Cometthe AI-integrated browser from Perplexity, has officially launched, opening a new chapter for the web surfing experience. More than just a regular search engine, Comet promises to redefine the way we interact with the digital world. This birth is considered a direct challenge to the giant Google Chrome in the global market.

What is Comet?

Last July, Perplexity stirred up the technology world when it introduced an AI browser called Comet. CEO Aravind Srinivas describes Comet as more than just a web surfing tool, but an “AI companion,” capable of turning every internet visit into an intelligent and seamless conversation.

The core difference is that Comet uses Perplexity’s reply engine as default. Instead of returning a long list of links, it provides AI-synthesized answers that are concise, in-depth, and accurate. This is Perplexity’s strategic step to reshape information search habits and compete fairly with Google.

Features that make Comet different

Perplexity has equipped Comet with a series of groundbreaking features, helping it stand out from traditional browsers:

• AI assistant is always on hand sidebar can answer all your questions about the content you are watching. For example, when you’re reading a complex document, this assistant can summarize the main ideas, explain confusing terms, or even assist with related actions like making an appointment.

• The ability to automate complex tasks saves you time, from comparing prices and shopping for products online, managing your schedule by syncing with Google Calendar, to shortening long articles with just one click.

• Smart search engine provides direct answers instead of a list of links, along with cite source clear to ensure authenticity. This helps users not have to open dozens of tabs to compare information like before.
• Switching from the old browser is incredibly simple, allowing users to import all their bookmarks, extensions, and settings in a single tap, removing barriers for new users.

Comet and the confrontation with Google Chrome

Although Google Chrome is dominating with more than 65% market share, Perplexity is still confident that Comet can change the game. Comet’s main weapon is the ability to “always be ready to answer” all users’ questions. When you enter a query, Comet not only provides results but also summarizes, cites sources and makes the entire search process transparent, thereby building strong trust in users.

Initial service packages and challenges

Comet offers a free version that allows users to experience an AI assistant in the sidebar along with tools for news personalization, project management and shopping assistance. To access more powerful AI models, users need to upgrade to a paid plan. Especially, the Max package for the price 200 USD/month also offers a dedicated email assistant.

Download Comet

However, shortly after launch, Comet faced its first challenge when Brave’s research team discovered a security vulnerability. This vulnerability allows a malicious website to inject hidden commands, causing the AI ​​to accidentally execute and risk exposing user data. Luckily, Perplexity quickly released a patch to fix this issue.

Conclusion

The launch of Comet is not only a step forward for Perplexity but also a signal that the future of web browsing is changing. With powerful AI features and a different direction, Comet promises to be a formidable competitor in the market. Is this the beginning of a new era of smarter and more efficient web browsing?

In the digital era, who is changing the way we work, and Codexadvanced programming support tools from Openai, is the most obvious evidence. Not only is a code writing tool, Codex is like a smart “technical assistant”, helping programmers save time, optimize processes and focus on creativity. With the ability to automate complex tasks, this tool is opening a new future for the software development industry.

About codex

Codex is the breakthrough of Openai, built based on the optimized AI model Codex-1inherited from strong O3 platform. This tool is trained with specialized data sets, allowing to create clean, accurate source code and meet the user requirements. According to TechCrunch, Codex not only helps to reduce the burden of repeated programming jobs, but also offers superior performance, minimizing errors and improving codes.

The highlight of Codex lies in the ability to integrate seamlessly with the code warehouse via github, operating in a safe and isolation cloud (sandboxed). This ensures security and allows it to handle programming tasks from simple to complex in about 1 to 30 minutes, depending on the difficulty.

Outstanding features of codex

Codex does not only stop writing code but also provides a series of smart features, helping programmers work more effectively:

• Create a new source code: The tool can automatically produce code for new functions based on specific requirements, saving design time from the beginning.
• Fix code error: Detect and fix errors quickly, minimize manual debugging time.
• Answering the question: Codex answers questions related to the code facility, from the overall structure to the details of each function, helping the programmer better understand the project.
• Proposals: Giving suggestions to edit code through pull request, supporting team development team effectively.
• Automatic testing: Code testing tools to ensure quality and early detection of potential problems.

Also, codex allows customization via file Agents.mdwhere users can add instructions on the standard programming or project context. This helps it work more accurately, meeting the expectations of the development team.

Instructions on how to use codex

Codex is integrated directly into the left toolbar of the web application Chatgptwith a simple and friendly interface. Users can use two main buttons:

• Code: Assign specific programming tasks.
• Ask: Ask questions about code or related issues.

