Image: seventyfourimages/Envato Elements AlexNet, which was released in 2012, is widely credited with sparking the modern AI revolution, particularly in the field of computer vision. Last week, the Computer History Museum in collaboration with Google made the source code for AlexNet publicly available on GitHub; this move gives researchers, developers, and AI enthusiasts a chance to dive into the foundational code that helped shape today’s AI landscape. What is AlexNet, and why does it matter? AlexNet was the deep-learning model that proved neural networks could significantly outperform traditional image recognition methods. Developed by Alex Krizhevsky, Ilya Sutskever, and their advisor Geoffrey Hinton at the University of Toronto, the model leveraged deep convolutional neural networks (CNNs) to classify images with unprecedented accuracy. The secret to AlexNet’s success wasn’t just its architecture — it was also the massive dataset (ImageNet) it was trained on and the use of GPUs for acceleration. At the time, neural networks were considered impractical due to high computational demands, but by harnessing NVIDIA’s CUDA-enabled GPUs, AlexNet changed that perception. When it entered the 2012 ImageNet competition, it dominated, achieving a top-5 error rate of 15.3% — nearly half of the second-place finisher’s score. The legacy of AlexNet in AI evolution Before AlexNet, machine learning models struggled to accurately recognize images, requiring manually crafted features and extensive rule-based programming. AlexNet took a different approach, using deep layers of artificial neurons to automatically learn patterns. This success was a turning point. Soon after, companies like Google, Facebook, and Microsoft ramped up investments in deep learning, leading to modern AI applications, from facial recognition to natural language processing. AlexNet’s influence extended beyond image recognition. Its core principles laid the groundwork for today’s AI models, including large language models (LLMs) like GPT and transformer-based architectures that power tools like ChatGPT. More must-read AI coverage Why open-sourcing AlexNet matters By making AlexNet’s original code publicly available, the Computer History Museum and Google are providing a rare window into one of AI’s defining breakthroughs. While modern AI models have evolved significantly, AlexNet remains a cornerstone of deep learning research. Having access to its source code allows: Students and researchers to analyze the model’s original implementation and learn how early deep learning frameworks were structured. Developers and AI engineers to experiment with the architecture and understand the principles that sparked AI’s rapid advancement. Historians and technology enthusiasts trace the evolution of machine learning from its roots to today’s sophisticated models. How to access the code The original 2012 version of AlexNet is now available on CHM’s GitHub page, preserving the exact implementation that transformed AI. While numerous versions of AlexNet have been recreated over the years, this release represents the authentic model that shifted the industry’s trajectory. source

Jai Prakash Sharma, Executive VP – Technology at Info Edge states that technology is a key business differentiator of digital-first businesses like online job portals, matrimony platforms, ed-tech solutions, and real estate marketplaces. He says, “My role evolved beyond IT when leadership recognized that platform scalability, AI-driven matchmaking, personalized recommendations, and data-driven insights were crucial for business success. For example, in the online job market, optimizing search algorithms and AI-driven candidate-job matching directly impacts user engagement and revenue.” For Shajy Thomas, Regional Head of Tech – APAC at Technicolor, technology actively contributes to shaping long-term business outcomes. “I am a key member of the council responsible for formulating the company’s business strategy and setting goals, followed by developing 1-year, 3-year, and 5-year plans. This ensures that our technology roadmap is fully aligned with our overarching business objectives and fosters a continuous cycle of innovation and efficiency.” source

