Latest News

NVIDIA, the leading technology company in artificial intelligence (AI), has partnered with the Korea Advanced Institute of Science and Technology (KAIST) to launch a joint research lab focused on developing AI models tailored to South Korea’s industries. The new laboratory will be established at KAIST’s Kim Jaechul Graduate School of AI in Seoul, aiming to advance agentic AI for the country’s future needs.

The collaboration brings together NVIDIA’s expertise in full-stack AI solutions with KAIST’s world-class scientific talent and research infrastructure. This partnership is expected to create a comprehensive academic AI research program that will drive innovation in Korea’s technology sector.

According to Bill Dally, chief scientist and senior vice president of research at NVIDIA, ‘Korea is home to leading AI researchers and has one of the most advanced technology ecosystems worldwide.’ The joint lab aims to provide a foundation for next-generation AI research, accelerating the development of models and agent systems tailored to Korea’s industries, language, and future.

Hyunwoo Kim, an incoming faculty member at KAIST’s Kim Jaechul Graduate School of AI, will lead the joint NVIDIA-KAIST lab upon joining the institution. He emphasized that ‘AI research is entering a new era requiring frontier talent, large-scale infrastructure, and deep collaboration across academia and industry.’ The partnership between NVIDIA and KAIST aims to address these needs by fostering ambitious work in Korea’s AI sector.

The joint lab will be equipped with state-of-the-art computing resources from local NVIDIA Cloud Partners. This infrastructure will provide researchers with direct access to the latest NVIDIA AI technology, enabling them to develop models optimized for the Korean language and specific use cases. The collaboration also includes funding for at least 10 KAIST researchers annually, who will receive internship opportunities at NVIDIA.

Additionally, NVIDIA plans to hire exceptional Korean researchers for full-time positions, creating new pathways for Korea’s top AI talent to pursue ambitious research and build long-term careers in the field. This initiative is expected to deepen global collaboration between academia and industry while attracting and retaining top AI scientists in Korea.

The $300 million partnership will include a significant investment of $50 million per year over an initial five-year period, providing substantial compute contributions for researchers at KAIST. The joint lab’s priorities will focus on developing models optimized for the Korean language and specific use cases using NVIDIA Nemotron open models to advance the country’s AI capabilities.

The collaboration between NVIDIA and KAIST marks a significant step towards accelerating AI innovation in Korea. By establishing this joint research laboratory, both parties aim to create a pipeline from academic discovery to enterprise and national AI deployments, ultimately driving economic growth and technological advancements in South Korea.

Amazon has introduced a new policy requiring third-party sellers to label any product images or videos that contain ‘AI-generated people’. This change comes after a recent law in New York mandated greater transparency around synthetic performers in ads.

The move is part of a growing trend as states begin to take steps towards requiring companies to disclose when advertising content has been created using AI. In fact, California recently started requiring large AI providers to embed watermarks in AI-generated images, video or other content.

Amazon’s policy change specifically addresses the use of photorealistic AI-generated people in advertisements. This requirement doesn’t apply to content featuring TV, movie and video game characters, or any real person who has been altered using AI.

The New York law, which took effect last month, requires companies to disclose if ‘synthetic performers’ are used instead of human actors in advertising. Governor Kathy Hochul described it as a ‘first-in-the-nation’ law, and her office noted that without notice about the content’s authenticity, AI-generated synthetic performers can make it harder for people to tell fact from fiction.

Amazon is increasingly using AI across its portfolio. The company has optimized listing titles and details so they’re more likely to be found by AI systems. They’ve also invested in a rebranded assistant called Alexa for Shopping, which uses AI to help customers shop.

More Amazon third-party sellers are now using AI tools like those from Meta to generate text, images and other content for their listings. This trend is partly driven by the company’s efforts to make listing creation easier with AI-powered features. With these tools, businesses can create more appealing and effective ads without needing extensive expertise in graphic design or video production.

The policy change will see Amazon add an indicator to its website informing consumers that images or other content feature AI-generated people ‘where applicable’. However, it’s unclear what criteria the company will use when deciding which listings get this label. It might be based on factors like how often a seller uses these tools or whether they clearly disclose the use of AI in their advertising.

Amazon didn’t provide an immediate comment about the new policy. The company relies heavily on third-party sellers, who account for more than 60% of goods sold on its marketplace. There’s no federal law requiring companies to disclose when advertising content has been created using AI.

Some tech giants have already started taking steps towards greater transparency around AI-generated content. Meta and TikTok have added labels to videos and images uploaded to their platforms, while YouTube includes such labels too. However, both TikTok and Meta have faced criticism for not doing enough to clearly label ads featuring AI-generated influencers selling dubious products.

TikTok has said it’s taken steps to ban accounts making misleading health claims, and Meta noted its efforts to label AI videos more clearly. The move by Amazon is a significant step towards greater transparency around AI-generated content in the advertising industry. With this change, consumers will be better equipped to make informed decisions when shopping online.

AI assistants like those from Google are becoming increasingly important for businesses looking to maximize their online presence. By using these tools, sellers can create high-quality ads that grab customers’ attention without requiring extensive expertise or resources.

