GoKawiil
Tech News
Google is rolling out the ability to upload files directly into the Gemini side panel in Chrome, through a new Add attachments option. Users can attach a document, such as a PDF, and ask Gemini to summarize it, extract details, or clarify sections without leaving the browser. The feature was first spotted by an X user and confirmed working in Chrome by Android Authority.
A group of climbers attempting Mount Shasta in Siskiyou County, California used Google's Gemini chatbot to plan their route, food and water needs, and were told the climb would take eight hours and that they should pack simple carbs instead of fats. The estimate proved drastically wrong: the ascent alone took 16 hours, leaving them descending in darkness, off-route, and with a dead phone after their AllTrails navigation app failed, ultimately requiring a nearly 24-hour rescue operation by search and rescue teams.
An Android Authority contributor describes a personal effort to reduce dependence on Google after growing frustrated with the company's push to inject Gemini AI features across its products. The writer began by testing search alternatives like DuckDuckGo and Startpage before settling on the paid search engine Kagi, while acknowledging some Google products remain hard to replace and older hardware will stay in use.
Google unveiled Gemini 3.8 Flash, its third Flash-tier model release in six weeks, alongside a new cybersecurity model for government and enterprise clients. The launch coincided with a federal judge rejecting the Justice Department's demand that Google divest its ad exchange, and with Berkshire Hathaway's Greg Abel publicly praising Alphabet's AI standing.
Google released two new AI models on Wednesday: a standard Gemini 3.8 Flash built for agentic tasks, coding, and multi-step reasoning, and Flash Cyber, tuned specifically for finding and fixing software vulnerabilities. CEO Sundar Pichai said 3.8 Flash beats many larger frontier models on the DeepSWE coding benchmark at lower cost, while Flash Cyber scored 86.2% on the CyberGym benchmark and found vulnerabilities across 20 programming languages with over 70% success in internal testing.
A Reddit user reported that their Google Home assistant began calling them 'Fish Lemon' after a casual conversation about cooking fish weeks earlier. The nickname has persisted across reminders, smart display labels, and direct queries about their name, resisting attempts to correct it despite their Google Account showing their real name.
Google has rolled out a Google Home app update that lets Gemini for Home handle voice requests that mix media playback and smart device controls in a single command, such as turning off lights while stopping music. The update also fixes previously reported bugs with setting alarms and adding items to Google Keep, and it now alphabetizes automation options while enforcing privacy settings on third-party routines.
Google has shipped Gemini 3.8 Flash, its third Flash-tier model in six weeks, built to handle complex agentic workflows more effectively than earlier versions. Alongside it, the company introduced Gemini 3.8 Flash Cyber, a specialized version aimed at vulnerability detection and automated patching, though that variant is currently restricted to participants in Google's Fairwind Program.
Google has launched Gemini 3.8 Flash, its third Flash-tier model release in just six weeks, alongside a specialized variant called Gemini 3.8 Flash Cyber aimed at vulnerability detection and mitigation. The company touts it as its strongest reasoning and coding model to date, with introductory API pricing of $0.75 per million input tokens and $3.75 per million output tokens through year-end before doubling.
An early build of Google's Device Intelligence app, discovered ahead of the Googlebook laptop launch, reveals an opt-in feature that automatically groups related local files into topic-based collections, such as trip planning documents. These collections can be viewed through desktop widgets, and Google states the sorting happens on-device without files leaving the machine or being used for AI training.
Google has launched Gemini 3.8 Flash, a successor to Gemini 3.7 Flash released just weeks earlier, which the company says performs more reasoning steps and iteratively calls tools on complex tasks. Input and output token pricing remains unchanged from the prior model, but Google cautions that the new model may consume more tokens overall to boost performance, particularly at higher effort settings. Early analysts note it beats rivals like Anthropic's models on coding, finance, and legal benchmarks, though it can end up pricier in practice due to higher token usage.
Google launched Gemini 3.8 Flash, its third Flash-tier model in six weeks, improving on coding, agentic tasks and multi-step reasoning while keeping the same pricing as 3.7 Flash ($0.75/$3.75 per million input/output tokens). Alongside it, Google introduced Gemini 3.8 Flash Cyber, a specialized version tuned for vulnerability detection and automated patching, accessible only to vetted security teams through a new invite-only Fairwind Program.