By Adam Lidgett ( March 24, 2025, 6:03 PM EDT) — The U.S. International Trade Commission has said it is launching a series of investigations into whether imports of products such as video game consoles, nose cleaning devices and semiconductors have infringed various U.S. patents…. 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 Hakimo announced today it has secured $10.5 million in Series A funding to expand its autonomous security monitoring platform. The Menlo Park-based AI security startup’s round was led by Vertex Ventures and Zigg Capital, with participation from RXR Arden Digital Ventures and existing investors Defy.vc and Gokul Rajaram, bringing the company’s total funding to $20.5 million. The company’s flagship product, AI Operator — an autonomous security agent that monitors existing security systems, detects threats in real-time, and executes response protocols with minimal human intervention — arrives as the physical security industry struggles with staffing shortages, rising costs, and false alarm fatigue. “Rising costs and the severe shortage of quality security guards are driving this shift,” explained Sam Joseph, Hakimo’s co-founder and CEO, in an exclusive interview with VentureBeat. “Companies save approximately $125,000 per year relative to a guard, so switching to Hakimo is a no-brainer to customers.” AI security agents that detect any threat describable in words Unlike conventional security monitoring services, Hakimo combines computer vision with generative AI to create a system that can detect any anomaly or threat that can be described in words. “Hakimo’s AI Operator continuously ingests live video feeds from existing security cameras and other sensors such as alarm systems and door sensors,” Joseph said. “Our AI Operator is trained to identify anomalies and threats—such as unauthorized personnel, tailgating, loitering, intrusion into restricted zones, or unusual behaviors.” When the system detects a potential threat, it reasons through pre-programmed response protocols, issues real-time audio warnings through on-site speakers, and escalates to human operators only when necessary. This multi-layered approach allows for comprehensive security coverage at a fraction of the cost of traditional guard services. The system’s ability to work with existing hardware serves as another differentiator. “Companies can deploy Hakimo in just minutes, directly integrating with their existing cameras and hardware—no new infrastructure required,” Joseph explained. How 24/7 AI monitoring outperforms traditional security guards Hakimo’s growth addresses significant challenges facing traditional security approaches, including police departments declining to respond to unverified alarms and a labor crisis in the manned guarding industry. “A guard physically cannot be at all places at all times, especially at places like campuses and large warehouses or even car dealerships,” Joseph said. “Hakimo, on the other hand, monitors cameras giving complete 100% coverage as long as there are enough cameras present at the site.” The CEO highlighted additional advantages: “Guards have to take mandatory breaks during their shifts whereas Hakimo’s AI Operator never takes breaks. Liability risk: Guards have a high risk of getting injured on the job, which causes a huge liability risk. Hakimo, with no on-site human presence, has no such risk whatsoever.” The company has tripled its customer base over the past year, now serving more than 100 clients across industries including multifamily apartments, car dealerships, construction sites, and Fortune 500 enterprises. In 2024 alone, Hakimo’s technology prevented thousands of security incidents, assisted law enforcement in multiple arrests, and even saved a life. “Our AI Operator detected a person collapsing on-site in the middle of the night with no one around. It was then escalated to our human operators who immediately dispatched emergency services (911), ultimately saving the individual’s life,” Joseph explained. Why investors are betting millions on AI-powered physical security The investment in Hakimo reflects growing recognition that physical security requires technological reinvention as traditional approaches like guard services and alarm monitoring increasingly appear antiquated and cost-prohibitive. With the new funding, Joseph plans to “expand and enhance our AI platform capabilities and scale operations to new markets and industries.” The company targets businesses across multiple sectors that need efficient physical security solutions. “Any business needing protection for physical assets or people benefits significantly, especially property management, car dealerships, construction sites, and campuses,” Joseph said, citing examples including a car dealership that eliminated monthly break-ins, a high-rise apartment building that saved nearly a million dollars on security guard spending, and a student housing property that saw incidents drop by 90% while reducing security costs to a third of traditional expenses. Creating the universal platform for intelligent video monitoring Hakimo was founded in early 2020 by Sam Joseph and Sagar Honnungar, both Stanford-trained AI experts with backgrounds in enterprise SaaS. They identified converging trends: exploding camera deployments, falling hardware costs, and rapid advances in computer vision. Their vision extends beyond current applications: “In five years, Hakimo aims to become the universal platform for intelligent video monitoring and understanding, effectively creating and leading a new category within the physical security industry,” Joseph said. As AI continues to advance, Hakimo’s approach may offer a preview of how intelligent systems will increasingly augment — and in some cases replace — human workers in security and beyond. “Our vision is that Hakimo will be the platform for all kinds of video monitoring,” Joseph said. “Anyone should be able to connect any camera to Hakimo and say ‘look out for X and do Y when you see X happening.’” source

