Insane AI Updates This Week You Nearly Missed!

This week has been wild with all the AI news and breakthroughs and I can't wait to discuss them with you. In this video, we'll cover the top AI updates of the week, like Fraud GPT, which is like the negative version of ChatGPT for hackers, an app called Text with Jesus, and Meta's grand gesture of releasing Llama 2 as open source. We'll also dive into Google's Robotics Transformer 2, a cool new AI feature from Adobe Photoshop Stability, AI's new model Stable Diffusion XL and LinkedIn's new AI to boost your profile and networking. Alright, let's start with some scary news. Cybersecurity experts have identified a new AI tool called Fraud GPT. Circulating on the Dark Web and Telegram channels since July 22, 2023, Fraud GPT has been advertised as an all in one solution for cybercriminals. Some of its features include crafting spear phishing emails, creating undetectable malware, generating phishing pages, identifying vulnerable websites, and even offering tutorials on hacking techniques. It is based on generative AI and ChatGPT technology, which means it can produce realistic and convincing content using natural language processing. The tool can also learn from its interactions with users and adapt to different scenarios. And this is not the first malicious AI tool out there. Just two weeks Agonext discovered WormGPT, a another AI powered bot that can spread malware through social media platforms. These tools pose a serious threat to online security and privacy. They can bypass traditional detection methods and trick unsuspecting victims into revealing sensitive information or downloading malicious files. So how can we protect ourselves from these AI powered attacks? Experts suggest that we need to use more advanced and innovative cybersecurity defenses that leverage AI for good. We also need to be more vigilant and aware of the signs of phishing and malware. And most importantly, we need to educate ourselves and others about the potential risks and benefits of AI technology. Now let's move on to some more uplifting news. Have you ever wanted to chat with Jesus? Well, now you can, thanks to a new app called Text with Jesus. This app is a revolutionary AI powered chatbot that allows you to engage with your faith through interactive conversations with Jesus Christ. This is the Apostles and other biblical figures. The app uses advanced AI and ChatGPT technology to create realistic and personalized dialogues based on the wisdom of the Scriptures. You can ask questions, seek guidance, share your thoughts, or just have a friendly chat with your favorite religious characters. The app also lets you explore the stories and teachings of the Old Testament through conversations with its prophets and other religious figures. It's designed for devoted Christians who want to deepen their connection with the Bible. It is also suitable for all ages and can be used as a family activity or a learning tool. The app is free to download, but requires a premium subscription to unlock access to the full range of figures. What do you think of this app? How do you feel about using AI to simulate religious figures? Let me know in the comments below. Now, if you're into design or editing, you probably know Adobe Photoshop as one of the most powerful and popular tools in the industry. Well, for Photoshop just got even more powerful and popular with its new AI feature, Generative Expand, which lets users add more to their images beyond the original size. Imagine you have a photo, but part of it is missing or cut off. Now, with this tool, you can expand the photo and the software will automatically fill in the missing parts for you. All you need to do is drag your photo to the size you want, hit the Generate button, and let Photoshop work its magic. If you want something specific in the added space, you can type in a prompt and the tool will try to include that in the image. Plus, the added content becomes a new layer in Photoshop, so if you don't like it, you can simply remove it. Adobe is also making sure that the content generated doesn't have anything inappropriate. They've put in checks to avoid any issues. Many might say that other platforms already had this feature, but having it directly in Photoshop, which is used by millions worldwide, is a big deal. And while right now you can't use this tool for commercial stuff, Adobe plans to change that later this year. They're also adding support for many more languages in Photoshop, making it even more user friendly. Speaking of Generative AI, there's another breakthrough that you need to know about. So Stability AI has launched a new image creating tool called Stable Diffusion xl, which they're calling their best one yet. This tool can make really detailed images, and it's available for everyone to use on platforms like GitHub, Clipdrop, and Dream Studio. When it comes to how it works, think of it like a sophisticated artist with 3.5 billion brushes, and it can now make images with better colors, shadows and lighting. The cool thing, it can even create images from simple sentences you give it. They're also working on a feature to create specific images, but that's still in the testing phase. The new version can also fix parts of an image or add more to