Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Organizations interested in deploying AI agents must first fine-tune them, especially in workflows that often feel rote. While some organizations want agents that only perform one kind of task in one workflow, sometimes agents need to be brought into new environments with the hope that they adapt.  Researchers from the Beijing University of Posts and Telecommunications have unveiled a new method, AgentRefine. It teaches agents to self-correct, leading to more generalized and adaptive AI agents.  The researchers said that current tuning methods limit agents to the same tasks as their training dataset, or “held-in” tasks, and do not perform as well for “held-out,” or new environments. By following only the rules laid out through the training data, agents trained with these frameworks would have trouble “learning” from their mistakes and cannot be made into general agents and brought into to new workflows.  To combat that limitation, AgentRefine aims to create more generalized agent training datasets that enable the model to learn from mistakes and fit into new workflows. In a new paper, the researchers said that AgentRefine’s goal is “to develop generalized agent-tuning data and establish the correlation between agent generalization and self-refinement.” If agents self-correct, they will not perpetuate any errors they learned and bring these same mistakes to other environments they’re deployed in.  “We find that agent-tuning on the self-refinement data enhances the agent to explore more viable actions while meeting bad situations, thereby resulting in better generalization to new agent environments,” the researchers write.  AI agent training inspired by D&D Taking their cue from the tabletop roleplaying game Dungeons & Dragons, the researchers created personas, scripts for the agent to follow and challenges. And yes, there is a Dungeon Master (DM).  They divided data construction for AgentRefine into three areas: script generation, trajectory generation and verification.  In script generation, the model creates a script, or guide, with information on the environment, tasks and actions personas can take. (The researchers tested AgentRefine using Llama-3-8B-Instruct, Llama-3-70B-Instruct, Mistral-7B-Instruct-v0.3, GPT-4o-mini and GPT-4o) The model then generates agent data that has errors and acts both as a DM and a player during the trajectory stage. It asses the actions it can take and then see if these contain errors. The last stage, verification, checks the script and trajectory, allowing for the potential of agents it trains to do self-correction. Better and more diverse task abilities The researchers found that agents trained using the AgentRefine method and dataset performed better on diverse tasks and adapted to new scenarios. These agents self-correct more to redirect their actions and decision-making to avoid errors, and become more robust in the process.  In particular, AgentRefine improved the performance of all the models to work on held-out tasks.  Enterprises must make agents more task-adaptable so that they don’t repeat only what they’ve learned so they can become better decision-makers. Orchestrating agents not only “direct traffic” for multiple agents but also determine whether agents have completed tasks based on user requests.  OpenAI’s o3 offers “program synthesis” which could improve task adaptability. Other orchestration and training frameworks, like Magentic-One from Microsoft, sets actions for supervisor agents to learn when to move tasks to different agents.  source

Molti sostengono che l’influenza politica dei ricchi sia un fattore determinante Questo comunicato stampa è stato tradotto in italiano dall’inglese, lingua originale di redazione. Un nuovo sondaggio del Pew Research Center condotto in 36 Paesi, pubblicato oggi, rileva una diffusa preoccupazione dell’opinione pubblica in relazione alle disuguaglianze economiche. Alla domanda relativa a cosa conduca a queste disuguaglianze, la maggior parte delle persone nei Paesi oggetto del sondaggio indica l’intersezione tra ricchezza e politica. Una mediana del 54% degli adulti dei Paesi esaminati afferma che il divario tra ricchi e poveri è un problema molto grande nel proprio Paese; una mediana del 30% afferma invece che è un problema moderatamente grande. Una mediana del 60% ritiene che l’eccessiva influenza politica dei ricchi contribuisca in larga misura a generare diseguaglianze economiche. Tra i sei fattori presi in esame, questo è quello che trova maggiormente il riscontro degli intervistati, posizionandosi in cima alla lista in 31 dei 36 Paesi in cui è stato svolto sondaggio. Il nostro sondaggio ha anche riscontrato un’ansia profonda a livello mondiale in merito al futuro dell’economia e un forte desiderio per una riforma economica. Una mediana del 57% degli adulti intervistati prevede che i bambini del proprio Paese, una volta cresciuti, staranno peggio dei loro genitori dal punto di vista economico, mentre una mediana del 34% sostiene che staranno meglio. In 15 dei 31 Paesi per cui sono disponibili le tendenze, la percentuale di pubblico che pensa che i figli staranno peggio dei loro genitori dal punto di vista economico è più alta oggi rispetto ai sondaggi precedenti alla pandemia. Nei Paesi in cui è stato condotto il sondaggio si registra un ampio sostegno alla modifica del sistema economico. In tutti i Paesi tranne tre (Singapore, Paesi Bassi e Svezia), la maggioranza delle persone afferma che il