Why you can trust TechRepublic At TechRepublic, we publish high-quality, independently researched reviews created by writers who are experts in the fields they cover. Our contributors include seasoned IT professionals, certified accountants, software developers, and industry consultants—people who have worked directly with the tools they evaluate. Every article is built on firsthand experience, in-depth testing, and a deep understanding of what businesses and tech teams actually need to make confident, informed decisions. QuickBooks Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Micro (0-49 Employees), Small (50-249 Employees), Medium (250-999 Employees), Large (1,000-4,999 Employees) Micro, Small, Medium, Large Features API, General Ledger, Inventory Management Quicken Business & Personal Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Micro (0-49 Employees) Micro Features Accounts Receivable/Payable, Invoicing / Billing, Mobile Capabilities, and more Top accounting software comparison QuickBooks Online: Best overall accounting software QuickBooks logo. Source: Intuit QuickBooks QuickBooks Online is my overall best accounting software because it’s one of the most versatile options for small and midsize businesses. Whether you’re managing a retail shop, nonprofit, or service-based company, its customizable features make it a great fit across industries. Another key reason is its widespread use, especially in the US, where it’s easy to find bookkeepers, accountants, or CPAs experienced with the platform. This makes hiring financial help simple, whether you need full-time support or occasional assistance. Additionally, QuickBooks Online has a moderate learning curve, meaning small business owners can handle bookkeeping without extensive training. For those who want more automation, it integrates with over 800 third-party apps, allowing seamless connections to payroll, e-commerce, and other essential tools. Overall, it’s one of the best business accounting software options for businesses that need flexibility, expert support, and powerful integrations. Pricing Simple Start: $35 for one userEssentials: $65 for up to three usersPlus: $99 for up to five usersAdvanced: $235 for up to 25 users Features Fully automated billing, invoicing, expense tracking, and income tracking Class and location tracking Useful expense categorization for easier business tax filing User-friendly mobile app with mileage tracking, receipt scanning, and on-the-go invoicing Sales and sales tax tracking Up to 25 users included with the highest-tier plan Contractor management Advanced features like automated workflows, Excel data syncing, and employee expense management Pros and cons Pros Cons Wide network of CPAs and accountants who are ProAdvisors or QuickBooks Certified. Has a moderate learning curve. Can track locations and classes in the Plus plan. TVersatile in different industries. Limits one organization per account. Can’t add additional users on top of the plan limit. Not as robust as its desktop counterpart in terms of industry-specific features Read our QuickBooks Online review FreshBooks: Best for project accounting FreshBooks logo. Source: FreshBooks FreshBooks is the best accounting software for managing projects efficiently. Its project management and time-tracking features make it easy to log billable hours, track expenses, and generate invoices—all within a streamlined platform. This is especially useful for businesses that bill clients based on time worked or project milestones. Another standout feature is its project-specific estimating tool, which lets you compare projected costs with actual expenses. This helps keep budgets on track and ensures financial transparency. Unlike QuickBooks Online or Wave, FreshBooks is built with project-based businesses in mind, making it a strong choice for those who need to track work progress alongside financials. However, if you require inventory or COGS tracking, I recommend QuickBooks Online, one of the best business accounting software options for small teams, or Xero for larger operations. Pricing Lite: $21; up to five billable clients Plus: $38; up to 50 billable clients Premium: $65; unlimited billable clients Select: Custom; unlimited billable clients and specialized features Features Unlimited invoices with a rich library of fully customizable invoice templates Basic income and expense tracking Integration with Gusto payroll and over 100 other third-party business apps Unlimited users (additional monthly fee) Client collaboration portal Built-in time tracking and project tracking Pros and cons Pros Cons Tracks project costs and compares actual with estimated costs Easy to use even for users without experience in accounting software Has built-in time tracking that works with its project accounting features for accurate billing User-friendly and intuitive Good for only one user for all plans NCharges per additional user, which can be expensive for larger teams Requires bank feed connection to use bank reconciliation Read our FreshBooks review Xero: Best for large teams Xero logo. Source: Xero Xero is QuickBooks Online’s toughest competitor. Regardless of that fact, I picked Xero because it offers a powerful alternative for businesses that need a robust, cloud-based solution. One of its biggest advantages is unlimited users, making it a great fit for large teams that require shared access to financial data without extra fees. Another standout feature is fixed asset tracking, something no other cloud-based software on this list offers. This makes Xero an excellent choice for businesses that need to monitor and manage long-term assets while maintaining accurate financial records. With its scalability and advanced automation, Xero is one of the best business accounting software options for growing companies. Pricing Early: $20; comes with 20 invoices and five bills Growing: $47; offers unlimited bills and invoices Established: $80; includes project accounting Features Automatic bank reconciliation Invoices and quotes (limited to 20 per month with the cheapest plan) Fixed asset tracking Cash flow tracking Unlimited users for easier collaboration Built-in inventory management with all three plans Pros and cons Pros Cons Is the only cloud-based software in this list that has a fixed asset manager Has strong inventory management features that can track COGS Integrates with more than 1,000 apps Offers unlimited users for all plans Is not as popular in the US compared to QuickBooks Has limited Xero Advisors across the US Has limited invoices and bills in the lowest plan Read our Xero review Wave Accounting: Best free plan Wave logo. Source: Wave I chose Wave because it’s one of the best accounting software for businesses, offering

  Image: StackCommerce TL;DR: This grade-A refurbished MacBook Pro (2020) with 16GB RAM, 512GB SSD, and a 10th Gen Intel Core i5 is just $449.99 (reg. $1,799). For professionals who demand reliability, speed, and sleek performance from their daily devices — but would rather not shell out nearly two grand — a refurbished 2020 MacBook Pro at $449.99 is a smart deal. It can also turn out to be a digital life upgrade on a budget. This model isn’t lightweight when it comes to capability. With a 10th Gen Intel Core i5 quad-core processor, 16GB of RAM, and a 512GB SSD, it’s built to handle demanding workflows like 4K video editing, software development, design, and audio production. Whether you’re running more resource-intensive software or juggling spreadsheets, code, and Zoom calls simultaneously, this MacBook Pro can deliver. The 13.3-inch Retina display brings sharp detail and accurate colors. This is ideal for photo and video work. Plus, True Tone technology adjusts the screen’s color temperature to match ambient lighting, reducing eye strain. The Magic Keyboard, a favorite among longtime Mac users, features the upgraded scissor mechanism, while the Touch Bar and Touch ID add convenience and efficiency to your workflow. Four Thunderbolt 3 ports make it simple to connect external displays, docked devices, or fast storage. And with up to 10 hours of battery life, it’s an excellent choice for hybrid professionals who split time between the office, coffee shops, or even airplanes. Refurbished to grade A condition, this MacBook Pro looks and performs like new, making it a solid pick for business pros, consultants, freelancers, and creatives who want Apple quality without the price tag that comes with a new machine. Don’t miss grabbing this grade-A refurbished MacBook Pro while it’s on sale for just $449.99 (reg. $1,799). StackSocial prices subject to change.   source

