News
Google DeepMind proposed a new approach to AI training
Google DeepMind researchers have introduced the concept of continuous AI learning, which will allow assistants to better understand users and…
Sora Trends (April 2025)
What users were generating in April: nostalgia, irony, dolls in boxes, odd angles, and memes with meta-awareness. A selection of…
ChatGPT receives a memory feature for internet search
OpenAI has added the ability for personalized search in ChatGPT, which takes into account the user’s conversation history.
ChatGPT has started addressing users by name
ChatGPT has unexpectedly started using users’ names in conversations, sparking discussions about privacy and personalization.
Google unveils Gemini 2.5 Flash model for developers
Google has launched testing of Gemini 2.5 Flash, which delivers fast processing of text, images, video, and audio with resource…
ChatGPT’s Accuracy in Guessing Photo Locations Amazes Users
OpenAI’s AI can now recognize cities and venues from photos, but this raises questions about personal data security.
LG TVs will track viewers’ emotions
LG uses AI to personalize ads on its TVs, taking into account viewers’ emotions and individual traits.
Copilot Vision is now available for free to all Edge users
Microsoft has made it possible to use Copilot Vision for free to analyze web pages in the Edge browser with…
Microsoft opens access to the compact AI model BitNet b1.58
BitNet b1.58 2B4T from Microsoft runs fast on simple devices and is already available for free use.
Grok gets a memory feature for better answers
Grok by xAI now remembers conversation details to provide personalized advice and lets users manage saved data.
Microsoft adds a new automation feature to Copilot Studio
Copilot Studio makes it easy to automate routine tasks on websites and in applications, even if they don’t have an…
OpenAI unveils two new AI models
OpenAI has released the o3 and o4-mini models, which combine text and visual capabilities to solve complex problems.
