Ad
Short

Meta is developing new AI models for images, videos, and text under the codenames "Mango" and "Avocado." The release is planned for the first half of 2026, according to a report from the Wall Street Journal citing internal statements by Meta's head of AI Alexandr Wang. During an internal Q&A with Head of Product Chris Cox, Wang explained that "Mango" focuses on visual content, while the "Avocado" language model is designed to excel at programming tasks. Meta is also researching "world models" capable of visually capturing their environment.

The development follows a major restructuring where CEO Mark Zuckerberg personally recruited researchers from OpenAI to establish the "Meta Superintelligence Labs" under Wang's leadership. The market for image generation remains fiercely competitive. Google recently released Nano Banana Pro, an impressive model known for its precise prompt adherence, and OpenAI quickly followed up with GPT Image 1.5 a few weeks later. Most recently, Meta introduced the fourth generation of its Llama series in April and is currently collaborating with Midjourney and Black Forest Labs on the Vibes video feed.

Short

European AI startup Lovable has raised $330 million in a Series B financing round, bringing its valuation to $6.6 billion. Google led the round, with additional backing from Nvidia, HubSpot, and Deutsche Telekom.Lovable, which allows users to build web applications through simple text prompts, has quickly become one of the fastest-growing startups in the world. The company says its Sweden-based team of 120 employees hit $100 million in annual recurring revenue just eight months after reaching its first million. This follows a massive summer for the company, where it secured one of the largest European Series A rounds on record - raising $200 million at a $1.8 billion valuation.

Lovable calls its mission the "Age of the Builder" - a time when people without coding experience, such as teachers, marketers, or founders, can turn their ideas into working software on their own. Despite its rapid growth, Lovable still trails market leaders like ChatGPT in total user numbers. The space is also becoming increasingly competitive: US rival Cursor recently raised $2.3 billion, highlighting both the intense competition in AI-assisted software development, but also the difference in investment volume.

Ad
Ad
Short

Mistral AI has released Mistral OCR 3, an updated version of its document analysis model. The system goes beyond basic extraction—it can interpret cursive handwriting, dense form layouts, and complex table structures, including linked cells. According to the company, this third version outperforms its predecessor in 74 percent of cases, showing particular strength in handling handwriting, scanned forms, and complex tables. OCR 3 also stacks up well against Deepseek's specialized character recognition model.

The model is available through an API or the Document AI platform introduced in May. Pricing sits at two dollars per 1,000 pages, with discounts available for bulk orders. The French company—which recently secured a large investment from chip manufacturer ASML—is using this release to solidify its position in document recognition, even as its current generation of open-weight language models trails behind commercial competitors from the US.

Short

OpenAI is launching a new learning platform for journalists and publishers called the "Academy for News Organizations." Developed in collaboration with the American Journalism Project and the Lenfest Institute, the initiative aims to teach media organizations how to effectively use artificial intelligence. The program offers on-demand training, practical examples for research and translation, and guidance on establishing internal AI guidelines. OpenAI says the goal is to help editorial teams work more efficiently, freeing up time for core journalistic work.

According to OpenAI, the Academy was developed with critical industry issues in mind, including concerns about job displacement and the reliability of AI-generated content. The platform builds on existing partnerships with major publishers like News Corp and Hearst, with plans to expand the offering next year. These educational initiatives might also be an attempt to smooth over tensions in the industry - while OpenAI courts publishers with tools and training, it is simultaneously battling copyright lawsuits from major media companies like the New York Times and Ziff Davis.

Short

OpenAI has started accepting submissions for ChatGPT apps, which will populate a new directory following a review process. These applications allow users to perform specific actions directly within conversations, such as ordering food. The directory is located in the Tools menu, and users can launch specific apps simply by using the "@" command. While a software development kit (SDK) is currently available in beta, the first batch of tested applications is scheduled to launch in early 2026.

On the security front, OpenAI requires that apps remain suitable for general audiences and request only essential user information. At this stage, developers can link from their ChatGPT apps to external websites or native apps to complete transactions for physical goods. However, the company is exploring additional monetization options—including for digital goods—and notes that it has been collaborating with PayPal for several months. This rollout follows October's Dev Day, where OpenAI introduced the Apps SDK alongside its AgentKit for autonomous AI agents.

Ad
Ad
Short

OpenAI is in early talks with investors about a massive funding round that could push the company's valuation to around $750 billion, according to The Information. The company could raise tens of billions of dollars, potentially as much as $100 billion.

The discussions are still in their early stages, and nothing is set in stone. At this valuation, the deal would mark a 50 percent jump from OpenAI's last share sale in October.

Amazon is also in talks to invest $10 billion or more. It's the kind of circular AI deal that's become common: Amazon hands OpenAI cash, and OpenAI turns around and spends it on Amazon's chips and cloud services.

According to The Information, OpenAI has reached an annualized revenue run rate of $19 billion, keeping the company on pace to hit its $20 billion target by year's end. The company is projecting $30 billion in revenue for 2026, rising to around $200 billion by 2030. But these ambitious growth targets come with an enormous cash burn of roughly $26 billion for this year and next.

Short

Renowned mathematician Terence Tao has proposed a new way to think about AI capabilities. On Mastodon, Tao questions whether true "artificial general intelligence" (AGI) is actually achievable with current AI tools. His alternative: "artificial general cleverness" (AGC).

According to Tao, "general cleverness" means the ability to solve complex problems using partly improvised methods. These solutions might be random, rely on raw computing power, or draw from training data. That makes them something other than true "intelligence," but they can still succeed at many tasks, especially when strict testing procedures filter out incorrect results, he says.

"This results in the somewhat unintuitive combination of a technology that can be very useful and impressive, while simultaneously being fundamentally unsatisfying and disappointing."

Terence Tao

In humans, cleverness and intelligence are linked, but in AI they're decoupled, Tao argues. The mathematician has recently spoken positively about how AI has sped up his own work.

Ad
Ad
Google News