Artificial intelligence (AI) has rapidly shifted from buzz to business necessity over the past year—something Zscaler has seen firsthand while pioneering AI-powered solutions and tracking enterprise AI/ML activity in the world’s largest security cloud. Enterprises are increasingly adopting AI tools to enhance productivity, automate workflows, and accelerate decision-making. However, cybercriminals are leveraging the same technology to scale sophisticated attacks, from hyper-realistic deepfakes to advanced phishing schemes. With AI fundamentally changing both how businesses operate and how cybercriminals attack, organizations must maintain a current and comprehensive understanding of the enterprise AI landscape. The just-released ThreatLabz 2025 AI Security Report examines the intersection of enterprise AI usage and security, drawing insights from 536.5 billion AI/ML transactions in the Zscaler Zero Trust Exchange. The report reveals how enterprises worldwide and across industries are using and managing AI/ML tools, highlighting both their benefits and security concerns. It examines rising risks associated with AI, from cybercriminals weaponizing AI to the security implications of recent AI advancements like DeepSeek, while providing best practices for mitigating these risks. 5 key findings: AI usage and threat trends The ThreatLabz research team analyzed activity from over 800 known AI/ML applications between February and December 2024. Here are the notable findings: 1. AI/ML usage surged exponentially: AI/ML transactions in the Zscaler cloud increased 36x (+3,464.6%) year-over-year, highlighting the explosive growth of enterprise AI adoption. The surge was fueled by ChatGPT, Microsoft Copilot, Grammarly, and other generative AI tools, which accounted for the majority of AI-related traffic from known applications. Zscaler Figure 1: Top AI applications by transaction volume 2. Enterprises blocked a large proportion of AI transactions: 59.9% of AI/ML transactions were blocked, signaling concerns over data security and the uncontrolled use of AI applications. As organizations work to establish AI governance frameworks, many are taking a cautious approach, restricting access to certain AI applications as they refine policies around data protection. 3. U.S. and India drive the most AI/ML traffic: The United States and India recorded the highest volume of AI/ML transactions in the Zscaler cloud, reflecting strong enterprise adoption and a growing focus on AI-driven innovation. Other top contributors include the United Kingdom, Germany, and Japan, each exhibiting different levels of AI/ML activity. 4. Finance & Insurance and Manufacturing dominate AI adoption: The Finance & Insurance (28.4%) and Manufacturing (21.6%) sectors generated the most AI/ML traffic. Following them, Technology, Healthcare, and Government are integrating AI at varying rates as they navigate the fine line between adoption and apprehension. Zscaler Figure 2: Industries driving the largest proportions of AI transactions 5. AI-driven cyber risks are escalating: Threat actors are leveraging AI to enhance phishing campaigns, automated attacks, and create realistic deepfake content. ThreatLabz researchers demonstrated how DeepSeek can be manipulated to quickly generate phishing pages that mimic trusted brands. Additionally, ThreatLabz uncovered a malware campaign in which attackers created a fake AI platform to exploit interest in AI and trick victims into downloading malicious software. Securing AI and staving off AI threats with Zscaler The ThreatLabz 2025 AI Security Report provides detailed guidance for enterprises looking to securely adopt AI while minimizing risks and stopping AI-powered cyberthreats. Enterprises must rethink security strategies to account for new vulnerabilities, expanded attack surfaces, and AI-fueled cyberattacks. Traditional security approaches reliant on firewalls and VPNs are woefully insufficient against the speed and sophistication of AI-powered threats. Enterprises must adopt a zero trust approach, eliminating implicit trust, enforcing least-privilege access, and continuously verifying all AI interactions. Zscaler’s zero trust architecture delivers Zero Trust Everywhere—securing user, workload, and IoT/OT communications—infused with comprehensive AI capabilities. Its AI models detect and disrupt advanced threats, blocking millions of attacks daily to enhance enterprise security outcomes and mitigate emerging risks. The report details how to stop AI-powered threats with Zscaler, including: Zero trust architecture: Reduce the attack surface by hiding applications and IP addresses from attackers and enforcing least-privilege access. AI-powered cyberthreat protection: Detect and block AI-generated phishing campaigns, adversarial exploits, and AI-driven malware in real time. AI-powered data classification and DLP: Use AI-driven classification to detect and protect sensitive data across Zscaler’s Data Fabric. AI-powered app segmentation: Stop lateral movement within networks, ensuring attackers cannot easily escalate privileges or access critical systems. AI-powered breach prediction: Preempt potential breach scenarios using generative AI and multi-dimensional predictive models. Real-time AI insights: Employ predictive and generative AI for actionable insights that enhance security operations and digital performance. AI visibility: Get in-depth visibility into AI application trends and interactions through interactive dashboards. Get the report Download the ThreatLabz 2025 AI Security Report for additional data-driven insights and analysis of AI’s impact on cybersecurity, with expert guidance to help enterprises securely embrace AI and mitigate its risks. source