Anthony Pompliano, a well-known expert in tech and business development, has pointed out that Meta excels at humanizing work, surpassing many large technology companies. This approach is not only evident in its products but also in how it communicates with users.

Pompliano recalled his time at Facebook when he noticed that product experiences couldn’t quantitatively change user emotions until the message was signed ‘From all of us at Facebook.’ After this humanized touch, emotional metrics began to shift. This approach was then scaled across various products and communications.

Today, Meta’s emphasis on ‘We bet on humanity’ stands in stark contrast to AI model labs that often discuss replacing jobs and controlling digital weapons. As a result, capital and user attention are more easily drawn to communicators capable of emotional connection.

Pompliano has extensive experience with business development at Facebook early in his career, later shifting to cryptocurrency investment and media, founding the Pompliano Company. He is adept at extracting insights from internal practices related to communication and user psychology.

Meta’s strategy involves expanding the humanizing technique of ‘collective signing’ across its entire product line and brand narrative. This approach has been validated through early A/B testing and reduces regulatory and public opinion pressure costs by creating emotional resonance. Additionally, it hedges against technological threat narratives with its ‘bet on humanity’ approach to attract advertisers and retain users.

Similar cases can be seen in Apple’s long-term emphasis on storytelling and designer stories, Nike’s personal narratives of athletes, and OpenAI’s early slogan shifting from a focus on benefiting all humanity to more technical expressions. Currently, tech communication is undergoing a phase of differentiation between AI threat narratives and humanized counter-narratives.

This shift represents a change in pricing power: user attention and brand trust are moving from pure technical capability to emotional connection capability. The mechanism behind this is that collective signing and ‘betting on humanity’ reduce feelings of alienation, making communication itself a key differentiator rather than an accessory.

Reports of Microsoft services going haywire started flooding in around 11 a.m. ET, with users taking to social media to vent their frustrations.

Downdetector’s outage map reveals that the problems are widespread and not limited to Teams and OneDrive; Xbox Live, SharePoint, Power Automate, and the Microsoft 365 Admin Center are also affected.

Microsoft has released an update on its service status page, confirming that it’s zeroed in on specific network paths as part of its investigation. The team is analyzing diagnostic data from affected infrastructure to pinpoint the source of the issue and devise a plan for mitigation.

In a subsequent post on X, Microsoft shed more light on what’s going on: it has rerouted connections across the most heavily impacted portion of its infrastructure and is working to scale these actions to minimize the problem’s effects.

While reports seem to have tapered off compared to earlier in the day, it appears that things are far from back to normal just yet. Microsoft continues to keep a close eye on the situation for any updates.

The company has kept users informed through its service status page and social media channels with a clear message: for the vast majority of users, services should be up and running by 4:07 p.m. ET (1:07 p.m. PT).

It’s not just Microsoft 365 services that are struggling; other areas like Power Automate are also impacted, which could have significant implications for businesses relying on these tools.

Microsoft is working diligently to resolve the issues as quickly and efficiently as possible, but it remains uncertain how long it will take to get everything back up and running smoothly. One thing’s clear: users will keep a close eye on the situation for any updates.

The Isle of Palms Police Department is getting a new vehicle, one that’s making waves in the community: a sleek Tesla Model S. According to police officials, this electric car will be used for official business and community events.

After equipping it with emergency lights, sirens, and other safety gear, the Tesla is now ready for duty. A big thank you goes out to Jeff and Melissa Polluta, who donated this eco-friendly cruiser and put smiles on everyone’s faces.

Police Chief Kevin Cornett will be driving the new Tesla when he needs an official ride. He’s thrilled about the donation because it lets his officers connect with residents in a more interactive way - which is exactly what they need to build stronger ties between law enforcement and the community.

The IOP PD expects this shiny new car to make its presence known at community events, giving officers a chance to engage directly with locals. With every function and gathering, these police officers will be one step closer to making their neighborhood an even safer place.

Schnucks Markets has announced plans to launch an artificial-intelligence-powered shopping assistant, set to become available on its app and website later this summer. The tool is being rolled out in partnership with VitalityIP, a digital nutrition guidance and health data platform that leverages AI technology.

The new agentic shopping assistant will provide customers with personalized recommendations based on their lifestyle preferences, including nutrition guidance, meal ideas, and product suggestions. This move marks the latest development in the growing trend of grocers adopting AI-powered tools to enhance customer experience.

Schnucks Chief Data and Information Officer Tom Henry explained that the company’s decision to introduce this feature was driven by customer demand for trusted guidance on various aspects of their shopping trips, such as meal planning, tracking health goals, balancing special dietary needs, and budgeting. The new platform aims to help customers make informed grocery purchase decisions aligned with their health and lifestyle objectives.

The AI assistant will be powered by over 6 billion lines of data from VitalityIP’s proprietary database, which was co-developed using customer and store inventory information. Schnucks emphasized that it maintains ownership of the customer relationship, shopper data, and brand experience while leveraging VitalityIP’s expertise in AI technology.

