The new image generation model inside of ChatGPT is a major upgrade from the previous Dolly 3 model. This new model is extremely good at generating accurate and photorealistic images. It excels at character consistency, allowing you to upload and restyle various images. Compared to the previous model, this new image generation is a significant improvement in quality and capabilities.

Some key highlights of the new image generation model in ChatGPT:

• Highly accurate and photorealistic image generation
• Excellent character consistency, enabling image restyling
• Versatile use cases, from creating Instagram ads to YouTube thumbnails
• Seamless integration within the ChatGPT platform

This new image generation model is one of the top AI upgrades in recent times, ranking among the best in the industry. It has opened up a wide range of possibilities for users to create visually compelling content with ease, directly within the ChatGPT interface.

Cling AI is a platform that allows you to turn any image into a video. This is a recent upgrade that the author finds particularly useful for creating short films.

The author provides an example of using Cling AI to bring an image generated in ChatGPT to life. They had created an animated-style tree waving gently with one of its branches in ChatGPT, and then used Cling AI to turn this into a video. The result is a visually appealing short clip, with the tree moving realistically and a pigeon added as well, all achieved with a single prompt in Cling AI.

The author notes that these types of short films created using AI tools like Cling AI are getting closer to becoming a reality for YouTube and other video platforms. The ease of transforming static images into animated videos makes the content creation process much more efficient.

Notebook LM, one of my favorite AI apps, recently added a new feature - mind maps. This powerful tool allows you to visually explore the topics within your knowledge base.

To use the mind map feature, simply click on the "Mind Map" button in your Notebook LM interface. This will generate a comprehensive mind map based on the content you've added to your notebook.

The mind map provides a clear overview of the different topics and subtopics you've covered. You can then click on any of the nodes to dive deeper and explore the related information. This makes it easy to identify connections between different areas of your research and uncover new insights.

For example, if you were studying the retail industry in 2025, you could click on the "Loyalty Programs" node to see more details on how retailers are approaching customer loyalty. From there, you can have a discussion with the AI and ask any questions you have, with the system providing references to the relevant sources.

The mind map feature in Notebook LM is a game-changer for organizing and navigating your knowledge. It's a powerful tool that can help you gain a deeper understanding of the topics you're exploring and uncover new connections and insights.

Manis AI is a new platform out of China that allows you to generate websites based on a simple text prompt. This is a powerful feature that can save a lot of time and effort in creating websites.

To use this feature, you simply need to provide Manis AI with a text prompt describing the website you want to create. For example, you could say "Create a website that focuses on AI education and AI tools." Manis AI will then generate the entire website for you, including the pages, content, and even the code.

The generated website includes modern formatting and layout, and Manis AI even does the research to include relevant information and popular AI tools. You can then view the generated website by clicking on the provided link.

One of the key upgrades to Manis AI is the addition of a knowledge base feature. This allows you to provide specific knowledge or content that Manis AI can use when generating the website. This can help ensure the website is tailored to your needs.

Manis AI is currently in a private beta, so you'll need to request an invite to access the platform. There are also paid plans available that provide more credits and the ability to run multiple tasks simultaneously.

Overall, the website generation capabilities of Manis AI are a impressive and time-saving feature, especially for those looking to quickly create AI-focused websites.

Refusion, one of the best music generation platforms, has recently rolled out two major upgrades that are worth highlighting.

The first upgrade is the ability to understand and work with song "stems". This means you can now swap out the vocals or other instruments in a generated song. For example, if you don't like the vocals, you can simply swap them out and replace them with your own. This gives you much more control and flexibility over the final song.

The second upgrade is the option to use "vibe" as a prompt. Instead of generating a new song from scratch, you can now take an existing song that you like the vibe of, and use that as the basis for a new song. This allows you to capture the overall feel and energy of a song, while still customizing the lyrics and other elements to fit your needs.

To try out these new features, simply head to the Refusion platform, upload or generate a song, and then use the "remix" options to swap out the vocals or change the vibe. The process is straightforward and the results can be quite impressive.

Overall, these upgrades to Refusion make it an even more powerful and versatile tool for music creation and customization. If you're looking to generate unique, high-quality music for your projects, be sure to give Refusion a try.

The biggest upgrade to Claude AI in recent weeks is the addition of search functionality. This makes Claude one of the most useful AI assistants, as it can now pull in relevant, up-to-date information from the web to assist with writing and research tasks.

With search enabled, you can start a query by searching for relevant information. Claude will then provide a summary of the key points, along with links to the source material. This allows you to quickly get a high-level understanding of a topic before diving deeper.