The processing process is clearly displayed through the Terminal Diary, reference documents and test results, making it easy for users to monitor the job status. Codex also supports copying environmental development into separate containers, ensuring flexibility and security.

Target users and safety

Currently, Codex is only available for users of packages Chatgpt Pro, Enterprise and Team. Openai plans to expand access to users Plus and Edu in the near future. In particular, Codex is designed with high safety, rejecting the requirements of creating malware or harmful software.

However, Openai recommends that users carefully check the code created by Codex before integrating into the project, to ensure compliance with programming standards and avoid security risks.

Experience now HERE

Conclude

Codex Ofenai is a revolutionary tool, providing superior automation and effective support for programmers. With friendly interface, diverse features and flexible customization, Codex is shaping the future of the software development industry. However, to make the most of the benefits, users need to check the code carefully and comply with the policy before use.

On April 29, 2025, billionaire Elon Musk announced the early beta version of Grok 3.5marking a new step in the global artificial intelligence race. With the ability to reason from the basic principle, Grok 3.5 promises to bring a breakthrough experience, not only reproducing information but also creating unique knowledge. Let’s find out the outstanding features of this version and the reason it is expected to change the way we use the technology.

What is Grok 3.5?

Grok 3.5 is the upgraded version of Grok 3, developed by Xai – The company is behind the mission to accelerate the discovery of human science. Introduced by Elon Musk, Grok 3.5 focuses on the ability to deduce from the “first-principles reasoning” (First-Principles Reasoning), which allows answering complex technical questions. Regardless of online data. This helps Grok 3.5 create a unique, accurate and creative answer, especially in areas such as aerospace, energy and chemical electrical.

Not only stopping at text processing, Grok 3.5 is also improved with multimodal features, including image analysis and creating interactive content based on user data. With the support of Colossus supercomputer and superior calculations, Grok 3.5 is being located to compete directly with the top AI models like Chatgpt.

The highlights of Grok 3.5

1. Inferred from the first principle

The most prominent feature of Grok 3.5 lies in the ability to infer independently, not based on the data available on the Internet.

For example, when asked about the missile or mechanical engine, GROK 3.5 can give answers based on logical arguments and internal knowledge, instead of synthesizing information from online sources. This not only ensures accuracy but also helps avoid “hallucinations” – the common problem in many other AI models.

2. Multimumical integration

Grok 3.5 upgrade multimodal processing capacity, allowing Image analysis and create interactive content. For example, if you provide a living room photo, Grok 3.5 can propose the appropriate interior size or even give creative design ideas. This feature opens up applied potential in many fields, from interior design to education and research.

3. Outstanding performance

Trained on the Colossus supercomputer with 200,000 GPU, Grok 3.5 use calculation amount 2.5 to 4 times higher Compared to the previous version. The result is outstanding performance on standard tests such as AIM (Mathematics) and GPQA (PhD science). These improvements make Grok 3.5 a powerful tool for researchers and engineers.

4. Integrating real -time data

Grok 3.5 Take advantage of data from the X platform to provide real -time update information. In addition, the reference memory feature allows it to memorize the context, providing a more seamless chat and personalization experience. The ability to process multi -language is also improved, helping to reach the global user.

When is Grok 3.5 released?

Currently, Grok 3.5 Beta version has been deployed from In early May 2025but only for users of Supergrok package – Xai’s premium registration package.

As expected, this version will be expanded to X Premium+ users and free packages within 4-6 weeks after the testing stage. Developers will also soon access Grok 3.5 via API, opening up the opportunity to integrate this AI into third -party applications.

Conclusion

Grok 3.5 is not only a new version of AI, but also a testament to Xai’s vision in creating artificial intelligence capable of creative and independent thinking. With outstanding features such as reasoning from the first principle, multimodal integration and high performance, the tool promises to open a new era for AI.

TikTok – The short video social network that is taking the world by storm, has just launched a promising new feature called “Sound Search” (Sound Search). This feature allows users to easily search for songs by singing, humming the melody or streaming the audio they are looking for. This marks a new step in the application of sound recognition technology on social networking platforms, and poses a direct challenge to giants like Apple's Shazam and Google's Hum to Search.

Find songs easily with Sound Search

With “Sound Search,” finding your favorite song has never been easier. Instead of having to remember the lyrics or the singer’s name, users just need to hum the melody or sing a short song, and TikTok will automatically recognize and display search results. Users can even stream audio from other sources, such as radio or TV, for TikTok to recognize the song.