Schnucks is set to be the first retailer to debut the VitalityIP platform, with several other grocery players having recently introduced their own AI-powered shopping assistants. This includes Albertsons, Kroger, and Associated Wholesale Grocers, among others.

The development of artificial intelligence has been one of the most significant technological advancements in recent years, with many hailing it as a revolutionary force that will transform human life. However, others have raised concerns about its potential impact on jobs, privacy, and even humanity itself. The question is not whether AI is good or evil, but rather what kind of system underlies its development, who owns this technology, for what purpose it is used, and in whose social class interests it operates.

The technological revolution has the potential to free human beings from arduous and repetitive work, reduce working hours, and increase productivity. This could lead to an improvement in everyone’s quality of life. Karl Marx recognized that technology was capable of serving society as a whole if used rationally and fairly. However, he also noted that under capitalism, technology is primarily employed to maximize profits and accumulate capital.

The problem with AI is not the technology itself but rather the relations of production that control it. Under capitalism, large corporations use AI to serve their interests at the expense of workers and society. The evidence for this can be seen in the rise of generative AI tools between 2022 and 2024. Tens of thousands of writers, designers, and translators have lost their contracts with companies not because they are less competent but because the algorithm is cheaper.

The use of AI has led to a transformation of intellectual production into a cheap commodity to increase profit margins. Companies such as CNET replaced their journalists with AI models to write articles, followed by BuzzFeed and others. This phenomenon can be observed in several manifestations including the dismissal of workers or reduction of wages through automation, increased surveillance of employees governed by algorithms, monopolization of data and knowledge by giant technology companies, and concentration of wealth and technological power in the hands of a small number of global monopolies.

A worker in an Amazon warehouse does not face a human manager but rather an algorithm that measures performance by the second, determines their picking quota, and automatically issues termination warnings if productivity declines. This has been documented by journalists from The Verge and The Guardian in extensive reports on what is known as ‘algorithmic management.’ Instead of humanity controlling technology, people begin adapting to the rhythm of the machine, living to serve it rather than the other way around.

The Marxist analysis of imperialism and monopoly suggests that the continuous development of capitalism leads to the concentration of capital in the hands of a small minority. This results in artificial intelligence becoming part of a global monopolistic system instead of being used as a means of improving human lives. Marxists also connect AI to the concept of ‘alienation.’ Under capitalism, workers become separated from both their labor and its product.

The feeling of alienation intensifies with technological development as people are transformed into mere appendages of machines, algorithms, and digital systems. This is precisely the danger posed by complete reliance on artificial intelligence without self-development or critical thinking. It may lead human beings to a state of intellectual laziness and cognitive dependence. Major capitalist elites seek to turn technology into an instrument that weakens individuals’ capacity for analysis and independent thought.

The issue with AI extends beyond labor and economic exploitation, also affecting the environment and natural resources. Modern artificial intelligence requires massive data centers that consume enormous amounts of electricity and water to cool servers. Additionally, rare minerals such as lithium, cobalt, and copper are extracted for manufacturing electronic chips. The extraction of these materials has devastating consequences on local communities.

For example, approximately 70% of the world’s cobalt is extracted from the Democratic Republic of Congo. Amnesty International has documented child labor in these mines where children under 12 work in deadly conditions for a few dollars a day while profits flow through supply chains ending at Apple, Samsung, and Nvidia headquarters. Lithium extraction operations deplete groundwater in Chile’s Atacama Desert and displace Indigenous communities from their ancestral lands.

The picture is not much different when it comes to data. Large AI models are trained on user-collected data without compensating them, then sold at prices that governments in the Global South cannot afford. Africa, South Asia, and Latin America export minerals, data, and human labor while importing a monopolized product over which they have no control.

The question is not only whether artificial intelligence consumes vast resources but also why these resources are consumed, who decides how to use them, and who bears the real cost of consumption? Marxists argue that capitalism treats nature as an open source of profit rather than as the foundation for sustaining human life. Production under capitalism is organized according to profitability and competition.

As a result, enormous resources may be wasted on applications whose primary objective is increasing profits such as targeted advertising, financial speculation, consumer surveillance, and corporate races toward monopoly while millions suffer from shortages of water, electricity, and basic services. This does not mean that communists reject technology or industrial development due to energy consumption but rather the chaotic and unplanned use of resources.

The subordination of technology to profit logic places a burden on poor peoples and workers to bear pollution costs and resource extraction. The transformation of scientific progress into a monopolistic instrument instead of collective possession weakens humanity’s capacity for analysis and independent thought. As long as the value of technology is measured by profits generated rather than what it contributes to humanity, the real question will not be ‘How far will artificial intelligence develop?’ but who owns this development and who pays its price?

OpenAI has launched its new ‘Health’ feature for users of its popular chatbot, ChatGPT. The feature allows users to securely connect their health information and use it to inform conversations with the AI assistant. This marks a significant development in the integration of healthcare data into conversational AI systems.

The Health feature is currently available to logged-in US-based users aged 18 and older on web and iOS platforms, across all plans including Free, Go, Plus, and Pro. Users can access the feature by opening ChatGPT’s sidebar or More menu and selecting ‘Get started’.

