Meta Invests $14.3 Billion in Scale and Hires Its CEO

Mark Zuckerberg's Meta Platforms Inc. has finalized a $14.3 billion investment in Scale AI — taking a 49% stake — and recruited the company's CEO, Alexandr Wang, 28, to join its AI team for producing data models. Scale offers data services to help companies, including Meta and OpenAI, train and improve their AI systems. It also builds custom AI applications for businesses and governments. The startup generated approximately $870 million in revenue in 2024 and is expected to reach $2 billion in revenue by the end of this year. Its new interim CEO will be Jason Droege. Zuckerberg was frustrated with Meta’s progress following the rollout in April of the company’s latest large language model, Llama 4. Not wanting to be left behind, the CEO of Meta took a more hands-on approach, making the recruitment of AI experts and scientists a top priority. He has been offering lucrative pay packages to attract top researchers from Alphabet Inc.’s Google and startup Sesame AI Inc. He has been following a playbook similar to Amazon.com Inc., Microsoft Corp., and Google, with arrangements designed to avoid the regulatory scrutiny that comes with significant acquisitions. Alphabet's Google, the largest customer of Scale AI, announced it would cut ties with Scale after rival Meta bought a 49% stake in the AI data-labeling startup. Google was paying Scale approximately $200 million this year for the human-labeled training data, which is crucial for developing technology, including the sophisticated AI models that power Gemini, its ChatGPT competitor, one of the sources said. Other major customers of Scale, including Microsoft, are also backing away. Elon Musk's xAI plans are exiting, too, arguing their concern to expose their research priorities and road map to a rival.

Hugging Face Unveiled Two Open-Source Humanoid Robots

AI Enterprise Search Glean Hits a Valuation of $7.2 Billion

Figma Announces the Beta Release of the Dev Mode MCP Server

Web Search, Built on Links, Starts to Shift Away Toward LLM Platforms

Web search, built on links, started to shift away from traditional browsers toward LLM platforms in 2025, according to a report by Andreessen Horowitz. The foundation of the $80 billion+ SEO market just cracked with Apple’s announcement that AI-native search engines like Perplexity and Claude will be built into Safari, said the VC firm. This put Google’s distribution chokehold in question. “A new paradigm is emerging, one driven not by page rank, but by language models. We’re entering Act II of search: Generative Engine Optimization (GEO),” stated the report. Page ranks are determined by indexing sites based on keyword matching, content depth and breadth, backlinks, and user experience engagement. However, today, it’s not about ranking high on the results page. LLMs are the new interface for how people find information. Visibility is obtained by showing up directly in the answers of LLMs like GPT-4, Gemini, and Claude. Users’ queries are longer (averaging 23 words vs. 4), sessions are deeper (averaging 6 minutes), and responses provide personalized, multi-source synthesis, remembering and showing reasoning, rather than just relying on keywords. Additionally, the business model and incentives have changed. Google monetizes user traffic through ads; users are paid with their data and attention. In contrast, most LLMs are paywalled, subscription-driven services. However, an ad market may eventually emerge on top of LLM interfaces, but the rules, incentives, and participants would likely look very different than traditional search. New monitoring platforms, such as Profound, Goodie, and Daydream, enable brands to analyze how they appear in AI-generated responses. Tools like Ahrefs’ Brand Radar track brand mentions in AI Overviews, enabling companies to understand how they’re framed and remembered by generative engines. Semrush has a dedicated AI toolkit designed to help brands track perception across generative platforms, optimize content for AI visibility, and respond quickly to emerging mentions in LLM outputs. referral traffic from the LLMs is still low overall <5%, but growing and perhaps better targeted? marketing and ad tech is and will evolve to fit this with the large LLM platforms building their own products + the AI software companies to cover the long tail (as we're seeing) https://t.co/itqzMBu40S — Seema Amble (@seema_amble) May 8, 2025 ChatGPT now refers 10% of new @vercel signups, which have also accelerated https://t.co/LzatDz8n8u — Guillermo Rauch (@rauchg) April 9, 2025

Mistral Announces Agents API, with Code Execution, Search, and MCP Support

Mistral last month announced a significant upgrade to its API, capturing the essential features now offered by leading large language models (LLMs). Mistral’s new Agents API resembles OpenAI's Responses API, which was issued in March 2025. It includes: Code execution. Mistral's new Code Interpreter mechanism, with Python in a server-side sandbox. Image generation. Mistral is using Black Forest Lab FLUX1.1 [pro] Ultra. Web search. Mistral offers two versions: a basic web_search classic search, and a and web_search_premium", which enables access to both a search engine and two news agencies: AFP and AP. Document library. It’s Mistral's version of hosted RAG over "user-uploaded documents". Model Context Protocol support. Users can now include details of MCP servers in their API calls. The same new feature rollout across OpenAI (May 21st), Anthropic (May 22nd), and now Mistral (May 27th) within eight days of each other.