loro sistema economico necessita di cambiamenti profondi (mediana del 52%) o di una riforma completa (mediana del 20%). Dei Paesi europei oggetto del sondaggio, gli adulti italiani sono i più propensi ad affermare che la presenza di alcune persone che lavorano più duramente di altre, unita alla presenza di alcune persone che nascono con maggiori opportunità, contribuisca in larga misura alle disuguaglianze economiche nel Paese. La maggioranza afferma che questo è il caso in ogni misura. Circa otto italiani su dieci (79%) ritengono inoltre che i bambini di oggi vivranno una situazione economicamente peggiore di quella dei loro genitori in futuro. Risultati chiave aggiuntivi del report: Fattori che si ritiene contribuiscano alle diseguaglianze economiche La maggioranza delle persone in quasi tutti i Paesi oggetto del sondaggio ritiene che tutti i sei fattori esaminati siano causa di diseguaglianze economiche almeno in parte. Tuttavia, esistono delle differenze in merito al fatto che ciascuno di essi vi contribuisca in misura notevole. Una mediana del 48% degli adulti sostiene che i problemi nel sistema educativo del proprio Paese contribuiscono in larga misura alle diseguaglianze economiche. Di tutti i Paesi presi in esame, lo Sri Lanka e la Turchia sono gli unici due Paesi in cui ciò viene considerato il maggiore fattore determinante. Circa quattro intervistati su dieci hanno affermato che la presenza di alcune persone nate con maggiori opportunità di altri (40%) e la presenza di alcune persone che lavorano più duramente di altre (39%) sono fattori che contribuiscono alle diseguaglianze economiche in larga misura. Un numero minore indica l’impatto dei robot e dei computer che svolgono il lavoro precedentemente svolto dagli esseri umani (31%) o le discriminazioni delle minoranze razziali o etniche (29%). I Brasiliani sono maggiormente propensi a ritenere che le discriminazioni razziali o etniche siano uno dei principali fattori che determinano le diseguaglianze economiche: lo afferma il 64%, la percentuale più alta in tutti i Paesi oggetto del sondaggio. Percezioni delle diseguaglianze e delle discriminazioni a livello globale Nel nostro sondaggio abbiamo chiesto agli intervistati la portata dei problemi di vari tipi di disuguaglianze nel loro Paese, tra cui il divario tra ricchi e poveri (una mediana del 54% lo considera un problema molto grave), le discriminazioni basate sulla razza o l’etnia di una persona (34%), la disparità di diritti tra uomini e donne (31%) e le discriminazioni basate sulla religione di una persona (29%). Molti degli intervistati nei Paesi oggetto del sondaggio li considerano problemi moderatamente gravi. Molte persone li considerano dei problemi molto gravi nel proprio Paese. In 35 Paesi su 36, la maggior parte degli intervistati afferma ciò in merito al divario tra ricchi e poveri rispetto a qualsiasi altro problema. Nel complesso, vi è meno preoccupazione circa le discriminazioni religiose rispetto agli altri problemi esaminati. Tuttavia, in cinque Paesi (Bangladesh, Francia, India, Nigeria e Sri Lanka), la metà o più delle persone intervistate ritiene che le discriminazioni religiose siano un problema molto grave. Le persone provenienti da Paesia reddito medio sono più propense a considerare tutte le forme di diseguaglianza come un problema grave rispetto alle persone dei Paesi ad alto reddito. (Consulti l’Appendice A per una classificazione dei Paesi a medio reddito e ad alto reddito.) Di seguito sono riportati i risultati chiave di un nuovo sondaggio del Pew Research Center che si è svolto tra il 5 di gennaio e il 22 maggio 2024 e ha coinvolto 45.103 adulti. Leggere il report completo (in inglese): https://www.pewresearch.org/global/2025/01/09/economic-inequality-seen-as-major-challenge-around-the-world/ LEGGERE ANCHE: Le valutazioni economiche in 34 Paesi sono più negative che positive Metodologia: https://www.pewresearch.org/2025/01/09/methodology-inequality/ Risultati principali del sondaggio: https://www.pewresearch.org/wp-content/uploads/sites/20/2025/01/pg_2025.01.09_inequality_topline.pdf source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Global venture capital investments rose to $368.5 billion in 2024, up 5.4% from $349.4 billion a year earlier, according to the first look at the Q4 2024 Pitchbook-NVCA Venture Monitor report. But the number of global deals in 2024 fell 17% to 35,686 from 43,320 a year earlier in 2023. AI deals as a percentage of all deals rose for the year, as you can see in the chart below. The 2024 global deals are down 50.9% from $751.5 billion in the peak year of 2021 and down 37% from 57,068 in deal count in 2021. AI deals are big part of the picture now. There were 8,343 global AI deals in 2024, down 3.6% from 8,661 in 2023 and down 16.6% from 10,007 in 2021. AI’s share of all global VC deals is at a new high. The value of those global AI deals in 2024 was $131.5 billion, up from 52% from $86.3 billion in 2023 and down 6% from $140.2 billion in 2021. AI and machine learning were 35.7% of global deal value in 2024, up from 24.7% in 2023. And AI and machine learning were 23.4% of the global deal count in 2024, up from 20% in 2023. In 2021, AI was 18.7% of global deal value and 17.5% of global deal count. Q4 global numbers On the global level in Q4, Asia Pacific’s venture market has struggled through the last few years, something that didn’t change in 2024, Pitchbook lead VC analyst Kyle Stanford said. Compared with Europe and the U.S., the amount of dry powder built up within the various markets across APAC was much smaller, further pressuring dealmaking over the past year. China, which has driven around half of the annual deal activity for APAC, has seen a material decline in activity, due to both economic challenges within the country, as well as the tensions with the U.S. government, which has curtailed activity by U.S.