According to OpenAI, more than 300 million people use ChatGPT each week with health-related questions, often scattered across patient portals, medical records, apps, and wearables. The new Health feature aims to provide a comprehensive view of users’ health information by allowing them to connect relevant data from various sources.

With the user’s permission, ChatGPT can draw on connected health information to help compare lab results with prior tests, summarize changes since their last appointment, or explore how sleep and activity relate to their routine. This can reduce the need for repeated gathering of details, making users feel more informed and empowered in managing their healthcare.

OpenAI has implemented layered privacy and security safeguards to protect connected health information. Conversations that use this data are not used to train its foundation models or target ads, regardless of the model-training setting chosen by the user. Connected medical records and Apple Health information receive additional encryption protections.

The company continues to improve its models behind health conversations in ChatGPT through dedicated training. GPT-5.6 Sol is OpenAI’s strongest model yet for handling complex health-related tasks, building on progress made with stronger performance on reasoning across details, clear communication, and careful judgment.

Physicians extensively tested the Health feature before release to measure real-world model performance and safety with connected health data. While these results represent meaningful progress, ChatGPT can still make mistakes and does not replace qualified medical professionals’ care and judgment. Users are encouraged to verify important information and discuss decisions with their healthcare provider.

The user has control over when ChatGPT uses the connected health information. By default, the AI assistant asks for permission before using this data to personalize a response. Users can approve each request or choose to always allow access, which turns off these permission prompts.

Health conversations are stored in the left sidebar of ChatGPT, where users can manage their accounts and review recent data and trends. Connected information may not be complete or current, so it’s essential for users to check important details against the original source.

When connecting health information, users should ensure that sharing is turned on with any wearable, fitness, or nutrition app linked to Apple Health. ChatGPT can use available information from these apps when allowed by the user.

Once connected, users can review conditions and medications, remove irrelevant data, and add details such as family health history. It’s crucial for users to tell ChatGPT about changes in their health status and verify important details against original sources.

The Health feature allows users to ask questions anywhere within ChatGPT, with the AI assistant using connected health information based on user permissions. Users can change these settings at any time in Settings > Plugins > Health or add @Health to a message to explicitly request health context.

OpenAI emphasizes that its primary goal is not to replace medical professionals but rather support users’ healthcare decisions by providing accurate and relevant information. The company continues to test safeguards through dedicated red-team exercises focused on connected information in the Health feature, using findings to strengthen protections.

Google has launched the AI & Economy ATLAS, an ongoing study that aims to provide a rich understanding of how artificial intelligence is being adopted and used in the economy. The first iteration of this large-scale project focuses on de-identified data from 15 million human-AI interactions across various Google products and tools, including the Gemini App, AI Mode, and the Gemini API. These platforms are used by over 1 billion people monthly, spanning more than 150 countries, 140 languages, 800 occupations, and 4,000 tasks.

Advisors are constantly looking for ways to save time and focus on high-value tasks. One key area of focus is automating administrative duties, which can be both tedious and time-consuming. According to Adam Blumenthal, Chief Growth Officer at OneSeven, there’s a pressing need to automate these tasks.

Blumenthal identifies five areas where automation can make a big impact: data entry, document management, client communication, compliance monitoring, and reporting. By using technology to handle these tasks, advisors can reduce the risk of human error, save time, and improve overall efficiency. This isn’t just about cutting costs; it’s also about providing better services for clients.

OneSeven’s operations are a testament to what can be achieved through automation. They’ve optimized their processes by leveraging technology in key areas. Advisors who partner with OneSeven benefit from this expertise, which enables them to focus on deeper client relationships and more valuable work.

Blumenthal emphasizes the importance of being proactive about automation rather than waiting until it becomes necessary due to increased workload or regulatory requirements. By staying ahead of these challenges, advisors can maintain a competitive edge and deliver better results for their clients.

Mavericks Inc. has just rolled out an update to its AI video generation service NoLang, adding a shape decoration feature that lets users add borders and background colors to text, images, and videos without needing design expertise.

NoLang’s new feature is designed to make it easier for companies to produce high-quality videos in-house by simplifying the process of decorating their content. This move comes as more Japanese businesses shift towards producing their own videos instead of outsourcing them.

A recent survey found that 69.6% of Japanese companies now produce videos internally, a significant jump from last year’s 29.7%. As this trend continues to grow, NoLang is positioning itself to meet the increasing demand for in-house video production tools.

The average cost for outsourced video production comes out to around ¥815,000 ($4,990), with a median of ¥540,000 ($3,305). However, costs can increase when companies add decorative elements like graphic design and narration insertion. The burden isn’t just financial – it’s also in the time-consuming process of adding decorations.

According to data from BGM Now, techniques like framing text and images to guide the viewer’s eye or highlighting important points with decorations are crucial for preventing drop-offs, especially in short social media videos where about 20% of viewers may leave within the first 10 seconds. NoLang addresses this need by allowing users to create decorated and highlighted videos without needing design skills.

The shape decoration feature is available in NoLang’s video editing interface, where users can select from four types: rectangle, circle, heart, and star. This selection process doesn’t require specialized coordinate manipulation or path editing; instead, it lets users easily add frame decorations to text, images, and videos.

