that have come through at the meetings that I’ve been at are that the early experiments on this, they cannot get respondents to be as racist or sexist or, frankly, as negative as human respondents,” he said. Academic research mostly agrees on this point. While there are some papers that show promising results when using LLMs to replicate polling data, most show that LLMs suffer from various quirks like producing too few “don’t know” responses and can seriously overpredict the favorability of politicians like Donald Trump and Kamala Harris. They also seem to struggle with too little variation between demographic subgroups, so the difference in predicted opinion between Democrats and Republicans, for example, is too small. When I asked Warner about these studies, his response to these papers was that just because academics can’t get synthetic sampling to work doesn’t mean that the technique doesn’t work in general. “Actually, the argument is, okay, yours does not [work]. That does not mean […] for this complex set of machinery, which uses a lot of investment, a lot of time, a lot of money, you can’t get it to work.” Cards on the table, I’m somewhat sympathetic to this argument because academics aren’t exactly great at making election forecasts. Usually, the people with skin in the game are the most accurate. Warner’s argument is that the approach Electric Twin takes — which includes, for example, making multiple predictions for each synthetic respondent using different models and prompts and subsequently averaging those to get a final prediction in a sort of ensemble forecast — produces better results than the simpler academic models. Warner shared a comparison between his method and the method from a recent academic paper with me, and Electric Twin was indeed able to get more accurate replication. But even still, he acknowledged that synthetic sampling “is not a crystal ball.” “If you asked me, do I think using other data sources will be more accurate than asking somebody who they will vote for, I would probably say no. But if you asked me ‘would your system be useful for our turnout modeling today?’ I would say yes.” For better or worse, it looks like the method is already getting more popular in the market research world. Most of the clients Aaru touts these days are businesses like EY and McDonald’s. And AI will probably begin to pop up in other parts of the political polling process. Pollsters are already using it to code open-ended survey responses, and some firms, like YouGov, are testing using LLMs to ask survey respondents questions. More worryingly, one danger to actual polls is that AI agents can be used to infiltrate online surveys. Most online polls use various checks to prevent that from happening, but there’s conflicting evidence on how effective those filters are and how prevalent AI agents currently are in online panels. If those agents ever become impossible to detect, it might spell the end of online polling, but the solution isn’t to replace all of your respondents with ChatGPT. Silver Bulletin is a reader-supported publication. To receive new posts and support our work, consider becoming a subscriber. Subscribe 1 That particular poll obviously doesn’t meet Silver Bulletin standards for aggregation. But we exclude all Public Sentiment Institute polls from our averages because we classify them as an amateur polling firm. 2 You could argue that Fink meant the next presidential election, but (a) I’m also confident we’ll still have real polls in 2028 and (b) in that case he should have asked an LLM to define “general election.” 3 Quick caveat: that’s reporting from a Democratic pollster. It’s possible that Republicans are more willing to use AI in political campaigns. 4 Indeed, Silver Bulletin does not include “polls” produced by MRP in our forecasts or averages, and we think it’s extremely misleading when their practitioners describe them in a way that suggests original data had been collected among a large number of states or Congressional districts. 5 A recent report from Aaru and EY did show two examples of a synthetic estimate being closer than a survey to a real-world benchmark — but I’d take those findings with a grain of salt because the report reads more like an ad and didn’t involve any sort of prediction being made ahead of time. 6 For comparison, our odds for Harris on the same day were 48.2 percent.
Qualcomm's Snapdragon X2 Elite Extreme, Tested: Big 18-Core Energy for Laptops, Incoming!
Debuting in Asus’ exciting 16-inch Zenbook A16, Qualcomm's 2026 flagship processor looks like a benchmark bombshell ready to shake up the laptop CPU field. We have the receipts.
Asian benchmarks jump after oil prices sink in response to the Iran ceasefire
Asian shares are surging as oil prices plunged after the U.S. and Iran agreed to a two-week ceasefire that includes the reopening of the Strait of Hormuz.
16 offensive tackle prospects who fit Packers experience benchmarks
Green Bay has coveted experience, ideally on the left side in their draft picks at offensive tackle. Here are 16 prospects who fit the bill in 2026.