What makes Sound Search TikTok different?

Beyond just identifying songs, “Sound Search” also takes full advantage of the TikTok video platform. After successfully identifying a song, the app will display a list of popular videos that use that sound clip. This makes it easy for users to discover creative and interesting videos related to the song they are looking for.

Currently, “Sound Search” is only being tested for some TikTok users in certain regions. It is unclear when this feature will be widely deployed globally. However, with the potential it brings, “Sound Search” promises to be a breakthrough in the user experience on TikTok, while creating fierce competition in the sound recognition application market.

See also: How to comment on TikTok with pictures

Comparison with Shazam and Hum to Search

Shazam

Shazam is an audio recognition app owned by Apple that has long been a top choice for users when searching for songs. However, Shazam does not currently support voice recognition. This app focuses on identifying songs playing from external audio sources.

Hum to Search

Google has also integrated a “Hum to Search” feature into the Google Assistant and Google Search apps, allowing users to search for songs by humming a tune. However, “Hum to Search” has not yet been integrated into video platforms like YouTube.

How to use Sound Search feature

If you are lucky enough to be part of the beta program, you can use the “Sound Search” feature by:

Step 1: Open the TikTok app.

Step 2: Click the microphone icon in the search bar.

Step 3: Select “Search for audio”.

Step 4: Sing, hum a tune, or stream the audio you're looking for.

Step 5: TikTok will automatically identify the song and display search results.

Advantages of Tiktok Sound Search

• Combine the power of video platforms when displaying popular video lists using songs, helping users discover diverse and interesting content.
• Friendly interface, easy to use, integrated directly into TikTok's search bar, convenient for users.
• Accurate and fast audio recognition, using advanced AI technology to identify songs from singing, humming or live streaming.
• Can be developed to recognize not only songs but also other types of sounds, such as sound effects, animal sounds, etc.
• With the popularity of TikTok, “Sound Search” has the potential to become the most popular sound search engine on the market.

See also: How save songs on TikTok to Spotify super simple

Conclusion

With the launch of “Sound Search“, TikTok is affirming its pioneering position in applying new technology and improving user experience. This feature not only helps users find songs more easily but also opens up new creative possibilities on this short video platform. In the future, “Sound Search” promises to continue to be developed and improved, contributing to consolidating TikTok's leading position in the race for global user market share.

Currently, Google Search is the most popular search engine in the world. The company has continuously upgraded this tool by applying AI, audio search and Google Lens. Now, Google has added more functionality Circle to Search, helping users do more tasks conveniently and quickly. So what is Circle to Search? Let's explore with AnonyViet!

What is the Circle to Search feature?

With the “Circle to Search” feature, users can search for everything they desire without having to switch back and forth between two applications. To enable this feature, users simply hold down the home button or navigation bar, then draw a circle, highlight or tap on the object they want to search for, whether it's clothes, food, furniture or anything else. anything else.

Similar to Google Lens, search results will appear right on the screen you are using. Furthermore, thanks to the advancement of AI, you can draw a circle around an individual object or piece of text, such as a dish or a sign, and ask related questions. This feature will answer you accurately and quickly.

What makes this new feature stand out?

This new feature will bring users many benefits, such as:

Reduce time and effort when searching for information

You don't have to enter keywords, open new tabs, or switch apps. Just draw a circle and get instant results.

Optimize user interaction experience

You can search by drawing, writing, taking a photo, or touching an object. You can also ask by voice or text and get intelligent answers from Google.

Easily explore the world around you

You can learn about anything you find interesting, from a flower to a structure. You can also compare prices, reviews, order or shop online easily.

Which devices will Circle to Search be available on?

On January 31, 2024, the new feature will be officially announced and appear on leading Android devices, including the brand new Google Pixel 8 series. Besides, this feature is also supported on smartphones of the series Samsung Galaxy S24 and you can own these high-end phones that same day.

Epilogue

Circle to Search is an important step forward in improving the user search experience, especially on mobile devices. Circle to Search not only helps users save time and effort, but also delivers more accurate and richer search results. What do you think about this new feature from Google?

Microsoft has officially launched the application Copilot on the iOS operating system, after being available on Android. This is a multi-purpose AI chatbot application that can help you answer questions, compose emails, summarize text, create images from text and many other features. You can experience this application on iPhone and iPad, with an interface optimized for large screens.