Users can toggle borders between ‘none’ and ‘present,’ with the option to adjust border color and width freely. They also have the ability to set background colors, enabling various expressions such as speech-bubble-style decorations surrounding text or framing images and videos with heart or star shapes.

The shape feature’s universal application across multiple setting locations within the video editing interface makes it easy for users to finish their videos with a consistent design tone throughout. This is especially beneficial in scenarios where short videos intended for social media, product introduction videos, internal training manuals, and entertainment content require decoration and highlighting.

Mavericks has been rapidly expanding NoLang’s editing functions, including the recent addition of a ‘transition’ function that adds motion to scene changes and a ‘seam fade’ toggle function. Users can choose from 18 types of transitions, such as slides, zooms, and wipes, to match their video’s tone.

The company also enhanced its tempo adjustment function by adding features for batch-applying intervals between scenes across all scenes and inserting pauses at any position within the script text of a scene. This series of functional enhancements aims to remove editing barriers from both AI and UI perspectives.

Japan’s domestic video production service market is estimated at ¥423.8 billion ($2.6 billion) for fiscal 2024, representing a significant increase from the previous year. The market is forecasted to expand to ¥540 billion ($3.3 billion) by fiscal 2027.

A Vimeo survey found that video production volume increased in 65% of organizations, with 73% expecting further increases while 48% struggle with scaling production while maintaining consistent quality. Even with AI-generated video production, the burden of corrections and fine-tuning after the initial draft is not insignificant.

PRIZMA’s survey revealed that 76.4% of respondents feel burdened by corrections and fine-tuning, and 60.8% make corrections four or more times on average. An Applib survey found that ‘complex operation methods’ were cited as a concern by approximately 41.1% of respondents.

Mavericks plans to continue expanding the functions of its video editing interface with NoLang, aiming to create an environment where anyone can produce videos that decorate and highlight text and images regardless of design skills.

The company intends to meet the growing decoration needs amid the shift to in-house video production and support video creation for a wide range of users. As more Japanese businesses move towards producing their own content, tools like NoLang’s new shape feature will become increasingly essential.

Mavericks’ rapid expansion of NoLang’s editing functions is designed to simplify the process of creating high-quality videos. The recent addition of transition and seam fade features has made it easier for users to add motion to scene changes without needing extensive design expertise.

The update includes 18 types of transitions, such as slides, zooms, and wipes, which can be matched to a video’s tone. Users can also batch-apply intervals between scenes across all scenes, reducing the time-consuming process of adding decorations.

Mavericks is positioning NoLang to meet the increasing demand for in-house video production tools by expanding its editing functions and simplifying decoration processes. The new shape feature is just one example of how the company aims to support users with varying levels of design expertise.

The domestic video production service market in Japan is expected to continue growing, reaching an estimated ¥540 billion ($3.3 billion) by fiscal 2027. As more businesses shift towards producing their own content, companies like Mavericks are working to provide tools that meet the increasing demand for high-quality videos.

Mavericks’ commitment to expanding NoLang’s functions and supporting in-house video production is part of its larger goal to make video creation accessible to anyone, regardless of design skills.

The challenges facing modern researchers are unprecedented. From simulating the intricate dynamics of fusion plasma to exploring vast search spaces for new materials and making sense of exabytes of data from experimental facilities, scientists today face extraordinary scale and complexity. Frontier AI can help address these challenges while accelerating groundbreaking scientific discoveries.

In December, Google DeepMind announced its commitment to the White House’s Genesis Mission – a national effort to harness AI and double the pace of American scientific discovery within a decade. The company has since provided early access to AI for science tools at 17 Department of Energy National Laboratories through an in-kind program. Today, at the DOE Genesis Mission Summit 2026, Google DeepMind is expanding this commitment by pledging $40 million in AI tokens and cloud credits for researchers supporting the Genesis Mission.

The expanded commitment provides two key benefits: first, it gives DOE’s Genesis Mission awardees access to Google DeepMind’s frontier AI for science portfolio. This includes AlphaEvolve – a Gemini-powered coding and discovery agent designed for advanced algorithms; AlphaFold 3 – a model predicting protein structure and interactions; AlphaGenome – a tool understanding DNA variation and its impact on biology and disease; WeatherNext – state-of-the-art weather forecasting models; and AlphaEarth Foundations, an AI model mapping our planet in unprecedented detail.

Secondly, Google DeepMind will provide Gemini for Government seats and tokens to tens of thousands of users across the DOE National Laboratories’ operations, research, and management teams. This secure platform supports work from the research bench to specialized user facilities serving the entire scientific community, providing a single foundation that the DOE mission can depend on.

The practical impact of the Genesis Mission is already coming to life across laboratory ecosystems. At Pacific Northwest National Laboratory (PNNL), senior scientist Dr. Henry Kvinge uses AlphaEvolve to map out massive mathematical systems too complex for humans to explore by hand. The AI uncovers hidden connections automatically, fast-tracking discoveries that would normally take years.

