GoKawiil
Tech News
Google published documentation outlining the intended uses and known limitations of Gemini 3.8 Flash, a model aimed at cost-effective, production-scale agent deployment for developers and enterprises. The company flagged persistent issues including hallucinations, occasional slowness or timeouts, and inconsistent knowledge freshness across domains despite a stated March 2026 cutoff.
Adobe has acquired Rilo, a Bengaluru-based marketing intelligence startup, through a licensing and team-acquisition deal, folding in its six-member team and technology. Founded in 2025 by IIT graduates Georgi Boby and Dhruv Jaglan, Rilo had raised $1 million at a $10 million valuation from backers including PeakXV and Day Zero Ventures before this exit. The startup's product will be shut down and its custom workflow-building tools absorbed into Adobe's offerings.
MrBeast is releasing a new video on September 5 showing his team racing across three extreme environments—jungle, desert, and Arctic—while relying on Google's Gemini AI to navigate dangers and survive harsh conditions. The project marks a branded collaboration between Google and one of YouTube's biggest creators to showcase Gemini's practical capabilities.
Google is reportedly preparing to launch Gemini 3.8 Flash imminently, a model DeepMind's internal testing found outperforms Anthropic's Opus on coding tasks. The release would follow Gemini 3.7 Flash, launched just weeks earlier, and comes as Google shelves plans for a separate Gemini 3.5 Pro model, pushing the next Pro-tier release to Gemini 4.0.
Google has upgraded Gemini's video analysis with a new 'agentic' approach that lets the AI selectively choose which frames, audio, or transcript segments to review instead of processing every frame. The company says this cuts token usage by up to 88% while improving accuracy by up to 7%, and enables tracking of object counts, movement, and sub-second events across multi-hour videos. The feature currently works through the Gemini API in AI Studio and the Enterprise Agent Platform, with plans to reach the Gemini app and YouTube's 'Ask YouTube' tool soon.
Meta unveiled Muse Voice Transcribe, its first real-time audio perception model, capable of transcribing over 20 speakers simultaneously while switching between more than 20 validated languages, including mid-sentence code-switching. CEO Mark Zuckerberg demoed the tool on X, noting it uses adaptive delay to balance speed and accuracy. The model is now available through Meta's Mac AI app, Muse Code, and its Model API at $3 per 1,000 audio minutes.
Internal testing of Google's upcoming Gemini 3.8 Flash model reportedly shows meaningful improvement in coding performance, an area where the company has trailed rivals Anthropic and OpenAI. The model is expected to be released publicly this week.
Google unveiled five new Android updates in a Tuesday blog post, including 'Motion Assist,' which displays a moving screen bubble to ease motion sickness in vehicles, and 'Guided Vision,' which uses the camera and Gemini AI to help blind and low-vision users read labels, menus, or identify objects. Additional updates let Gemini help users track misplaced items and add new personalization options to Google Messages.
A new study from Google Research and Technion shows that when models like GPT-5 and Gemini-3 hallucinate, the underlying fact is usually already stored in their parameters rather than missing entirely. The researchers found these models encode 95-98% of tested facts, but often fail to surface them during generation due to retrieval, not storage, limits. They propose 'fact-level profiling' to separate whether a fact is encoded versus reliably retrievable across different phrasings.
Google's September Android Drop introduces a feature letting users verbally tell Gemini where they placed untracked items, such as passports or keys, with the option to attach a photo. These notes are stored in a new 'Remembered' tab within the Find Hub app, and can also be added manually without using voice commands. The feature will roll out soon to devices running Android 16 or later in regions where both Gemini and Find Hub operate.
Google released its monthly Android feature drop this week, adding tools like a motion sickness reducer for screen use in moving vehicles, shared notes within chat threads, camera-based object identification, and customizable Google Messages chat themes. The update also lets users ask Gemini to remember where they placed everyday items, such as passports or keys, storing that information in Google's Find Hub tracking app.
Google has rolled out its September 2026 Android Drop, a bundle of feature updates spanning several Android apps. Highlights include a new Gemini capability that lets users verbally ask the assistant to remember where they placed objects, saving that location in Find Hub, plus a Gemini Live Guided Vision mode that describes surroundings aloud and suggests reframing for dim or small text. Google Messages also gains built-in Google Keep list sharing and officially launches customizable Chat Themes with matching text bubble colors.