-headquartered firms. Just 20.4% of deal count occurred in Asia, the lowest proportion in the past decade. Globally, AI has continued to dominate the headlines and investment focus of investors despite some noting that the investment activity is not sustainable long-term. Whether or not that true is trivial in the current moment. Just over half of all VC invested globally during Q4 went to an AI-focused company. Its true that amount was heavily influenced by the likes of OpenAI, Databricks, xAI, and other well-known companies raising for share buybacks and investment into chips and computing energy needs, but the most important factors is the level of capital availability for AI compared with other sectors, Stanford said. The proportion of total deals going to AI companies has consistently increased over the past couple years as large corporates and investors alike move to harness the expected efficiencies of the next tech wave, he said. Global VC investments and deal counts by year. “VC-backed exits have not been strong historically for APAC, though many markets are still too young to develop a healthy exit environment,” he said. “The lack of exits across many of the regions has kept many foreign investors weary of increased activity during the market slowdown. Japan has been an outlier in terms of count, as many IPOs within the country have helped drive returns to investors. In 2024, 19% of the global VC-backed exits originated in Asia-based companies.” Fundraising has been slow globally, as new commitments dropped just over 20% YoY. The lack of exits has had a large impact on fundraising for Asia as LPs have been less inclined to reup commitments at this time. 2024 marked the lowest year for new commitments since 2018, and was the lowest year for closed funds in the market in the past decade. North America and Europe similarly struggled to secure new commitments to venture funds. Q4 U.S. deals U.S. Dealmaking remained relatively robust in the fourth quarter of 2024 from a count perspective, and increased slightly by 3.7% compared to a year earlier, Pitchbook and the NVCA said. In the quarter, AI deals accounted for nearly half (46.4%) of total US deal value. Stanford said it seems counterintuitive to the narrative in the market over the past few years, but is indicative of holdover of certain mechanics of venture from a few years ago. “What has happened is that the excess of dry powder from the high fundraising years of 2021 and 2022 have kept many investors active in the market despite the lack of returns,” Stanford said. “With the slow fundraising years of 2023 and 2024, we should likely see this relative robustness start to deteriorate as fund run through their available capital and aren’t able to raise a subsequent fund.” AI deals by year has been rising sharply. Artificial intelligence continues to be the story of the market, and drove a near majority of dollars for VC in 2024, he said. OpenAI, xAI, Anthropic, and others have become synonymous with outsized deals in venture, and seemingly operate in a different funding environment than most VC-backed companies who continue to struggle with lower capital availability, Stanford said. But the lack of exits remains the story of the venture market, even as the outlook is more hopeful, he said. Just $149.2 billion in exit value was created during 2024, largely coming from a handful of IPOs. Unicorns, which hold around two-thirds of the U.S. VC market value, have held tight as private companies, creating pressure on investors and limited partners with the lack of distributions. Merges and acquisitions were was also “silent in 2024,” with few large deals to note, Stanford said. A more acquisition-friendly environment in 2025 could set the stage for a renewed M&A market, especially if a soft-landing for the economy can be fully engineered, he said. In the U.S., fundraising was dominated by large, established firms. Thirty firms accounted for more than 68% of total

A new survey by payroll tech firm Yellow Canary found that just 22% of Australian businesses have adopted proactive payroll compliance technology. Still, more may follow as they seek to reduce the legal and business risk of underpaying employees. Intentional employee underpayments were made a criminal offense on Jan. 5 following amendments to Australia’s Fair Work legislation, with individuals and businesses now potentially liable. While unintentional mistakes will not attract criminal penalties, Yellow Canary estimates underpayments represent between 1% and 3% of total headcount costs across the market. The Yellow Canary survey of 533 compliance leaders in Australia found the rising risk around underpayments is driving more tech buyers toward proactive payroll compliance tools: 23% plan to adopt technologies in the next one to two years. 21% of businesses plan to implement these tools in the next 12 months. 17% said they were satisfied with manual compliance processes. 