Dr. Kvinge notes the importance of abstraction in modern math and how combinatorics offers concrete models making geometry and algebra easier to grasp. He believes AlphaEvolve is perfect for this search due to its ability to automate exploration using LLMs’ broad mathematical knowledge, ‘automate the exploration of countless angles.’ The discoveries are already shaping future research.

At the National Laboratory of the Rockies (NLR), researchers utilize Gemini to fundamentally change how they interact with physical laboratory hardware. Dr. Steven R. Spurgeon leads a pioneering program in autonomous materials discovery using Gemini instruments. He explains that deploying Gemini has cut microscope calibration time from over 90 minutes to about 13 minutes, reducing manual steps needed to focus an image from as many as 50 down to two.

This has given back time and attention to the science itself, enabling genuinely autonomous workflows observing, reasoning, and deciding in real-time. This capability helps explore parts of the material design space that could not be reached through manual operation alone. The Genesis Mission represents a chance to transform research and science across America by providing access to advanced AI tools.

By accelerating breakthroughs across critical energy, security, and scientific challenges, Google DeepMind aims to help scientists overcome these hurdles. To learn more about how these AI capabilities can support your research initiatives, join the upcoming Google Public Sector Summit in October.

LLM inference is often routed through a well-tested path, but building your own runtime can be beneficial when you need to customize or optimize it. This article provides a step-by-step guide on how to build an LLM runtime from scratch using CUDA and C++.

Trust is the foundation upon which the legal profession operates. Every contract, opinion, case analysis, and research document relies on accuracy, authenticity, and accountability. However, with the increasing use of generative AI tools in professional environments, lawyers and researchers face a new challenge: distinguishing between human-created content and AI-generated or modified material.

As AI becomes more prevalent across various areas of legal practice – from summarizing lengthy documents to preparing first drafts – concerns arise about originality, reliability, and professional responsibility. This is where Lynote.ai steps in with its AI detection and humanization platform, designed to help professionals maintain authenticity while leveraging the efficiency of modern AI technology.

For legal professionals, identifying AI-generated content has become increasingly important. A document represents analysis, expertise, and judgment; knowing its origin can uphold credibility when reviewing submissions, evaluating research materials, or managing internal documents. Traditional AI detection tools often rely on surface-level patterns, which may not be effective against sophisticated AI-generated text.

Lynote.ai’s approach differs significantly from these traditional methods. Its AI detection technology analyzes deeper writing patterns rather than just looking for obvious AI signals. This allows the platform to identify content generated by leading AI models like GPT-5 and Gemini while recognizing text modified through AI rewriting or ‘humanization’ tools.

Understanding how AI detectors work provides valuable insight into why advanced detection requires more than basic keyword analysis. Effective AI detection examines linguistic structures, consistency, contextual patterns, and other signals that distinguish machine-generated writing from authentic human expression. Lynote.ai’s multilingual capability is another notable advantage; it can analyze AI-generated content across languages such as English, Spanish, French, Portuguese, and German.

Beyond detection lies the challenge of ensuring AI-assisted writing maintains a natural voice and communicates ideas effectively. This is where AI humanization technology becomes valuable. Basic rewriting tools often operate like traditional article spinners, replacing words without understanding meaning behind the text – resulting in unnatural writing or loss of original message.

Lynote.ai’s AI humanizer focuses on a more advanced approach: analyzing logic and context before transforming content into a natural human-like style. The goal is not to change meaning but improve readability, flow, and authenticity. This can be particularly useful for legal writers preparing educational articles, client communications, internal explanations, or research summaries.

The platform supports outputs from popular AI systems including ChatGPT, Gemini, DeepSeek, and Claude. Its customizable bypass modes and support for more than 80 languages allow users to adapt content according to different audiences and communication requirements. Professionals exploring solutions in this area often compare options when searching for the best humanizer because not every tool provides the same level of contextual understanding.

Lynote.ai focuses on maintaining original intent while creating content that feels naturally written. This combination of AI detection and humanization is especially relevant for industries where trust and professionalism are critical – such as law, where authenticity can make or break a case.

The growth of AI writing assistants has created many opportunities but also a gap between efficiency and authenticity. Many AI platforms excel at generating information quickly but may not evaluate whether that information sounds genuinely human or meets professional communication standards. Traditional AI detectors struggle with edited AI content while basic humanizers produce text that feels artificial because they focus mainly on surface-level changes.

Lynote.ai attempts to address both challenges in one ecosystem: its AI detector helps users evaluate content authenticity, and its AI humanizer transforms AI-assisted drafts into more natural writing. This approach is crucial for industries where trust and professionalism are paramount – like law, where accuracy, originality, and professional integrity are essential.

The future of AI and authenticity in legal communication looks promising with platforms like Lynote.ai leading the way. By combining advanced detection capabilities with intelligent content transformation, professionals can navigate the changing relationship between humans and artificial intelligence more effectively. For lawyers, researchers, and organizations that rely on trustworthy communication, identifying AI-generated content and refining AI-assisted writing can become an important advantage in today’s digital landscape.

