Tech • AI • Robotics • Game
AI News: Stay Ahead of 99% With the Latest AI Updates (Grok, Gemini)
8/10TL;DR
OpenAI, xAI, Meta, Crea, and Google have introduced a wave of AI updates spanning voice, automation, image editing, subscriptions, and new Gemini tools, with several features already live and others still rolling out.
KEY POINTS
ChatGPT voice expands
OpenAI has updated ChatGPT voice to sound more natural in longer conversations and added 13 new voices. The changes are rolling out to all users rather than being restricted to paid tiers. The update is likely to matter most for regular voice users, including people practicing spoken languages who need smoother back-and-forth interaction.
Grok gets a separate inbox
Grokbot can now claim and use its own email inbox for signups, newsletters and confirmations instead of relying on a personal address. That could reduce clutter in services such as Gmail when AI agents are used to register for sites or handle routine online tasks. The feature points to a broader push toward more autonomous agent behavior.
xAI bundles products into one plan
xAI has launched an X subscription that combines Grok, Grokbot, Cursor, and X Premium under a single package starting at $8 a month. The offer is aimed at users juggling multiple AI subscriptions, especially those who want one plan for general AI use, automation, coding and platform perks. Final value will depend on tier limits and included features.
Facebook adds direct feed controls
Facebook is introducing recommendation controls that let users explicitly request more or less of certain content in the main feed and Reels. Instead of depending only on likes, watch history and skipped posts, users will be able to type preferences such as more travel clips or fewer celebrity updates. The feature has been announced but is not yet visible on every account.
Crea 3 focuses on precise edits
Crea 3 is being positioned as an image editor for targeted changes rather than full regeneration. Users can create an image, describe one specific edit, and have only that part altered while the rest stays intact. Typical examples include changing the color of a liquid in a bottle or replacing a single object, with edits said to take under 8 seconds.
Gemini 4 Argon enters the app
Google has added Gemini 4 Argon directly to the Gemini app for Google AI Pro and Ultra subscribers. AI Pro starts at $19.99 per month, and the model is intended for heavier work such as summarizing long reports, extracting key findings and processing large amounts of information. Its inclusion in the app removes the need to switch to a separate tool.
Gemini introduces task mode
A new switch between chat and task modes is designed to let Gemini do more than answer prompts. Task mode can carry out multi-step requests across Google apps and continue running even after a user closes a laptop, while chat mode remains the standard conversational format. The feature is coming to Android first for AI Pro and Ultra subscribers.
Google reorganizes Gemini tools
Gemini now includes a customized menu that brings together its skills and more than 40 connected apps in one place. Listed integrations include services such as Instacart and OpenTable, making it easier to find available tools without digging through settings. Most connected apps are currently limited to the United States.
Free Gemini users may lose manual model choice
Google is also changing model selection for free accounts by moving them to an auto setting, where Gemini chooses the model automatically. The rollout appears incomplete, as manual selectors are still visible on some free and unsigned sessions. If fully implemented, the change would simplify the interface but reduce user control.
Three more US-only Google tests
Google also announced Ask YouTube, which answers questions using information drawn from videos, support for building personal Android apps in Google AI Studio, and a universal cart for shopping across retailers in one place. All three are currently US-only. Together they show Google extending AI beyond chat into search, app creation and commerce.
CONCLUSION
This week’s updates show AI platforms competing on usability as much as raw model performance. The immediate trend is clear: more natural interfaces, more automation, and tighter integration into everyday apps and services.
Ask a question