15% were curious about more proactive payroll technologies but had no plans to implement them. “The introduction of the Closing Loopholes Acts, including the criminalisation of wage theft, marks a pivotal moment for Australian businesses,” Yellow Canary Managing Director Marcus Zeltzer said in the report. “Our research reveals while many businesses are making payroll compliance a top priority, a significant number are still relying on flawed manual processes or have not conducted thorough reviews.” SEE: Best practices for maintaining payroll compliance Payroll teams are concerned they are not paying staff correctly Almost half (48%) of those surveyed by research house Lonegran Research on behalf of Yellow Canary said they had been making payroll compliance a top priority ahead of the Closing Loopholes law. However, 93% of local businesses with at least 50 employees still said they had at least one area of concern regarding potential employee underpayments in their organisation as the law came into force. Additionally, 17% expressed uncertainty about paying their staff correctly, while 19% suspect an underpayment issue may exist but have not confirmed it. Several key drivers of payroll underpayment concerns were identified in the research report: 39% of respondents had concerns with staying current with legislation and obligations, demonstrating the complexity of remaining compliant in an evolving regulatory environment. 37% cited concerns around a lack of internal communication, noting that collaboration and information flow across departments reduce errors and inconsistencies in payroll processes. A further 32% had concerns with time and resource constraints for payroll audits and historical reviews. Meanwhile, the reliability of payroll software in ensuring compliance was a concern for 31%, as was aligning rostering or time and attendance processes, which are often managed through system integrations. SEE: 8 best payroll software for Australian Businesses Only 7% of respondents said there are no areas of concern regarding potential underpayments. However, Yellow Canary said it was unclear if this reflected genuine assurance or lack of awareness, given it had found some non-compliance in 100% of clients in its work reviewing $70 billion in wages. More Australia coverage Proactive compliance and AI could improve payroll scorecard Australia has experienced widespread problems with underpayments — affecting large private and public sector organisations — in many cases due to Australia’s complex system of payment awards. The Yellow Canary report found many employers still rely on “less reliable” methods: 31% still conduct manual audits with spreadsheets. 32% review pay code configurations. 37% use sampling for payroll checks. SEE: A step-by-step guide to doing payroll (the right way) “While businesses may feel confident in their manual methods, these processes are flawed, prone to error, limited in scalability, and unable to keep up with the increasing complexity of compliance,” the report said. Adopting proactive payroll compliance technologies is expected to help reduce the problem by replacing more manual review processes with regular tech-supported audits of workforce payroll data. The incorporation of AI could support these efforts — but some businesses remain skeptical More than half (59%) of Australian businesses with 50 or more employees are optimistic about the potential of introducing artificial intelligence into their payroll compliance frameworks in the future. AI is not yet commonly used in payroll compliance in Australia, but the report said that the evolution of technology shows “great potential for being integrated into existing processes.” For instance, AI can be used to analyse payroll data patterns, and identify anomalies — such as incorrect pay codes, underpaid employees, or misclassifications — to provide payroll teams with real-time insights. However, 27% of respondents remain either skeptical of AI’s ability to improve payroll compliance or believe AI will introduce more challenges and complicate payroll processes in the future. “Businesses must navigate challenges such as integration issues, data privacy concerns, and resistance to change before widespread adoption [of AI],” the report said. source

By Adam Lidgett ( January 10, 2025, 5:34 PM EST) — A California federal judge has held that German microscope company Leica Microsystems Inc. didn’t infringe a patent issued to the University of Michigan that covers a new way of measuring fluorescence…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Nvidia today announced foundation models running locally on Nvidia RTX AI PCs that supercharge digital humans, content creation, productivity and development. GeForce has long been a vital platform for AI developers. The first GPU-accelerated deep learning network, AlexNet, was trained on the GeForce GTXTM 580 in 2012 — and last year, over 30% of published AI research papers cited the use of GeForce RTX. Jensen Huang, CEO of Nvidia, made the announcement during his CES 2025 opening keynote. Now, with generative AI and RTX AI PCs, anyone can be a developer. A new wave of low-code and no-code tools, such as AnythingLLM, ComfyUI, Langflow and LM Studio enable enthusiasts to use AI models in complex workflows via simple graphical user interfaces. NIM microservices connected to these GUIs will make it effortless to access and deploy the latest generative AI models. Nvidia AI Blueprints, built on NIM microservices, provide easy-to-use, preconfigured reference workflows for digital humans, content creation and more. To meet the growing demand from AI developers and enthusiasts, every top PC manufacturer and system builder