Google Makes Its Visible AI Watermark Optional - SynthID Stays On Either Way
Google's Gemini watermark toggle went live August 14, letting users hide the visible mark on AI images, videos, and songs without affecting SynthID.
The biggest stories in technology - software, hardware, and the platforms shaping how we live and work.
Google's Gemini watermark toggle went live August 14, letting users hide the visible mark on AI images, videos, and songs without affecting SynthID.
WhatsApp is rolling out pinned message notifications on iPhone, making it easier to notice important messages without opening a conversation first.
ChatGPT Google Drive integration rolled out, letting paid subscribers edit Docs and Sheets inside the chat, with edits saving to the real file.
OpenAI Ultrafast launched August 13 as a new API tier for GPT-5.6 Sol - 14x faster than standard, 750 tokens per second, powered by Cerebras.
Learn how to enable RCS on iPhone to text Android users with high-resolution photos, read receipts, typing indicators, and better group messaging.
Grok 4.6 launched August 12 for long-running agents at $2 per million tokens, with a 1753 ELO on knowledge work and double usage promo in Cursor at launch.
Grok Bot is xAI's early beta - AI agents with their own cloud computer that sign into tools, finish jobs overnight, and coordinate in group chats.
Bluesky has added a new option that lets users hide reposts from specific accounts in their feeds while continuing to follow those users.
YouTube Partner Program changes take effect February 1, 2027 - Shorts revenue needs 10 million views per 90 days, and YPP entry rises to 8,000 watch hours.
Venmo Google Play support launched August 10. Link a Venmo account in Play Store payment settings to use it for apps, games, subscriptions, and tips.
Starting August 14, Claude Code auto mode runs by default for Pro, Max, and Team. Auto mode blocked 89% of dangerous commands vs 13.6% by humans in testing.
OpenAI is pausing some Astra activities while it strengthens security controls after evaluations raised concerns about the model's cybersecurity capabilities.