By Tom Zanki ( March 24, 2025, 7:55 PM EDT) — Crypto-friendly trading platform eToro Group Ltd. on Monday publicly filed its long-awaited plans for an initial public offering, represented by Skadden Arps Slate Meagher & Flom LLP and underwriters’ counsel Latham & Watkins LLP…. 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

In the summer of 2018, a devastating fire tore through Greece’s Attica region. In its aftermath, four scientists from the University of Cambridge and Imperial College London — including two Greeks — set out to answer a simple yet crucial question: why did some houses burn while others were left untouched? That question set off years of research into computer simulations for predicting wildfire risk and behaviour. In December 2023, the scientists founded Pinepeak to bring their idea out of the lab and into the market. Pinepeak’s wildfire prediction technology combines machine learning algorithms trained on satellite data with cutting-edge physics-based modelling originally developed for jet engine research at the University of Cambridge.  Just as turbulence and air movement dictate how fuel ignites in an engine, they also play a crucial role in the spread of wildfires. By modelling the chaotic nature of wildfire spread and how it interacts with other factors like terrain and weather, Pinepeak can arrive at more accurate forecasts.  “Catastrophic wildfires are escalating due to climate change, but current risk assessment methods fall short of predicting how fires interact with their environment and spread from an ignition point towards communities,” Dr Savvas Gkantonas, CEO and co-founder of Pinepeak, told TNW.  3 free tickets to TNW Conference? Get them now! For a limited time, groups can get up to three extra free tickets! Book now and increase your visibility and connections at TNW Conference Pinepeak COO and co-founder Dr Daniel Fredrich (left) and CEO and co-founder Dr Savvas Gkantonas. Credit: Pinepeak Pinepeak’s algorithms forecast the probability that a wildfire will occur with up to 90% accuracy. They also predict how fires will spread — where they will go, what they will impact, and when all that might occur. The tool can do all this at an extremely high granularity, delivering wildfire risk forecasts down to the level of an individual home located anywhere in the world. Gkantonas said that this kind of intelligence is “highly valuable” to everyone from property insurers to national governments and firefighters.  Now, Pinepeak is looking to scale up. The startup has just been admitted to an accelerator that might enable it to do just that. Turning wildfire predictions into a successful startup Pinepeak is one of eight startups to have been picked for the University of Cambridge’s second START Accelerator programme. Alongside a £40,000 upfront investment, the accelerator gives startups like Pinepeak access to mentorship from proven investors and founders.  Gkantonas said the programme  is “helping accelerate our growth by connecting us with early customers, domain experts, and seed investors, enabling us to grow faster and smarter.” Last month, the company also nabbed $322,000 (£250,000) in pre-seed funding.  Other startups in the cohort include Podromic, which uses AI to predict dementia, and Trismik, which provides testing solutions for large language models (LLMs).  The startups are well located for success. Cambridge is the world’s third most valuable deep tech hub, only below Boston (Harvard, MIT, and Boston University) and the Bay Area, according to Dealroom data. Last year’s START Accelerator saw 11 companies participate, including lithium-sodium battery startup Molyon, which recently raised $4.6mn (£3.5mn), as TNW previously reported.  The search for Europe’s next big startups is the guiding mission of TNW Conference, which takes place on June 19-20 in Amsterdam. Check out out our initial list of speakers and our early agenda for a taste of what’s to come. Tickets for the event are now on sale. Use the code TNWXMEDIA2025 at the check-out to get 30% off the price tag. source