Outstanding features of Microsoft Copilot

Microsoft Copilot is an AI chatbot application developed by Microsoft, based on OpenAI's GPT-4 platform. This application can understand users' natural language and answer questions related to many different fields, from education, health, tourism, to technology, business, culture,…

In addition, Microsoft Copilot can also help you perform office tasks such as drafting emails, summarizing documents, creating presentations, writing articles, writing programming code… You just need to enter your request , and the application will automatically generate suitable content.

A unique feature of Copilot is the ability to create images from text, through integration with the DALL-E3 text-to-image generator. You can ask the application to draw a picture according to your ideas, or create an illustration for the article.

For example, this is 1 of 4 results when I asked Copilot to draw a dragon image for the new year 2024:

I asked Copilot to change the dragon to make it cuter, and this is the result:

How to use Microsoft Copilot on iOS

To use Microsoft Copilot on iOS, you need to download the app from the App Store or via THIS LINK and sign in with a Microsoft account. You can then start chatting with the app by entering a question or request.

You can also use voice to communicate with the application, by tapping the microphone icon in the lower right corner. The app will automatically convert voice to text and respond to you.

Review of Microsoft Copilot on iOS

Microsoft Copilot is a potential AI chatbot application, providing users with many useful and interesting features. The application is capable of understanding the user's natural language and answering questions quickly and accurately.

Epilogue

Overall, Microsoft Copilot is an AI chatbot application worth experiencing, especially if you are a technology lover and want to explore the possibilities of AI. You can download the app from the Apple App Store and try it today.

Google launches a series of Christmas mini games: This is a special gift that Google gives users on the occasion of the year-end festival. You can experience diverse, rich and interesting games on your browser, no need to download or install. These games not only help you entertain, but also train your intelligence, reflexes and agility.

Google launches a series of Christmas mini games – Follow Santa

Following Santa Claus is a Christmas game developed by Google, allowing users to experience attractive and useful games on the theme of Christmas. To help children explore more comfortably, the game uses familiar characters, such as reindeer, Santa Claus and elves.

There are also games that help you learn about the different ways people around the world celebrate Christmas.

How to play Follow Santa Claus

To play Follow Santa on Google, you just need to follow these simple steps:

Step 1: Go to Google.com or your browser's search bar.

Step 2: Enter the keyword “Christmas” or “Christmas”.

Step 3: A red gift box will appear in the left corner of the screen.

Step 4: Click on the gift box.

You will find many exciting games to explore.

Benefits of the game Follow Santa Claus

Following Santa Claus is not only an entertaining Christmas game, but also a useful tool to educate and nurture children's souls. Games help children:

• Improve knowledge and skills in various subjects, especially programming, a very important skill in today's digital age.
• Develop logical thinking, creativity and problem solving, through games that challenge the brain and imagination.
• Expand your perspective and understanding of the world, by learning about different cultures, histories and geographies.
• Strengthen feelings and bonds with family and friends, through sharing and participating in Christmas activities together.

Epilogue

Google launches a series of Christmas mini games – This is a commendable initiative by Google, aimed at bringing joy, relaxation and a festive atmosphere to users. You can play these games anytime, anywhere, just need an internet connection and a browser. You can also invite friends and family to join, for added bonding and fun. Have you tried playing Follow Santa on Google? Hurry and experience it and share your feelings with us!

NotebookLM is a new note-taking application from Google, integrating the Gemini Pro AI language model. This application allows users to create text content based on custom, unlimited and secure data sources.

Introducing NotebookLM

NotebookLM is considered a “personalized AI” because it only uses information from documents provided by users, not taking data from the Internet or external sources. This application can help users write an article, create an outline, summarize a document, or support learning effectively and accurately.

Google announced NotebookLM at the I/O 2023 event earlier this year under the name Project Tailwind. Google later changed the brand name and said the technology was based on Gemini Pro – a newly launched AI language model from Google, said to be more powerful than OpenAI's GPT-4.

What's different about NotebookLM?

NotebookLM differs from current AI chatbots in that it does not use big data information on the Internet, but only retrieves information from users' customized document sets. This helps it deliver text content tailored to users' specific needs, without being influenced by irrelevant or inaccurate information online.

Besides, NotebookLM also ensures the security and privacy of users' personal data, as this data is not used to train the AI ​​model.

NotebookLM is now available to users in the US, and Google has also been providing access to pre-registered users for the past five months.

You can try out NotebookLM by visiting the app's page here (You remember use VPN change IP to US)

How to use NotebookLM

The Notebook appears at the top of the chat box with a tabbed grid user interface that helps you save NotebookLM responses, snippets from your sources, and notes you've created.