Qualcomm Snapdragon X2 Elite Extreme Analysis, Benchmarks & Efficiency - Serious rival for Apple and a problem for AMD & Intel
Notebookcheck review/analysis of the Qualcomm Snapdragon X2 Elite & X2 Elite Extreme mobile ARM processors.
The MacBook Neo is proof that Google’s Pixel plan was right all along
Apple's MacBook Neo validates the Google Pixel strategy by emphasizing user experience and ecosystem growth over raw hardware benchmarks.
Galaxy S26 FE Is Up To 33% Slower Than Exynos 2600-Powered Galaxy S26 In Benchmark Leak, Highlighting Samsung’s Lack Of Interest In This Lineup
Samsung is once again treating its non-flagship smartphones with poor specifications, with the Galaxy S26 FE shipping with an inferior SoC
Gemma 4 and what makes an open model succeed
Having written a lot of model release blog posts, there’s something much harder about reviewing open models when they drop relative to closed models, especially in 2026. In recent years, there were so few open models, so when Llama 3 was released most people were still doing research on Llama 2 and super happy to get an update. When Qwen 3 was released, the Llama 4 fiasco had just gone down, and a whole research community was emerging to study RL on Qwen 2.5 — it was a no brainer to upgrade. Today, when an open model releases, it’s competing with Qwen 3.5, Kimi K2.5, GLM 5, MiniMax M2.5, GPT-OSS, Arcee Large, Nemotron 3, Olmo 3, and others. The space is populated, but still feels full of hidden opportunity. The potential of open models feels like a dark matter, a potential we know is huge, but few clear recipes and examples for how to unlock it are out there. Agentic AI, OpenClaw, and everything brewing in that space is going to spur mass experimentation in open models to complement the likes of Claude and Codex, not replace them. Especially with open models, the benchmarks at release are an extremely incomplete story. In some ways this is exciting, as new open models have a much higher variance and ability to surprise, but it also points at some structural reasons that make building businesses and great AI experiences around open models harder than the closed alternatives. When a new Claude Opus or GPT drops, spending a few hours with them in my agentic workflows is genuinely a good vibe test. For open models, putting them through this test is a category error. Something else to be said about open models in the era of agents is that they get out of the debate of integration, harnesses, and tools and let us see close to the ground on what exactly is the ability of just a model. Of course, we can’t test some things like search abilities without some tool, but being able to measure exactly the pace of progress of the model alone is a welcome simplification to a systematically opaque AI space. Share The list of factors I’d use to assess a new open-weight model I’m considering investing in includes: Model performance (and size) — how this model performs on benchmarks I care about and how it compares to other models of a similar size. Country of origin — some businesses care deeply about provenance, and if a model was built in China or not. Model license — if a model needs legal approval for use, uptake will be slower at mid-sized and large companies. Tooling at release — many models release with half-broken, or at least substantially slower, implementations in popular software like vLLM, Transformers, SGLANG, etc due to pushing the envelope of architectures or tools. Model fine-tunability — how easy or hard it is to modify the given model to your use-case when you actually try and use it. The core problem is that some of these are immediately available at release, e.g. general performance, license, origin, etc. but others such as tooling take day(s) to week(s) to stabilize, and others are open research questions — with no group systematically monitoring fine-tunability. In the early era of open models, the days of Llama 2 or 3 and Qwen pre v3.5, the architectures were fairly simple and the models tended to work out of the box. Some of this was due to the extremely hard work of the Llama, Qwen, Mistral, etc. developer teams. Some is due to the new models being genuinely harder to work with. When it comes to something like Qwen 3.5 or Nemotron 3, with hybrid models (either gated delta net or mamba layers), the tooling is very rough at release. Things you would expect to “just work” often don’t. I’ve been following this area closely since we released Olmo Hybrid with a similar architecture, and Qwen 3.5 is just starting to work well in the various open-source tools that need to all play nice together for RL research. That’s 1.5 months after the release date! This is just to start really investing more into understanding the behavior of the models. Of course, others started working on these models sooner by investing more engineering resources or relying on partially closed software. The fully open and distributed ecosystem takes a long time to get going on