is launching NIM-ready RTX AI PCs. “AI is advancing at light speed, from perception AI to generative AI and now agentic AI,” said Huang. “NIM microservices and AI Blueprints give PC developers and enthusiasts the building blocks to explore the magic of AI.” The NIM microservices will also be available with Nvidia Digits, a personal AI supercomputer that provides AI researchers, data scientists and students worldwide with access to the power of Nvidia Grace Blackwell. Project Digits features the new Nvidia GB10 Grace Blackwell Superchip, offering a petaflop of AI computing performance for prototyping, fine-tuning and running large AI models. Making AI NIMble How AI gets smarter Foundation models — neural networks trained on immense amounts of raw data — are the building blocks for generative AI. Nvidia will release a pipeline of NIM microservices for RTX AI PCs from top model developers such as Black Forest Labs, Meta, Mistral and Stability AI. Use cases span large language models (LLMs), vision language models, image generation, speech, embedding models for retrieval-augmented generation (RAG), PDF extraction and computer vision. “Making FLUX an Nvidia NIM microservice increases the rate at which AI can be deployed and experienced by more users, while delivering incredible performance,” said Robin Rombach, CEO of Black Forest Labs, oin a statement. Nvidia today also announced the Llama Nemotron family of open models that provide high accuracy on a wide range of agentic tasks. The Llama Nemotron Nano model will be offered as a NIM microservice for RTX AI PCs and workstations, and excels at agentic AI tasks like instruction following, function calling, chat, coding and math. NIM microservices include the key components for running AI on PCs and are optimized for deployment across NVIDIA GPUs — whether in RTX PCs and workstations or in thecloud. Developers and enthusiasts will be able to quickly download, set up and run these NIM microservices on Windows 11 PCs with Windows Subsystem for Linux (WSL). “AI is driving Windows 11 PC innovation at a rapid rate, and Windows Subsystem for Linux (WSL) offers a great cross-platform environment for AI development on Windows 11 alongside Windows Copilot Runtime,” said Pavan Davuluri, corporate vice president of Windows at Microsoft, in a statement. “Nvidia NIM microservices, optimized for Windows PCs, give developers and enthusiasts ready-to-integrate AI models for their Windows apps, further accelerating deployment of AI capabilities to Windows users.” The NIM microservices, running on RTX AI PCs, will be compatible with top AI development and agent frameworks, including AI Toolkit for VSCode, AnythingLLM, ComfyUI, CrewAI, Flowise AI, LangChain, Langflow and LM Studio. Developers can connect applications and workflows built on these frameworks to AI models running NIM microservices through industry-standard endpoints, enabling them to use the latest technology with a unified interface across the cloud, data centers, workstations and PCs. Enthusiasts will also be able to experience a range of NIM microservices using an upcoming release of the Nvidia ChatRTX tech demo. Putting a Face on Agentic AI Nvidia AI Blueprints To demonstrate how enthusiasts and developers can use NIM to build AI agents and assistants, Nvidia today previewed Project R2X, a vision-enabled PC avatar that can put information at a user’s fingertips, assist with desktop apps and video conference calls, read and summarize documents, and more. The avatar is rendered using Nvidia RTX Neural Faces, a new generative AI algorithm that augments traditional rasterization with entirely generated pixels. The face is then animated by a new diffusion-based NVIDIA Audio2FaceTM-3D model that improves lip and tongue movement. R2X can be connected to cloud AI services such as OpenAI’s GPT4o and xAI’s Grok, and NIM microservices and AI Blueprints, such as PDF retrievers or alternative LLMs, via developer frameworks such as CrewAI, Flowise AI and Langflow. AI Blueprints Coming to PC A wafer full of Nvidia Blackwell chips. NIM microservices are also available to PC users through AI Blueprints — reference AI workflows that can run locally on RTX PCs. With these blueprints, developers can create podcasts from PDF documents, generate stunning images guided by 3D scenes and more. The blueprint for PDF to podcast extracts text, images and tables from a PDF to create a podcast script that can be edited by users. It can also generate a full audio recording from the script using voices available in the blueprint or based on a user’s voice sample. In addition, users can have a real-time conversation with the AI podcast host to learn more. The blueprint uses NIM microservices like Mistral-Nemo-12B-Instruct for language, Nvidia Riva for text-to-speech and automatic speech recognition, and the NeMo Retriever collection of microservices for PDF extraction. The AI Blueprint for 3D-guided generative AI gives artists finer control over image generation. While AI can generate amazing images from simple text prompts, controlling image composition using only words can be challenging. With this blueprint, creators can use