Once you have the necessary information, you can then ask Claude to rewrite or expand on the content in its own words. This combines the power of search with Claude's strong writing abilities, making it a formidable tool for tasks like:

• Research and report writing
• Summarizing complex topics
• Generating original content based on source material

The new user interface also makes Claude easier to navigate, with a cleaner design and better organization of features. Overall, the search upgrade positions Claude as a top choice for general-purpose AI assistance, complementing the strengths of other platforms like ChatGPT.

Hen, a powerful AI platform, has recently introduced two exciting new upgrades:

1. Generate Avatar from Text Prompt: Instead of choosing from pre-made avatars, you can now create your own avatar by simply providing a text prompt. This feature allows you to clone yourself and even have your avatar speak in different languages, with the lip-sync accurately matching the audio.

2. Create User-Generated Ads: Hen's new "User Generated" tab offers a range of pre-designed avatars that can be used to create engaging video content, such as TikTok posts, Instagram ads, and more. You can simply type in your script, and Hen will generate a short, customized video for you.

To try out these new features, you can access the Hen platform and navigate to the appropriate sections. Under the "User Generated" tab, you'll find the options to create your own avatar and generate user-generated ads. Experiment with different prompts and scripts to see the impressive results Hen can deliver.

The biggest upgrade that came out in the last week of March was the release of Gemini 2.5 and Gemini 2.5 Pro. These new models completely outperform every single large language model, including the best models from ChatGPT.

Gemini 2.5 Pro is said to be best for coding and complex prompts. You can try it right now for free inside Google AI Studio, and it's also available in the Google Gemini platform if you upgrade your plan.

The Gemini 2.5 Pro model has been put through a test called "Humanity's Last Exam," which is the hardest exam these models have faced. In this test, the best version of OpenAI 4.5 currently has only a 6% accuracy, while the DeepCR1 model is at 8%. However, Gemini 2.5 Pro is at an impressive 19% accuracy, almost tripling the performance of the other top models.

When comparing Gemini 2.5 Pro against other benchmarks, it consistently beats every single model, including the GPT-3 mini-high. This makes Gemini 2.5 Pro one of the most capable language models available right now.

If you want to test the latest models, you can do so in Google AI Studio. Simply click the dropdown menu to access the newest versions, including Gemini 2.0 0 Flash, which has image generation capabilities. This model can perform tasks like adding chocolate to existing images and recreating people in the image, though it may not be an exact replica due to the design choices to prevent deep fakes.

Overall, Gemini 2.5 Pro is a powerful language model that sets a new standard for performance, making it a must-try for anyone interested in the latest advancements in AI.

Grock 3 recently introduced a new feature called "Deeper Search" that takes their deep research capabilities to the next level. With this update, users can now dive even deeper into their research queries, exploring a wider range of sources to gather more comprehensive and insightful information.

The Deeper Search function allows you to find detailed information on topics such as recent funding rounds and current valuations of top AI companies. Instead of just performing a regular deep search, the Deeper Search option breaks down the prompt into more tangible elements, allowing Grock 3 to scour through a larger number of websites and sources to provide a more thorough and detailed output.

Compared to other platforms that offer deep research or deep search features, Grock 3's Deeper Search stands out for its speed and efficiency. While some tools may look at over 150 websites for a single query, Grock 3 finds that after around 37 websites, the necessary information has been gathered, resulting in a faster turnaround time without sacrificing the depth of the research.

The Deeper Search feature on Grock 3 is a valuable asset for users who require comprehensive and up-to-date information on various topics, particularly in the fast-paced world of AI and technology. By leveraging this enhanced research capability, users can make more informed decisions and stay ahead of the curve in their respective fields.

The recent upgrades across various AI platforms have brought significant advancements, revolutionizing the capabilities of these tools. The integration of image generation within ChatGPT has been a game-changer, allowing users to create highly accurate and photorealistic images with just a few prompts. The ability to swap vocals or instrument stems in Refusion's music generation platform has further enhanced the creative potential of these AI-powered applications.

The introduction of search functionality in Claude has made it a more versatile and practical alternative to ChatGPT for general use cases. The impressive performance of Gemini 2.5 Pro in complex benchmarks, outperforming even the best language models, is a testament to the rapid progress in the field of AI.

Additionally, the deeper search capabilities in Grock have provided users with more comprehensive and in-depth research, making it easier to gather valuable insights. The changes to the free trial period and the release of the new "AI-Powered Presentations" course in the author's platform further demonstrate the continuous efforts to enhance the user experience and provide valuable resources.

These upgrades across the AI ecosystem have undoubtedly expanded the possibilities for users, empowering them to leverage these powerful tools for a wide range of applications, from creative endeavors to research and analysis. As the AI landscape continues to evolve, it is essential to stay informed and explore the latest advancements to maximize the potential of these transformative technologies.