You can select multiple notes and have NotebookLM summarize them, combine them into a single note, or create other broader content, like outlines, study guides, or convert your notes into an email newsletter, text structure or marketing plan draft. You can also export to Google Docs with keyboard shortcuts.

NotebookLM also allows you to select note sources (from the sidebar) to chat specifically with only these groups of information, with a source limit per notebook of up to 100.

Epilogue

NotebookLM is an innovation in the field of note taking software. This product exploits the capabilities and potential of artificial intelligence language models to assist users in retrieving necessary information from their existing content and text. Have you experienced NotebookLM yet?

Shutterstock launches Creative AI – A revolutionary new product in the field of digital art. What is Creative AI? What new features are there? Let’s find out in this article!

Shutterstock launches Creative AI – AI photo editing tool

Shutterstock, one of the world’s largest stock photo repositories, has just launched a new feature called Creative AI. This is a tool that allows users to edit stock photos right on Shutterstock’s website using artificial intelligence.

Creative AI functions

Shutterstock launches Creative AI, providing six main functions, helping users customize stock photos as desired:

• Magic Brush: Allows you to add or remove objects in an image by drawing on the image.
• AI Image Generator: Allows creating new images from some keywords or an original image.
• Expand Image: Allows you to expand the size of an image without losing details or distorting the image.
• Variations: Allows creating many different versions of an image by changing color, lighting, perspective, etc.
• Design Assistant: A virtual assistant capable of chatting with users and giving suggestions on how to design photos.
• Background Remover: Allows you to delete the background of the image and replace it with another background.

How to use Creative AI

To use Creative AI, users need a Shutterstock account and are free to use for the first month. Users can then select a stock photo from Shutterstock’s photo library and click the “Edit with creative AI” button to start editing.

If users want to use photos created or edited by Creative AI for commercial purposes, they will have to pay additional copyright fees to Shutterstock according to each type of use.

Epilogue

Shutterstock launches Creative AI – A huge step forward in the history of digital art. With Creative AI, you can create unique and high-quality digital works with just a few simple steps. You can also explore and get inspired by millions of works generated by Creative AI. Have you tried this invitation feature yet? Let us know your opinion!

Project Res Up is a video upscaling AI tool introduced by Adobe during the MAX Sneaks 2023 event. It uses diffusion-based sampling technology to transform low-quality videos and GIFs into high-quality content. AnonyViet will introduce you to Project Res Up, how it works, and its potential applications in the field of video editing and creative content.

Introducing Project Res Up – AI video sharpening tool

Adobe is one of the leading companies in the field of graphic design and video processing. Recently, the company introduced a new tool that uses artificial intelligence (AI) to improve the quality of low-resolution videos or GIFs. The tool is called Project Res Up and was demonstrated at the Adobe MAX 2023 event.

How Project Res Up works

Project Res Up uses a technique called diffusion-based sampling to increase video resolution and sharpness. The technique relies on an AI model trained on millions of photos to learn how to recreate details lost during video compression. When applied to video, Project Res Up will analyze each frame and add new pixels to make the image clearer.

Result

Adobe tested Project Res Up with a variety of videos, from classic movies to animal clips. The results show that this tool can increase video resolution three times and significantly improve details such as wrinkles, fur, flowers, etc. Images sharpened by Project Res Up look very cool. natural and not distorted or artificial.

With this tool, the company took a clip from the movie The Red House (1947) and increased its resolution from 480 x 360 to 1280 x 960 pixels. In sharpened video, details in movies have been significantly improved. Cracks, blurriness, and noise on the screen have been eliminated, colors have been enhanced and sharpened, and the actors’ faces have been made clearer.

Not only that, Adobe has also applied this tool to poor quality videos such as animal videos. This tool reduced blurriness and added some subtle details like the elephant’s skin folds in the video were sharpened. The highlight is that these folds are restored naturally and don’t look artificial like many other tools used to sharpen videos.

Prospects

Present, Project Res Up is still in the testing phase and has no official launch date yet. However, Adobe has said it intends to integrate the tool into its video processing applications, such as Adobe Premiere Pro or Adobe Express. This will be a useful feature for those who want it Improve the quality of the video without having to use professional filming equipment.

In addition, Project Res Up can also be used to restore precious movies or GIFs that have faded over time.

Epilogue

Project Res Up is a potential and innovative AI video upscaling tool from Adobe. It not only helps you improve the quality of old videos and GIFs, but also opens up new possibilities in the field of creative content. You can use Project Res Up to sharpen images and increase video resolution easily and quickly.