Trump AI video duping people
In a sophisticated cyber scam, over 200 people across Karnataka, mainly in Bengaluru, Tumakuru, Mangaluru, and Haveri, were duped of more than Rs 2 crore through an AI-generated video of Donald Trump promoting fake investments in Trump Hotel rentals. Victims, lured by promises of high returns and work-from-home tasks, were first asked to pay Rs 1,500 to create accounts. They were shown fake earnings on an app to gain trust, prompting larger investments. One Haveri-based advocate lost nearly Rs 6 lakh before realising the fraud. Police are investigating.
Shopping magic
Google Cloud and Myntra have teamed up to launch ‘Dream Room Inspirations,’ an AI-powered tool that lets users visualise personalised home decor using text prompts. Powered by Google’s Imagen 3 model on Vertex AI, the feature enables furniture shoppers to generate images of rooms that match their desired aesthetic, making interior design more accessible and interactive. The tool is part of Myntra’s push to enhance customer engagement through immersive, generative AI experiences, offering shoppers inspiration and clarity while choosing home decor, all from the convenience of a simple text-based input.
Amazon’s audio feature
Amazon is testing a generative AI-powered audio feature that offers short-form product summaries through the “hear the highlights” button on select product pages in its shopping app. These AI-generated audio clips synthesise product details, reviews, and web data into conversational, easy-to-digest summaries, helping customers save time and shop on the go. Initially rolled out to select US users, the feature is designed for products that require more consideration. It builds on Amazon’s broader AI ecosystem, including tools like Rufus, Shopping Guides and Review Highlights, which the e-comm platform says are all aimed at making shopping experience faster, smarter, and more personalised through conversational technology.
Power hungry
Artificial intelligence could consume nearly half of all data centre electricity by the end of 2025, as per new analysis by tech sustainability expert and founder of the Digiconomist tech sustainability website Alex de Vries-Gao. Published in Joule, the study highlights how energy-hungry AI chips, particularly from Nvidia and AMD, are driving soaring demand. The International Energy Agency also predicts AI systems may soon match Japan’s current energy usage. While advances in chip efficiency and geopolitical constraints could slow growth, rising demand for “sovereign AI” and ventures like OpenAI’s Stargate risk deepening fossil fuel dependence. Experts call for greater transparency in AI’s energy impact, warning of growing environmental concerns.
Google’s ‘AI Mode’ in search
At its 2025 I/O conference, Google unveiled “AI Mode”, marking a leap in reshaping internet search with conversational, expert-like interactions. Powered by its Gemini 2.5 model, AI Mode is now available to US users, alongside previews of tools that book tickets, interpret live video, and integrate with Gmail. While enhancing user satisfaction, AI Overviews have reduced click-throughs to other websites by nearly 30%, raising concerns for web publishers. Google is also re-entering the smart glasses market with AI-powered spectacles. As AI alternatives grow, Google asserts its dominance, receiving 136 billion monthly visits, far outpacing ChatGPT, and betting big on a future led by AI.
Medical error check
Medical errors, particularly medication mistakes, remain a critical threat to patient safety. More than 138 million patients are harmed every year by doctors’ errors, the World Health Organization (WHO) warned in 2022. In the same year, in India, research from the National Library of Medicine, showed an alarming annual incidence rate of up to 5.2 million cases related to medical malpractice across various healthcare settings in India. As per a report, it injures 1.3 million Americans annually. At UW Medicine in Seattle, nurse anesthetist John Wiederspan and anaesthesiologist Dr Kelly Michaelsen are pioneering AI-powered smart eyewear that scans and verifies drug labels in real time to prevent deadly vial swap errors. Their system, boasting 99.6% accuracy in simulated settings, promises to reduce mistakes without adding to clinicians’ cognitive load. While experts urge caution about privacy and over-reliance on AI, the technology’s potential to save lives and ease pressure in high-stress medical environments marks a hopeful advance in healthcare safety.
Blood test sans needles
India’s first AI-powered blood test, Quick Vitals, is transforming diagnostics by offering needle-free results in under 20 seconds via facial scans. Using photoplethysmography, the app measures haemoglobin, blood pressure, heart rate, and more, making it ideal for mass screenings. First launched at Hyderabad’s Niloufer Hospital, it is now expanding to Maharashtra and remote regions. Experts say the app’s speed, affordability, and accuracy within WHO standards could revolutionise community healthcare. A clinical trial on 1,000 children aims to validate its use for early diagnostics nationwide.
Microsoft’s Aurora & weather forecasting
Microsoft’s new AI weather model Aurora is making waves in forecasting by delivering accurate 10-day predictions within seconds. Trained on physics-based models and massive Earth system datasets, Aurora’s flexibility extends beyond weather to areas like air pollution and renewable energy markets. Already in use at a major European weather centre, it showcases the growing shift toward AI-powered forecasting. While faster and more adaptable than traditional models, Aurora still needs human oversight and physical data grounding. Experts remain optimistic but cautious, citing challenges like extreme weather prediction and high energy costs.