Huge Claude Update

Welcome to this new edition of AI News You Can Use your trusted source for the latest AI developments, tools, and technologies that are shaping our world. From new generative features in AI to revolutionary gaming experiences, here's what you need to know.

Anthropic Makes Big Moves with Claude 3.5 and More

Anthropic continues to make waves in the AI space with a series of notable updates. This week, they rolled out several new features for their Claude model, including an exciting enhancement that gives Claude the ability to interact with visual elements in PDFs. Unlike other language models that only analyze text, Claude can now process images, graphs, and visual data within PDF documents— a huge leap forward, especially for scientific and technical content.

If you're working with complex PDFs, this is a game-changer. And for developers, this feature is also available through the API, allowing for custom applications that process visual PDFs automatically.

In another surprise move, Anthropic raised the pricing for their Claude 3.5 model. This is the first price hike in the LLM space in quite some time, raising questions about future pricing trends and what this means for the competition. But for now, users of Claude's web interface won’t see these price changes, as they currently apply only to API usage.

Anthropic also introduced a new voice input feature for their mobile app, allowing users to speak directly to Claude to input text. While this feature doesn't yet support advanced voice outputs, it's a convenient option for those who prefer speaking over typing.

OpenAI’s Latest: ChatGPT Gets a New Domain

OpenAI has taken a small but meaningful step to streamline access to its ChatGPT tool by purchasing the domain chat.com. Now, typing in the address redirects directly to ChatGPT, simplifying the process and making it even easier to use. While this is more of a minor usability improvement, it's part of OpenAI’s ongoing efforts to make its AI accessible and user-friendly.

ChatGPT Search: A Better Way to Find Information

If you've been following along, you know we've been diving into the new ChatGPT-powered search feature. After using it exclusively for a week (and avoiding Google altogether), the results were surprisingly positive. ChatGPT's search quality is on par with, if not better than, other tools like Perplexity, offering high-quality answers for simple queries—whether it's currency conversion or weather forecasts.

While the experience isn't as instantaneous as Google, it offers a deeper, more contextual understanding of your queries. The hope is that, with more time and user data, ChatGPT's search function will evolve into a powerful alternative to traditional search engines. It’s a personal assistant for your searches—one that could soon know what you actually want to find, not just what you type.

Google’s AI-Powered Search Expands

Not to be outdone, Google has been expanding its AI-driven search features. Their "AI overviews"—a response that sits above traditional search results—are being rolled out globally. Initially met with some skepticism due to issues with accuracy, the feature has seen notable improvements.

This move signals that Google is ready to take on the AI-powered search race head-to-head with OpenAI's ChatGPT. Although many of us are still relying on traditional search for now, AI-based search engines are increasingly offering a more personalized, holistic way to find answers.

X AI’s Gro: A New Developer Playground

X AI has introduced a new developer playground for its Gro model, offering an API with $25 of free credits to get started. The platform’s clean interface and simplicity make it a standout, but the actual performance of Gro still lags behind top models like OpenAI’s GPT and Anthropic’s Claude. However, the integration with Twitter’s firehose of real-time data gives it a unique edge, and with backing from investors like Nvidia, Gro could be a serious contender in the AI space soon.

AI-Generated Minecraft: A First of Its Kind

In a groundbreaking move, we’re witnessing the very first fully generative game engine that allows you to play a game like Minecraft without any pre-existing code or data running in the background. Every element—every block, interaction, and scene—is generated in real-time as you play. This could revolutionize not only gaming but digital experiences as a whole. Although the initial execution might not match up to traditional Minecraft, the potential for this type of real-time, dynamic world creation is enormous.

While still in early stages, this technology could pave the way for a completely new type of interactive digital experience, where everything is generated on the fly based on player interactions.

AI Video Tools Get More Advanced: New Camera Controls in Runway Gen 3

AI video creation continues to evolve, and this week we’re seeing even more refined controls in platforms like Runway’s Gen 3. With the introduction of detailed camera control, users now have the ability to manipulate shots in a way that wasn’t possible before—such as adjusting the intensity of camera movements and angles. This is a huge step forward for AI-driven video content, providing creators with new tools for storytelling and cinematography.

Google’s "Learn About" Experiment: A New Way to Learn with AI

Google is always experimenting with new AI tools, and their latest initiative, Learn About, is a great example. This AI assistant helps you learn new skills by creating a custom learning plan based on your interests. Whether you want to dive into prompt engineering or another area, the tool offers a structured approach to learning, providing interactive prompts and explanations. While currently only available in the U.S., this could be a valuable resource for anyone looking to acquire new skills in a more guided, engaging way.