🔥Meta Challenges OpenAI, Again

Today we will discuss:

• Sponsored: Kin AI—Your personal peace companion for emotional growth

• Meta takes aim at OpenAI—Chatbot app and Llama API announced at LlamaCon to rival closed AI models 

• Don’t miss this week’s Workflow Wednesday—AI-powered decision making with Consensus, INK, and OpenAI + Trello + Zapier for smarter choices; now available at only $20/month! 

• Google’s new AI language experiments—Three new experiments use Gemini to teach real phrases, slang, and vocabulary from your surroundings

• In other AI news today—Trump officials consider changes to Biden's AI chip export rule, Freepik launches an AI image generator trained on licensed data, Musk teases Grok 3.5 after Alibaba’s Qwen3 stirs buzz, and Microsoft CEO reveals up to 30% of the company’s code is AI-generated

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

💛Your personal peace companion - meet Kin

Kin is your private AI companion for emotional growth, helping you make sense of your thoughts, so you can own the moments that matter (and your personal data). 

Whether you’re looking to recognize patterns, gain clarity, or simply be heard, Kin is here to grow with you.

And as part of the OpenTools.Ai community, you’re getting an early look at one of the most meaningful AI tools of 2025.

➡️ Discover Kin today

At LlamaCon, Meta debuts chatbot app and developer API to challenge OpenAI’s stronghold

Key Points 

• Meta launched an AI chatbot app with a social feed and a streamlined Llama API to expand the use of its open models.

• The company is openly positioning itself as the alternative to OpenAI, favoring openness over proprietary systems.

🦙News - At its first-ever LlamaCon on Tuesday, Meta rolled out two major AI products.

The company introduced a consumer-facing Meta AI chatbot app and a developer-focused Llama API that lets people access Llama models in the cloud. While both are designed to boost adoption of Meta’s open-source AI models, the bigger picture seems to be about taking on OpenAI directly.

The chatbot app feels like a response to OpenAI’s rumored social network. It includes a social feed where users can share their chats, and it tailors responses based on your activity across Meta apps. The Llama API, on the other hand, is aimed at developers. It lets them build tools and apps using Llama models with just a single line of code, removing the need for third-party cloud services.

👨‍💻Open models, open strategy - Meta’s approach is clear; promote open-source, challenge closed systems. In a conversation with Databricks CEO Ali Ghodsi, Mark Zuckerberg said he sees other open model labs like DeepSeek and Qwen as collaborators, not competitors. He believes mixing the best parts of different models will eventually lead open systems to surpass closed ones in quality.

🤓A bonus regulatory advantage - There could also be a legal angle. The EU AI Act offers benefits for companies that release free and open-source models. While there’s debate over whether Llama fully qualifies, Meta continues to frame it that way.

For now, Meta seems focused on growing the open AI ecosystem, even if that means skipping the race for the most advanced model in favor of long-term influence.

Read the full story here. 

We just dropped the 17th edition of Workflow Wednesday—built for AI-minded professionals who want to do more than just read about AI.

This week’s theme: AI-Powered Decision Making

Make smarter decisions without the guesswork: Use Consensus for research-backed answers, let INK shape your content strategy with AI insights, and automate smarter project choices using OpenAI, Trello, and Zapier.

Join here

Google experiments with AI to personalize how you learn languages

Key Points 

• Tiny Lesson helps you learn phrases tailored to specific real-life situations, like losing your passport or asking for help.

• Slang Hang generates local-style conversations and explains informal phrases so learners sound more natural, not overly formal.

• Word Cam uses your phone’s camera to label real-world objects in your target language and suggest related vocabulary. 

🤖News - Google is rolling out three new AI-powered language learning experiments designed to help people learn languages in a more flexible, personalized way. 

🤹‍♂️Here's the lowdown -

1. Learn what you need, when you need it: The first experiment, called Tiny Lesson, focuses on helping people learn specific phrases based on real-life situations. For example, if you’ve lost your passport, you can describe the scenario and get relevant vocabulary, grammar tips, and example sentences like “I don’t know where I lost it” or “I want to report it to the police.”

2. Speak like a local, not a textbook: The second experiment, Slang Hang, is all about helping users sound more natural when they speak. It creates realistic dialogues, like a street vendor chatting with a customer, and introduces informal expressions as the conversation unfolds. You can hover over tricky terms to understand their meaning and context, though Google warns that the AI may occasionally fumble slang.

3. Point, click, learn: The third experiment, Word Cam, uses your phone’s camera to identify objects around you and label them in your target language. It also suggests related words, which can be helpful for everyday vocabulary that often gets overlooked. For instance, you might know the word for “window,” but not for “blinds.”

These experiments are available through Google Labs and currently support over 20 languages. Google says the goal is to explore how AI can make language learning more dynamic, useful, and tailored to everyday situations.

Read the full story here.