A federal judge in San Francisco has approved a record-breaking settlement between AI company Anthropic and authors who accused the firm of downloading their copyrighted books without permission. Judge Araceli Martínez-Olguín signed off on the agreement on July 20, marking it as the largest copyright class action settlement in history.

The lawsuit centered around how Anthropic acquired its book collection – specifically using pirated libraries LibGen and PiLiMi to build its database. The court had previously ruled that training AI models with copyrighted material is lawful under fair use provisions, but this case focused on the method of acquisition rather than the practice itself.

Anthropic’s unauthorized actions have led to a significant payout for authors and publishers whose works were affected. Under the settlement terms, those who can prove their books ended up on Anthropic’s ‘Works List’ are eligible to claim around $3,000 per book – roughly four times the typical minimum award for copyright infringement cases.

Over 91% of eligible titles have already been claimed by authors and publishers, with more than 440,000 books affected. As part of the agreement, Anthropic must delete the pirated files it downloaded from these libraries, effectively removing the infringing materials from its system.

The settlement does not release Anthropic from liability for future lawsuits over what its chatbot generates or new claims going forward. Judge Martínez-Olguín made clear that the agreement does not absolve Anthropic of responsibility for potential harm caused by its AI models. This means authors and publishers may still pursue action against the company in the future.

The court rejected all 54 objections filed by class members and third parties, including requests to expand the list of covered works or add non-monetary remedies like source attribution. The judge deemed these proposals beyond the scope of this lawsuit, sticking to the original agreement between Anthropic and the authors.

With the order finalized on July 20, the case is officially closed – although the court will continue to monitor how the settlement funds are distributed among affected authors. This process should provide some clarity for those impacted by Anthropic’s actions in acquiring its training data.

The University of Houston has been selected by the US Department of Energy (DOE) to lead a nearly $750,000 collaborative research project that will harness artificial intelligence (AI) to accelerate the design of next-generation nuclear and fusion energy systems. Funded through the DOE’s Genesis Mission, this project brings together researchers from UH, Lawrence Livermore National Laboratory, and the University of Pennsylvania to develop AI-powered engineering tools capable of dramatically reducing the time needed to model complex heat-transfer systems critical to advanced nuclear reactors and future fusion technologies.

The award builds on UH’s growing leadership in AI and energy innovation. By combining AI with fundamental laws of physics, researchers aim to improve the accuracy and speed of computer models used to accelerate the development of reliable, carbon-free energy technologies. This project is a prime example of how data analysis tools can be leveraged to drive breakthroughs in energy research.

The interdisciplinary project focuses on molten salts, whose unique characteristics make it uniquely suited for applications in energy production and storage technologies like nuclear reactors and hybrid energy systems. The goal for the project, titled “Physics-Informed AI Surrogates for Turbulent Forced Convection in Energy Systems,” is to develop a faster, more reliable AI tool that predicts how heat moves through turbulent liquid coolants used in advanced nuclear-fission reactors and future fusion systems.

According to Myoungkyu Lee, a professor at UH’s Cullen College of Engineering, molten salts are attractive for several advanced reactor and fusion blanket concepts but are also among the hardest ones to model. Today’s standard engineering tools were developed for fluids like water and air, and can significantly misjudge heat flow in molten salts. The most accurate simulations are too computationally expensive for everyday design work, so engineers currently bridge the gap with conservative safety margins.

The project aims to cut down this guesswork by developing an AI-assisted model that reproduces results much faster while still following actual laws of fluid flow and heat transfer. Unlike many AI systems, this model would flag areas where its predictions may be unreliable. The goal is to develop a tool that runs much faster than today’s most detailed simulations while keeping errors small.

The award adds to UH’s growing portfolio of energy research, and the broader commitment to bringing together world-class researchers, cutting-edge technologies, and collaborative partnerships with industry, government agencies, and laboratories to advance AI applications across energy, engineering, manufacturing, and other critical industries. This latest award further strengthens UH’s leadership in energy innovation.

The project will receive $180,000 from the DOE, while Lawrence Livermore National Laboratory and University of Pennsylvania will each receive $380,000 and $190,000 respectively. By combining some of its greatest assets – leadership in energy and engineering, world-class research, and national partnerships – UH is solidifying its role as a leader in AI applications for businesses.

According to Ramanan Krishnamoorti, Vice President for Energy and Innovation at UH, this project combines the university’s strengths. ‘Being part of this effort will continue to solidify UH’s role as The Energy University,’ he said.

The development of reliable, carbon-free energy technologies is a pressing challenge that requires innovative solutions. By harnessing AI to accelerate design processes, researchers can make significant strides in advancing next-generation nuclear and fusion energy systems.

Diffusion transformers have become increasingly popular for generating high-quality images, but they often come with a significant memory footprint and latency overhead. To address this issue, researchers have developed various quantization methods that reduce the precision of model weights and activations while maintaining image quality. One such method is SVDQuant, which has been integrated into Nunchaku, an inference engine designed for diffusion transformers. Now, thanks to the recent integration of Nunchaku Lite in Diffusers, users can load pre-quantized checkpoints without requiring a custom pipeline or separate inference engine.