it, kind of like what Adobe does. However, there's some concern about how people might misuse this tool, especially since it's open to everyone. Some are worried about creating fake videos or using artists works without permission. Stability AI responded by saying they're working to make the tools safer and they respect artists rights. They're even facing some legal challenges because of these concerns. Joe Penna from Stability AI mentioned they're committed to improving the tool's safety. Plus, this tool will also be available on Amazon Web Rock, a popular platform for AI tools. Imad Mostaq, the head of Stability AI, is proud of this new release and believes it'll benefit the AI world Alright now. Meta has just made a grand gesture by releasing its own large language model, llama2, as open source. That means anyone can download, use and modify LLAMA 2 for free for any purpose. LLAMA 2 is a successor to Meta's LLAMA 1 language model, which was released in the first quarter of 2023. It is pre trained on 2 trillion tokens of public data and is designed to enable developers and organizations to build generative AI powered tools and experiences. It can generate natural language responses based on user inputs as well as images, code, music and more. There are three versions available at the moment ranging from 7 billion to 70 billion parameters. The larger the model, the more powerful and diverse it is. Llama 2 also has a fine tuned version, Llama 2 Chat, which leverages over 1 million human annotations to create more engaging and coherent conversations. Conversations and it's not only open source, but also open innovation. Mieta has partnered with Microsoft, Amazon, Hugging Face and other tech companies to make Llama 2 accessible and compatible with various platforms and frameworks. They have also established an open innovation AI research community to promote transparency and collaboration between academic partners doing LLM research. But why is Meta doing this? According to Nick Clegg, Meta's President of Global affairs, the company believes in an open approach to AI that gives everyone a stake in the AI development process. By opening llama2 to the public, Meta hopes to foster more innovation, diversity and responsibility in the AI field. Meta also hopes to address some of the ethical and social challenges that come with large language models such as bias, privacy and accountability. Llama 2 is a bold move by Meta to democratize AI and empower the global community of creators and research. It also shows that Meta is confident in its position as a leader in the AI industry and is willing to share its resources and expertise with others. Lately, every week's AI news seems to include Google now they've unveiled its new Robotics Transformer 2 RT2, an AI system that can translate Internet based knowledge into robotic actions. It is based on transformer models which are a type of neural network that can learn from sequential data such as text or speech. It can take natural language instructions from humans or web sources and convert them into executable plans for robots. Basically, RT2 is like a bridge between the Internet and robots. This AI can read or listen to something online like a sandwich recipe and then tell a robot how to make it. But it can also learn from its own experience and improve its performance over time because it uses reinforcement learning techniques to reward or penalize the robot based on the outcome of its actions. RT2 can also use self supervised learning techniques to learn from unlabeled data by generating its own labels. This is a big deal because it means robots can learn new things directly from the web without us always guiding them. It's a step closer to having robots that can do a range of tasks, whether that's helping around the house or assisting with shopping. Google dreams of having robots that make our lives easier and with RT2 they're definitely on the right path. Plus they hope this sparks more cool discoveries in the world of robotics and AI. Alright, finally, let's talk about LinkedIn's new AI powered tool that aims to help users improve their professional profiles and networking skills. The tool is called LinkedIn coach and it uses natural language processing and sentiment analysis techniques to analyze users profiles, messages, posts, comments and endorsements. Think of it like a buddy who gives you advice on making your profile stand out or helping you figure out what to say when reaching out for job opportunities or just saying thanks. Plus, it can give you a nudge if it thinks you're posting too much, using the wrong tone or missing out on making connections. But while it's super helpful, not everyone's a fan. Some folks feel it takes away the unique touch from their profiles, making them sound too cookie cutter. Others worry it might be a bit too nosy peeking into their private data. And there's a concern that relying on this tool might mean we're not thinking for ourselves as much, especially when it comes to networking. Alright, that concludes this week's AI News. If you enjoyed this video, don't forget to hit the like button, subscribe to my channel and click the Bell icon to stay updated on all my future uploads. Thanks for tuning in and I'll catch you in the next one. Sam.