some new models. All of this is lead-in for the most important question for open models — how easy is it to adapt to specific use-cases? This is a different problem for different model sizes. Large MoE open-weight models may be used by entities like Cursor who need complex capabilities in their domain, e.g. Composer 2 trained on Kimi K2.5. Other applications can be built on much smaller models, such as Chroma’s Context-1 model for agentic search, built on GPT-OSS 20B. The question of “which models are fine-tunable” is largely background knowledge known by engineers across the industry. There should be a thriving research area here to support the open ecosystem model. The first step is to understand characteristics of different base and post-trained models to understand what they look like. The second step is to tune pretraining recipes for open models so they’re more flexible. Interconnects AI is a reader-supported publication. Consider becoming a subscriber. Subscribe For The ATOM Project and other Interconnects endeavors, we’ve put in substantial effort to measuring adoption trends in the open ecosystem. Everything takes a long time to unfold after a model is first publicly available — and adaptability is why. What we know for sure now, when Qwen has been going from strength to strength with its releases, is that technical staff across the industry has gotten comfortable working with Qwen models. Countless research methods and datasets were made to work with Qwen. It’ll take patience for any other model family to get to this point — a patience I’m not sure many open model builders have. This takes us to Gemma 4, Google’s latest open models. Gemma 3 was released more than a year ago, in March of 2025, and is a bit underrated. Gemma 4 comes in 4 sizes for now, with a bigger, MoE model of over 100B total parameters rumored but not released yet. The models we have today come in sizes of ~5B dense, 8B dense, 26B total 4B active MoE, and 31B dense. I’m most excited that they’re finally adopting a standard Apache 2.0 open source license. This’ll massively boost adoption. The standard of better licenses for strong open-weight LLMs was set by mostly Chinese open model labs in the last 1-2 years, and now U.S. companies are following suit. I will personally be so happy if the horrible Llama licenses and Gemma terms of service were an ~18-month transient dynamic of the industry being nervous about releasing strong open models. The Gemma 4 scores look very solid, the small models have incredible benchmark scores (especially in general domains like LMArena) and the 31B model rivals the recent Qwen 3.5 27B, which is the leading member of that class. The ~30B size range is an important one, as it’s accessible both to researchers and to enterprises looking to deploy the model in real use-cases. Where the 7B model scale is the default for tinkering and research, a 30B model is the default for seeing if an open model can unlock substantial value in your specific workflow — a good mix of intelligence, low price, tractability for downstream training, etc. This takes us back to the above adoption criteria I mentioned for open models and the bigger question — do I think Gemma 4 will be an overwhelming success? Previous Gemma models have been plagued by tooling issues and poorer performance when being finetuned. Gemma 4’s success is going to be entirely determined by ease of use, to a point where a 5-10% swing on benchmarks wouldn’t matter at all. It’s strong enough, small enough, with the right license, and from the U.S., so many companies are going to slot it in. I’m cautiously optimistic that Gemma 4 is going to work better here. Winds are shifting for open models built in America. We saw GPT-OSS go through a bumpy launch to become an overwhelming success. There’s a collective energy around the likes of Reflection, Arcee, Nemotron, Gemma, Olmo, and peers that show substantial demand for building new stacks around open models. There’s capital to be spent on AI stacks across the economy by those who want more ownership of everything, including the model. After launching The ATOM Project 240 days ago, the conversation is shifting into the next stage. Summer of 2025 was a crisis moment where the U.S. AI scene realized it can’t wait and figure out open models after building AGI. The two markets will capture different areas and proceed in parallel. Now that more companies in the U.S. are releasing strong models, we need to improve the ecosystem so that these models are easy to use, understand, and build value around. It’s the hard work to build another inflection point in these adoption plots I’ve been updating consistently, but that’s the work to be done. Join me in it. More data coming soon! Here’s a sneak peek:
17 wide receiver draft prospects who fit Packers production benchmarks
There are production trends among Green Bay's previous draft picks at receiver, and there is a healthy list of prospects who fit the bill in 2026.