Belgian and Dutch Tesla owners can now enjoy the company’s advanced voice assistant, Grok, after months of waiting. The feature was first introduced to other European markets earlier this year, but it has finally been rolled out to these two countries as part of the latest summer update.

The new voice control system replaces the previous rigid list-based interface with a more conversational approach. Users can now interact with their cars in a more natural way, setting destinations by describing them rather than dictating addresses. This marks a significant improvement over the old system, which required precise phrasing and repetition to work correctly.

Grok is available on eligible versions of the Model S, Model 3, Model X, and Model Y, provided they have an AMD infotainment chip installed since around 2021. Unfortunately, owners with older Intel-based hardware will not be able to access this feature, as no software update can change their car’s underlying architecture.

The Summer Release brings a range of new features beyond just navigation. Grok can now consult the owner’s manual and provide advice on vehicle maintenance and usage, answer general questions, look up music and podcasts, and adjust climate control settings. The manual function is particularly useful, as many owners rarely read their car’s manual but often have unanswered questions.

Grok ships with five different voices and a range of personalities to suit various tastes. While some options are designed for children, others cater specifically to adults, including modes described by Tesla as ‘unhinged’ or ‘sexy.’ However, it remains unclear whether these more provocative settings belong on the dashboard, especially considering European regulators may take an interest in this aspect.

To use Grok effectively, owners need either a Premium Connectivity subscription or a stable Wi-Fi connection. In practice, most users will likely opt for connectivity while driving, as accessing Wi-Fi from a moving vehicle is not very reliable.

It’s worth noting that Grok does not replace any part of Autopilot or Full Self-Driving systems and only provides information through voice commands. The assistant focuses on answering questions, searching for music, and setting destinations without taking control of the car’s driving functions.

The introduction of Grok in Belgian and Dutch Teslas highlights an interesting aspect of AI assistants in cars: their most useful features often lie in their mundane capabilities rather than flashy personalities or voices. By providing straightforward answers to common queries and streamlining navigation tasks, these systems can genuinely improve the driving experience.

A new tool is being rolled out by Substack, a platform that allows writers and creators to share their work with readers. This tool can scan posts, notes, replies, and comments to estimate how much text may have been written using artificial intelligence or AI assistance.

The detection system uses an external company called Pangram, which specializes in identifying AI-generated content. The integration of Pangram’s technology is aimed at increasing transparency on the platform as AI becomes more prevalent online.

According to Chris Best, Substack’s cofounder and CEO, the primary issue with AI-generated content is not its quality but rather the mismatch between a reader’s expectations and reality when they unknowingly invest their attention in something created without human thought. This phenomenon has been referred to as ‘Claudefishing’.

The new tool can analyze content longer than 100 words by allowing readers to choose the option to scan for AI text from the three-dot menu in the top-right corner of a post.

Substack is also introducing an optional feature that allows creators to explain their writing process with a statement called ‘How I make this’. This addition aims to provide more context and transparency about the content being shared on the platform.

Writers can now use Pangram’s technology to scan their drafts, giving them the option to report any inaccurate results. However, it is essential to note that Pangram can only detect whether AI was used in creating the text, not its quality or the level of human care involved.

Chris Best emphasizes that the goal of this integration is to make it easier for readers to choose where they invest their time and attention. He argues that when readers are unsure about the authenticity of content, it undermines trust in authorship and can have negative consequences for writers who use AI tools responsibly.

A new tool has been made available to help users of the AI platform Claude better understand how artificial intelligence is being used in various industries and occupations. The Anthropic Economic Index connector allows anyone with access to Claude to explore data on AI usage directly, providing insights into which jobs are most affected by automation and what tasks are being taken over by machines.

The Anthropic Economic Index itself measures the actual use of AI in the economy, gathering information that has already proven useful for researchers, journalists, and policymakers. However, its creators want this valuable resource to be accessible not just to experts but also to anyone curious about how AI fits into their daily work or personal life. With the new connector, users can ask Claude questions like ‘Which occupations rely most heavily on AI?’ or ‘What are the common ways people in Colorado use Claude for?’

The data behind these answers comes directly from the Anthropic Economic Index, providing a comprehensive picture of how AI is being used across different sectors and regions. Users can start by asking broad questions about their industry or occupation, then drill down into specifics to get more detailed information. They can also ask Claude to show them the underlying data supporting any answer they receive.

It’s worth noting that the Anthropic Economic Index reflects patterns in Claude usage rather than providing a comprehensive view of the entire labor market. As users explore the data, Claude will remind them about these limitations and point them back to the source information for further context. The full datasets remain freely available on the website, alongside the new connector.

The Anthropic Economic Index connector is now live within claude.ai, requiring only a minute or so to set up. Users can find it in the connectors menu by searching for ‘Anthropic Economic Index’ and enabling it – this works with any conversation using any Claude model without needing additional installation. This tool marks an important step towards making complex data more accessible to those who need it most.

Users of claude.ai can now explore how AI is being used across various industries, occupations, and regions. The Anthropic Economic Index connector provides a valuable resource for anyone looking to understand the impact of automation on their field or daily life.