You’re running into errors or glitches in Microsoft Outlook and suspect the issue may be due to corruption in your personal folder file, which houses all your email and other content. To help track down the cause of the problem, Microsoft offers the built-in Inbox Repair tool, also known as ScanPST. SEE: Windows, Linux, and Mac Commands Everyone Needs to Know (free PDF) (TechRepublic) The ScanPST tool can scan a Personal Storage Table, PST, or Offline Storage Table, OST, file to diagnose and repair errors in the file. These file formats are used in Outlook to store data locally for personal backups and offline access, respectively. If the tool finds any corruption, it offers to repair the errors. Here’s how it works. Before it attempts to repair a file, the tool automatically creates a backup. However, you may want to have your own backup as an additional safety precaution. To find the location of the PST or OST file, open Outlook and click the File menu. Click the button for Account Settings and then select the command for Account Settings. Image: TechRepublic At the Account Settings window, click the tab for Data Files. Examine the path for the file you wish to scan and open it in File Explorer. Close Outlook. Then, simply create a backup copy of the file. If you bump into an error about the file being locked when trying to back it up, ensure Outlook and any applications that use or integrate with Outlook are closed. If necessary, open Task Manager to check for any programs that need to be shut down. Image: TechRepublic PINBOX: 96019 In File Explorer, browse to the folder that contains scanpst.exe to launch the tool. The location varies slightly based on your flavor of Outlook and whether it’s the 32-bit or 64-bit version. For the 64-bit version, start by browsing to C:Program FilesMicrosoft Office. For the 32-bit version, browse to C:Program Files (x86)Microsoft Office. From there, drill down to the following locations: Microsoft 365, Outlook 2021, Outlook 2019, and 2016 – ..rootOffice16. Outlook 2013 – ..Office15 Outlook 2010 – ..Office14 Outlook 2007 – ..Office12 If you cannot locate the file through a specific path in File Explorer, simply search for scanpst.exe. Double-click the file. The window lists the path for the PST or OST file. If it’s pointing to the wrong file, click the Browse button and select the correct file. Image: TechRepublic With the correct PST or OST file listed, click the Start button. The tool goes through eight phases. Assuming the file is corrupt, the tool will stop at some point and tell you that it found errors in the file. Clicking the Details button may or may not provide more information. Either way, click the Repair button. Image: TechRepublic The tool will then display a notice telling you when the repair is complete. Click OK. Run the tool again to see if your PST or OST file now passes the test. If additional errors are found, click the Repair button again. Image: TechRepublic At some point, the tool may indicate that it has found only minor inconsistencies in the file and that repairing it is optional. Instead of repairing it again, you may want to check the app’s log file to view the scan results. To do this, go to the folder containing your PST or OST file. Double-click the log file that starts with the same name as your mailbox. The log file contains plain text, so you can read it in Notepad or a similar text editor. Then, open Outlook and try to replicate the behavior that caused the problems in the first place. If Outlook is working properly, then you’re set. If not, you may want to try another repair or consider other reasons for the glitches plaguing Outlook. How can I tell if my PST file is corrupted? Common indicators of corrupted PST files include: Error messages, such as “the file [filename].pst cannot be opened.” Outlook fails to open or crashes when accessing the PST file. Emails, contacts, or calendar entries are missing or cannot be accessed. Delays when opening or navigating within folders stored in the PST file. Emails show strange characters or incomplete data. SEE: How to Scan and Repair Corrupted System Files in Windows 11 How can I prevent future PST file corruption? PST file corruption can be prevented with the following best practices: Keep PST files as small as possible: Archive old emails, delete unnecessary items, and split large PST files into smaller ones. Close Outlook correctly: Click File > Exit, and don’t shut down your computer abruptly when it’s running. Install the latest Outlook updates and antivirus software: Keeping software up-to-date ensures bug fixes, while antivirus software prevents malware that can corrupt PST files. Upgrade faulty or unreliable hardware: These include hard drives or USB drives for storing PST files. Enable Auto-Archive: Auto-Archive reduces the size of your active PST file by moving older items to an archive. Don’t use PST files for processing large amounts of data: For example, frequent import/exports and large attachments. How can I repair PST files without ScanPST? You don’t necessarily need to use ScanPST to repair corrupted files. One alternative way is to create a new PST file and then import the data. Go to File > Account Settings > Data Files > Add, then choose “Outlook Data File (.pst)” and save the new file. Go to File > Open & Export > Import/Export and select “Import from another program or file”. Then, choose ‘Outlook Data File (.pst)’. Browse your corrupted PST and select “Do not import duplicates,” then complete the wizard to move any recoverable data to the new PST file. You could also restore from a recent backup of the corrupted PST file if you have one. Sometimes, corruption is limited to certain add-ins or settings, rather than the PST file itself. In this case, you can access your non-corrupted file in Outlook’s Safe Mode, where such settings are disabled.