Hingham Schools Eye Business Curriculum Expansion as Class Size Concerns Mount
HINGHAM - March 23, 2026 - The Hingham School Committee moved to modernize its high school offerings Monday night, reviewing a comprehensive plan to expand the “small but mighty” Business Department while simultaneously grappling with data that places Hingham as an outlier in elementary class sizes. Director of Social Studies Andy Hoey presented a two-year program review that highlighted the department’s success in community partnerships but warned that high student demand is testing current staffing limits. Meanwhile, committee Chair Jen Benham presented independent research showing Hingham is the only benchmark district with second-grade classes hitting a 25-student cap, sparking a directive for the Policy Subcommittee to review class-size standards. Editors note: South Shore News is going paid in April, subscribe now to make sure you don’t miss a story. Reach out for group or organizational pricing. Subscribe The Full Story The meeting opened with a look into the future of Hingham High School’s (HHS) curriculum. Andy Hoey, who oversees the Business Department, presented the first formal program review for the department since 2009. [15:09] Hoey characterized the department as a vital elective hub that currently serves between 14% and 27% of the HHS student body in any given year. [31:31] “We are very proud of our business department here at Hingham High School... while we’re very proud of it, we are always striving to make improvements and enhancements.” [15:55] — Andy Hoey, Director of Social Studies A key highlight of the presentation was the announcement of a new AP Business with Personal Finance course launching next year, pending enrollment. [20:08] The review team found that while the program is “well-aligned” to national standards, it remains a high-school-only offering. Hoey outlined an action plan to explore “vertical articulation,” potentially introducing business or financial literacy units at the middle school level to fill a “gap” identified between fifth and eighth grade. [41:48] However, the department faces “smoke and mirrors” regarding student GPA, according to Hoey. Currently, most business courses are “unleveled” and do not count toward a student’s GPA, which survey data suggests may deter some students from enrolling. [28:15] The committee discussed the possibility of offering Honors or College Prep (CP) credit for advanced courses, such as investing or marketing, to better reflect their academic rigor. [38:27] The discussion took a sharp turn toward elementary education during an enrollment update. Superintendent Dr. Katie Roberts noted a net gain of four students mid-year, but it was Jen Benham’s “rabbit hole” data analysis that caught the committee’s attention. [50:35] Benham researched 13 benchmark towns—representing 34 elementary schools—and found that Hingham was the only district with a second-grade section reaching 25 students this year. [51:36] Benham’s findings suggested that while many districts lack formal class-size policies, Hingham’s current cap of 25 for second grade may be too high. “I think if we could at least have 23 at a cap for grade two,” Benham suggested, noting the importance of that grade level for literacy development. [53:09], [54:55] Dr. Roberts cautioned that smaller classes require more staff, which is a significant hurdle following a budget season where nine general education positions were cut across the district. [55:56] The committee also finalized the 2025 Town Report, which was updated to be more “student-centered” and highlight staff early in the document. [11:12] Additionally, a series of food service policies were approved to align with Massachusetts’ universal free school meal mandate, including new protocols for meal modifications due to allergies or religious needs. [59:23] Share Why It Matters The meeting underscored a tension between Hingham’s desire to offer elite, specialized programming and the foundational pressure of class sizes. The expansion of the Business Department—including potential Microsoft certifications and AP courses—aims to prepare students for a “real world” where credit scores and financial literacy are paramount. [46:20] However, the revelation that Hingham’s second-grade classes are larger than those in all neighboring benchmark towns suggests a potential strain on early education that could impact long-term student performance if not addressed by the Policy Subcommittee. Official Minutes & Data Key Motions & Votes Motion: To approve the Hingham Public Schools 2025 Town Report. [13:16] Vote: Unanimous (7-0) ([01:13:28]) Motion: To approve the Memorandum of Agreement (MOA) for summer counselor work regarding course placement. [57:33] Vote: Unanimous (7-0) ([57:50]) Motion: To approve Policy EFBA (School Food and Nutrition/Meal Modifications). [01:00:32] Vote: Unanimous (7-0) ([01:00:44]) Motion: To approve Policy EFC (Universal Free School Meals). [01:00:54] Vote: Unanimous (7-0) ([01:01:07]) Public Comment No members of the public spoke during the meeting or via the remote link. [01:02:30] What’s Next The Policy Subcommittee will take up the review of elementary class-size caps over the next year to determine if the Grade 2 limit should be lowered from 25 to 23. [52:18] The Business Department will monitor enrollment for the new AP Business course and continue exploring dual-enrollment partnerships with Quincy College. [38:13] Source Video: Harbor Media South Shore News is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Subscribe
Consumer Discretionary - Casino Operator Stocks Q4 Recap: Benchmarking Golden Entertainment (NASDAQ:GDEN)
Quarterly earnings results are a good time to check in on a company’s progress, especially compared to its peers in the same sector. Today we are looking at Golden Entertainment (NASDAQ:GDEN) and the best and worst performers in the consumer discretionary - casino operator industry.
