🧠Gemini Gets Recall Ability

Today we will discuss: 

• Gemini can recall past conversations—Google’s chatbot now remembers previous chats, offering a more seamless and personalized experience

• Anthropic’s hybrid AI model—A new AI that blends deep reasoning with speed and outperforms OpenAI’s o3-mini in key areas

• In other AI news today—Google will test AI to enforce age restrictions, Arm launches its own chip with Meta as an early customer, and the UK’s AI Safety Institute is now renamed the AI Security Institute to reflect its new focus

• Saved the best for last—9 must-try AI tools

Google’s Gemini can now recall past conversations

Key Points 

• Gemini can now remember details from past conversations, so users don’t have to repeat themselves.

• This feature goes beyond general memory by recalling specific past interactions for more natural conversations.

• Users can review, delete, or disable stored chats, and Google won’t use them to train AI models.

🤖News - Google’s Gemini AI chatbot is getting a significant upgrade—it can now remember details from previous conversations. Announced in a blog post on Thursday, this new capability allows Gemini to summarize past chats or recall information from earlier discussions. Instead of repeating yourself or searching through old threads, you can pick up right where you left off.

🤔How is this different from the memory feature? Google has previously introduced a general memory feature that helps Gemini remember user preferences—like how you like to be addressed or your favorite foods. However, this feature doesn’t automatically recall full chat histories. Gemini’s new ability makes conversations more fluid by referencing specific past interactions.

🧐What about privacy concerns? Google is giving users control over their data, allowing them to review, delete, or set time limits for stored conversations. The recall feature can also be turned off entirely through the “My Activity” page in Gemini. Google has also made it clear that user conversations are not used to train its AI models.

👨🏻‍💻Rollout - The recall feature is rolling out now for English-speaking users subscribed to Google One AI Premium ($20/month). Google plans to expand support to more languages and enterprise users in the coming weeks.

Read the full story here. 

Anthropic prepares to launch hybrid AI model that outperforms OpenAI’s o3-mini in key areas

Key Points 

• Anthropic’s new AI model will switch between deep reasoning and fast responses for greater flexibility.

• The model is said to outperform OpenAI’s o3-mini-high in certain programming tasks and excels at analyzing large codebases.

☕News – Anthropic is preparing to launch its next major AI model, which is described as a “hybrid” capable of switching between deep reasoning and faster responses. 

Expected to arrive in the coming weeks, the model reportedly outperforms OpenAI’s o3-mini-high “reasoning” model on certain programming tasks. It’s also said to be particularly strong in analyzing large codebases and other business-related benchmarks, suggesting a focus on enterprise applications.

Anthropic CEO Dario Amodei hinted at the company’s approach earlier this week. “We’re generally focused on trying to make our own take on reasoning models that are better differentiated,” he said in an interview Monday. He also pushed back against the idea that standard models and reasoning models are entirely separate categories, suggesting a more integrated vision for AI.

🤓What’s more? Alongside the model, Anthropic is reportedly introducing a “sliding scale” pricing system, giving developers control over costs. Since deep reasoning tasks require more computing power, this approach could help balance performance and affordability.

Read the full story here.