OpenAI Claims It Reached $10 Billion In Annual Revenue

WWDC 25: Apple Focuses on a New User Interface and Fails to Deliver a More Personalized Siri [Watch]

A Y Combinator Startup Launches an On-Demand CTO and Founding Engineer

Manus Introduces Slides, Which Creates Structured Presentations on Google

This month, Manus introduced a feature that allows users to create structured slide presentations instantly. With a single prompt, Manus generates tailored slide decks. Interestingly, slides can be edited on the same screen and later exported as Google Slides. An example can be seen here. Manus slides can now be exported as Google Slides! Create with Manus, collaborate everywhere! https://t.co/JlLI9MMoAb pic.twitter.com/MIt0Mwiwu1 — ManusAI (@ManusAI_HQ) June 2, 2025

Google Infuses in Gemini 2.5 Its Family of Models Fine-Tuned for Education, LearnLM

To make the learning process more active, engaging, and effective, Google announced at the I/O 2025 annual event that it's infusing its family of models, fine-tuned for education, with Gemini 2.5 LearnLM. The Gemini’s multimodality feature allows remixing information into any format — audio, video, images, and text. Another Google tool that has been enhanced is NotebookLM, which enables users to upload sources for research and makes the system an expert, presenting the outcome as audio Audio Overviews and Mind Maps.   Additionally, Google announced that it will soon introduce a feature that enables users to convert the content of their notebooks into educational videos. document.createElement('video'); https://iblnews.org/wp-content/uploads/2025/06/Search_Live_-_AI_Mode_ora36b4.mp4 Google continues to enhance its new search modality, called AI Mode, with advanced reasoning, multimodality, web links, the ability to ask follow-up questions, and soon, Deep Search. In April, Google gave U.S. college students a free Gemini upgrade through 2026 final exams. The company is now expanding its offerings to students in Brazil, Indonesia, Japan, and the United Kingdom. Students in these countries will receive free access to the Google AI Pro plan for 15 months, helping them fine-tune their writing, study for exams, and get homework help, along with 2 TB of free storage, NotebookLM, and more. Students globally will also have the ability to create custom quizzes to help them prepare for exams by simply asking Gemini to “create a practice quiz…” on any topic, or base them on uploaded documents such as class notes. The quiz experience provides hints, offers explanations for both right and wrong answers, and provides a helpful summary at the end, highlighting areas of strength as well as those that may benefit from further study. Later this year, the search giant will introduce Sparkify, which will turn users' questions or ideas into short animated videos through the latest Gemini and Veo models.   With Project Astra, Google is prototyping a conversational, personalized tutor that can help with homework. The tools walk users through problems step-by-step, identify mistakes, and even generate diagrams to help explain concepts if they get stuck. This research project will be coming to Google products later this year. Android Trusted Testers can sign up for the waitlist to see a preview. https://iblnews.org/wp-content/uploads/2025/06/draw_example.mp4

......

Today's Summary

Sunday, November 23, 2025

Education technology today is marked by rising AI adoption among educators and innovative personalized learning approaches.

Video News

Loading videos...

Today in AI & EdTech

Sunday, November 23, 2025

AI is transforming the education technology landscape as more teachers adopt intelligent tools, driving forward and adaptive learning experiences.

AI & EdTech Videos

OpenAI Launches Educational GPT Model

Adaptive Learning Platforms Show 40% Improvement

Microsoft Education Copilot Beta Launch

Today in Education

U.S. Department of Education Announces New Funding for STEM Programs

The initiative aims to support science, technology, engineering, and mathematics education.

Global Education Summit Highlights Digital Learning Innovations

Leaders from around the world discuss the future of remote and hybrid learning models.

New Study Shows Benefits of Early Childhood Education

Research indicates significant long-term academic and social advantages for students.