Google Claims Android Is Now Faster Than iPhone for Web Browsing
Google today said that Android has set a new record for mobile web performance, making it the fastest mobile platform for web browsing. The newest Android devices have set new records on web performance benchmarks like Speedometer and LoadLine, which Google attributes to "deep vertical integration across hardware, the Android OS, and the Chrome engine".
Google says Android with Chrome is ‘fastest mobile platform for web browsing’
Google shared today that Android and Chrome have set “new performance records” for mobile web browsing in two benchmarks...
Amazon all-day delivery test could reshape logistics
As part of the plan, Amazon benchmarked similar services from Walmart, Instacart, DoorDash, and UberEats, internal document shows.
The S&P 500 is flashing a warning sign. How to protect your portfolio.
The S&P 500 crossed below its 200-day moving average last week, a signal that the benchmark index could be headed into a downward trend.
Apple's M5 Pro & M5 Max easily beat the RTX 5090 in PugetBench creator benchmarks
MacBook's are very popular among content creators and the two new SoCs M5 Pro as well as M5 Max manage excellent results in the PugetBench tests for Adobe apps and DaVinci Resolve. They also clearly beat the mobile GeForce RTX 5090.
Intel Core Ultra 7 270K Plus & Core Ultra 5 250K Plus Official Benchmarks: Faster Than 14900K & Comparable To Ryzen X In Gaming, Much Faster In Apps, Superb Value
Intel shared official benchmarks of its Core Ultra 7 270K Plus & Core Ultra 5 250K Plus in gaming, apps, and more, against AMD Ryzen CPUs.
Exclusive: $599 MacBook Neo easily beats $1,119 Dell laptop on Windows 11 benchmarks - yes, macOS may well be the best platform ever for Microsoft's OS
Despite running a mobile CPU and Windows in a virtual machine, the MacBook laptop excelled in single-core CPU performance
AMD "Medusa Point" APU Early Benchmarks Match "Strix Point" at Half the Clock Speed
AMD is preparing to launch its "Medusa Point" APU in early 2027. However, more benchmarks are emerging to showcase what the actual SoC can do as AMD and its OEM partners test the chip. In the latest Geekbench v6 run, AMD's 10-core, 20-thread "Zen 6" chip appeared with the AMD Engineering Sample numb...
Paradigm Launches Surface, an AI-Native Platform That Connects Culture to Business Performance
Paradigm, the talent and culture intelligence company, today announced the launch of Surface, an AI-native platform that gives talent and culture leaders a unified view of what's shaping their organizations, benchmarks their efforts against peers, and translates insights into action with the industry's first agentic AI.
Apple’s Smaller M5 Pro MacBook Pro Suffers From Severe Thermal Constraints, With Larger 16-Inch Model Being 30% Faster With Just 3 Extra Cores
In one CPU-focused benchmark, the 14-inch MacBook Pro is unable to cool the M5 Pro properly, leaving the 16-inch M5 Max model to register a major performance difference
Here's what the Samsung Galaxy S26 Ultra's Privacy Display does to its battery life
Benchmarks and tests to see what's going on under the surface
Scientists built the hardest AI test ever and the results are surprising
As AI systems began acing traditional tests, researchers realized those benchmarks were no longer tough enough. In response, nearly 1,000 experts created Humanity’s Last Exam, a massive 2,500-question challenge covering highly specialized topics across many fields. The exam was engineered so that any question solvable by current AI models was removed. Early results show even the most advanced systems still struggle — revealing a surprisingly large gap between AI performance and true expert-level knowledge.
Healthcare spending surges past state benchmark
Spending on healthcare in Massachusetts jumped by 5.7 percent in 2024, according to new state report
Nuts & Bolts: Lightning host Blue Jackets on Women in Sports Night
The Bolts are back on home ice to kickoff a three-game stand at Benchmark International Arena