Scientists in the UK have forged 5.5 tonnes of a new kind of steel capable of withstanding the searing heat and intense neutron radiation of nuclear fusion, the same reaction that powers the Sun and stars. The breakthrough is another boost to Europe’s growing flock of fusion energy startups. A UK Atomic Energy Authority (UKAEA) working group called NEURONE produced the reduced-activation ferritic-martensitic steel, or “RAFM” for short. It marks the first time that RAFM has been produced on an industrial scale in Britain. “This is really positive and potentially has relevance for all fusionenergy projects,” Ryan Ramsey, COO at British startup First Light Fusion, told TNW.    Fusion reactors superheat hydrogen atoms to extremely high temperatures, forming a charged gas called plasma. Using magnetic fields or lasers to compress the plasma, they force the atoms to fuse, releasing huge amounts of energy that can be used to generate electricity.  When running, the plasma inside a fusion energy machine reaches temperatures of 150 million°C — temporarily making them the hottest points in our solar system. Giant magnets suspend this plasma in mid-air — keeping it away from direct contact with the metal walls. The walls are also cooled to stop them from overheating. Nevertheless, no ordinary steel is up to the task.  The 💜 of EU tech The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now! “The biggest problem isn’t the heat, it’s neutron damage,” said Ramsey. Neutron radiation can quickly degrade the inner walls of a nuclear reactor. “If you don’t manage that, then you’ll be shutting down the fusion reactor regularly to replace the walls, which means you’re not producing power during that time frame,” he explained.    The inner walls of fusion reactors, like the retired JET tokamak machine pictured here, must withstand searing heat and intense radiation. Credit: EUROfusion NEURONE’s new steel can withstand high neutron loads and temperatures of up to 650°C, potentially improving the operational efficiency of future fusion powerplants.  For startups like Oxford University spinout First Light, the development marks another step towards the moonshot goal of building a commercially viable fusion reactor.   NEURONE forged the steel using an electric arc furnace, which runs on electricity instead of coal, housed at the Materials Processing Institute (MPI) in Middlesbrough. The UKAEA said that its new forging method could make producing RAFM up to 10 times cheaper than was previously possible.  “The production of 5.5 tonnes of fusion-grade RAFM steel lays the foundation for cost-effective manufacturing of these types of fusion steel for future commercial fusion programmes,” said David Bowden, who heads up the NEURONE programme.   Despite huge progress, fusion energy has always seemed to be that “20-years-away” technology. But the tides might be changing. According to a poll at the International Atomic Energy Agency’s (IAEA) forum in London last year, 65% of industry insiders think fusion will generate electricity for the grid at a viable cost by 2035, and 90% by 2040. source

Budget planning tools help IT leaders build an accurate estimate of future income and expenses in a detailed enough way to make sound operational decisions. That sounds simple enough, yet in actual practice creating a realistic budget is a time-consuming task that many IT leaders dread.  AI has the ability to analyze historical finance data, usage patterns, project expenditures, and related inputs to better forecast the future, says Tyler Higgins, managing director of management and technology consulting firm AArete, via email.  When teamed with automated data collection, AI has the potential to enhance many budget modeling processes, says Anurag Sahay, managing director and global lead of AI and data sciences at digital engineering firm Nagarro. In an online interview, he notes that AI can also improve extrapolation and forecasting to assess resource needs, extract key insights from unstructured feedback, and optimize decision-making models for the best planning outcome and “what-if” scenarios.  Multiple Benefits  AI-supported budget planning offers both direct and indirect benefits. “The direct benefits are streamlining and shortening the budgeting process,” Higgins says. “The ideal outcome is a predictive budgeting process that contains powerful scenario planning tools and improved accuracy.”  Related:What Could Less Regulation Mean for AI? The most exciting part about using AI in IT budget planning is how it can shift the entire mindset from cost-cutting to value-building, says Jeff Mains, founder of Champion Leadership Group, a business training and coaching provider. Traditionally, budgets were seen as ways to manage resources and avoid overspending, but with AI we’re talking about a tool that identifies opportunities for innovation, he explains via email. “It doesn’t just keep you within budget — it shows you where strategic investments in IT can drive growth.” Mains says he uses AI to not only forecast expenses, but to create dynamic budget models that adjust in real-time based on shifting business needs and external factors. “It’s about creating a budget that grows with you, rather than just containing costs.”  AI-driven predictive analytics and benchmarking tools are already available for parts of the overall IT budget process, says Steven Hall, chief AI officer at technology research and advisory firm ISG. In an email interview, he notes that several technology business management tools, such as Apptio, provide deep insights and scenario planning to analyze current spending patterns and run savings and growth scenarios. “These platforms are integrating GenAI capabilities to provide even deeper insights and look for savings by integrating usage, external benchmark, and demand data to plan better IT spending.”  Related:AI-Driven Quality Assurance: Why Everyone Gets It Wrong First Steps  Higgins says the best way to begin using AI budget planning is to pick a specific use case and explore its potential. “We’re still in the infancy of AI, yet use cases keep growing,” he notes. “Instead of biting off everything at once, pick a few use cases and ensure that your baseline operational, financial, and usage data is sufficient, clean, and well structured.” Higgins suggests establishing an objective for each use case, then deploying a pilot AI project to determine if it’s delivering the anticipated output.  When embedded into IT financial platforms, AI budgeting will provide deeper insight into opportunities as well as create the ability to model various scenarios for growth, Hall says. “These evolving capabilities will also provide leaders with actionable insights and identify specific actions to address budget challenges.”  The best approach is to take the long view, Mains says. “AI can deliver immediate insights, but its real power comes when it’s integrated into long-term strategic planning.” He suggests selecting a single area of volatile IT spending, such as cloud services or software licenses, and allowing AI to analyze usage patterns in order to offer smarter budget recommendations. “From there, you can gradually scale AI’s role, aligning its outputs with broader business goals.”  Related:6 AI-Related Security Trends to Watch in 2025 Risks and Benefits  AI’s biggest benefit is predictive accuracy. It’s not just about saving time — it’s about knowing where your IT investments will have the highest impact six months from now, or even a year down the road, Mains says. The biggest risk is treating AI as a silver bullet. “The human element is still critical,” he warns. “Without context and strategic insight, even the most advanced AI models can miss the mark.”  Hall notes that AI models are only as good as the data they’re fed, and poor-quality or incomplete data can easily result in inaccurate budget forecasts. “Implementing AI tools also requires an upfront investment in technology and talent, which can be a barrier for smaller organizations.”  Looking Forward  The hardest part of most AI-driven projects, including budgeting, is getting started, Higgins observes. “These tools are never going to be perfect at first, but they will get better, and the results will be tangible for every organization.”  source

It’s a good idea to establish a governance policy supporting the framework. This includes the creation of landing zones, defining the VPN, gateway connections, network policies, storage policies, hosting key services within a private subnet and setting up the right IAM policies (resource policies, setting up the organization, deletion policies). Creating awareness of the policy of least privilege and addressing frustrations when cloud users ask for more to play with, and as a cloud CoE team, you are rightfully holding your ground that comes with it. The cloud CoE team should collect feedback from the users and tweak policies along the way as they deem fit. As the enterprise user community matures through this learning curve, the CoE team has a pivotal role to play in engaging them proactively, supporting them, meeting their needs and helping them address their pain points to lay a strong foundation for building a robust cloud infrastructure that is scalable to deliver business value.  Partnering with the enterprise architecture team in this stage of the cloud journey can speed up cloud maturity and buttress the foundation further on solid bearings. The cloud CoE team of architects should work with the EA to align with the reference architecture patterns that the CoE team would like the application teams/product teams to follow in their solution design. This strategic and collaborative work serves as a blueprint for the architects to ‘show and tell’ the concept of designing cloud-effective solutions and shape mindsets towards the ‘making money with the cloud’ vision. Every company that wants to succeed in scaling AI solutions and capabilities today needs to get this design thinking working for them. Otherwise, it’s like the story of ‘sour grapes’ with high cloud cost bills coming their way with little value added in the AI domain as it has been for a lot of companies in 2024.  3. WALK FASTER: Develop cloud adoption planning and migration roadmap   With the CoE and EA teams working in tandem, it’s time to engage the business stakeholders/product owners and develop the plan and roadmap. Understanding business constraints and priorities, and evaluating domains that have opportunities to expand their revenue and grow exponentially upon moving to the cloud are the key parameters to keep in mind. I recommend a structured approach for this phase of the journey.  source