AI SHOCKS AGAIN: GitHub's New AI Tools, , Amazon's AI Secret Project Theia

GitHub's Copilot now auto-fixes code vulnerabilities. Google integrated health advice into Fitbit. Amazon's Project Theia lets you shop by pointing at things in the real world. Microsoft formed a new AI division led by Mustafa Suleyman and Karen Simonyan.
Chapters
So this video is packed with this week's exciting AI news, from Google's new health focused AI and Fitbit wearables to GitHub's tool for fixing code vulnerabilities and the unveiling of Stability AI's 3D animation tool. We'll also explore the discovery of new galaxies by Citizen Scientists With AI's help, the emergence of new AI focused job roles Due to the EU Artificial Intelligence Act, Amazon's Project Thea for enhanced online shopping, YouTube's new feature for AI generated content, Ubisoft's AI powered NPCs, a groundbreaking AI system for predicting breast cancer and Microsoft's new AI division led by Mustafa Suleiman and Karen Simonyan. So Google is stepping up its game in the health sector by integrating AI with its Fitbit wearables. The tech titan is working on a health focused AI aptly named Personal Health Large Language Model, which aims to offer personalized health and fitness advice directly from your wrist. This innovation is a result of collaboration between fitbit, now under Google's umbrella, and Google Research. By leveraging data from Fitbit and Pixel devices, the AI promises to deliver tailored suggestions for improving your health routines. Imagine getting customized tips on adjusting your workout intensity based on your sleep quality, all thanks to the sophisticated analysis by this new AI coach. While it's still in the development phase, Google teases that the AI's capabilities are promising, showcasing proficiency in sleep medicine practice tests. Alongside this Fitbit project, Google is not shy about its ambition in healthcare AI, rolling out other AI tools aimed at supporting medical professionals with diagnostics and patient care. Alright now. GitHub has rolled out a cutting edge AI tool designed to automatically detect and fix security flaws in code. This innovation is now in beta. Harnesses the power of GitHub's existing tools Copilot and CodeQL to scan code in real time and patch up vulnerabilities on the spot. Remarkably, it's capable of addressing over two thirds of detected issues without needing direct input from developers. Covering a wide range of problem types across popular programming languages like JavaScript, TypeScript, Java and Python. Available to GitHub Advanced Security customers, this tool not only aims to ease the burden on developers, sparing them from the drudgery of manual bug fixes, but also allows security teams to shift their focus toward more strategic defense measures. This smart Solution draws on GitHub's semantic analysis engine CodeQL for its core functionality, combined with heuristics and the GitHub Copilot APIs, powered by OpenAI's GPT4 to propose fixes. Although it promises a high accuracy rate, GitHub concedes that a small fraction of the suggestions might not fully grasp the code's context or the nature of the vulnerability, highlighting the balance between AI assistance and human oversight in software development okay, next stability AI has launched a new AI tool called Stable Video Diffusion 3D, transforming simple images into 3D animations. This advanced technology, built on the widely used Stable Video Diffusion model, now allows for creating animated views of objects or scenes from just a single image. It introduces a more dynamic perspective by generating 3D meshes, offering a realistic consistency within videos. There are two versions, SV3DU for orbiting videos from a static image, and SV3DP, which adds camera path movement for more detailed animations. This innovation opens up possibilities for more immersive virtual experiences and product displays, and making the futuristic concept of a holodeck more tangible. The training for this model utilized a select portion of the Objiverse dataset, emphasizing transparency in its development process alright now Citizen scientists teamed up with AI to uncover 430,000 new galaxies, including 30,000 of the rarest kind ring galaxies. This achievement comes from the Galaxy Cruise project, supported by 10,000 volunteers using the Subaru Telescope, an advanced optical infrared telescope in Hawaii. Together, they analyzed vast amounts of space data that were too much for astronomers to handle alone. AI played a crucial role, efficiently classifying galaxies in less than an hour, a task impossible without the two years of data collection by citizen astronomers. This collaboration not only showcases the power of community and technology in cosmic exploration, but also emphasizes the diversity of galaxies, from the common spiral to the exceptionally rare ring galaxies, revealing the intricate tapestry of our universe. Okay, next. So as AI technology advances, especially with the European Union setting the stage with the EU Artificial Intelligence Act, a variety of AI focused roles are emerging from artificial intelligence engineers who develop and program AI algorithms to prompt engineers who refine communication between humans and AI. These positions highlight the evolving relationship between technology and the workforce. Cybersecurity analysts are now more in demand, with a particular emphasis on those familiar with AI to protect against data breaches. Additionally, AI technology specialists or consultants are being sought after to navigate the legal and ethical frameworks of AI use, ensuring compliance with the EU AI Act. Furthermore, the role of an AI ethics officer is gaining prominence to address the ethical considerations in AI application, ensuring fair and unbiased use of technology in hiring processes and beyond. Right now there's a big need for AI experts globally, with Europe standing out. So if you have expertise in AI, it's now the best time to look for a job in this field. Okay, now there's buzz that Amazon is under wraps developing a new AI tool project Thea, designed to make shopping online even easier. Imagine you're looking for something specific to buy on Amazon. With this new tool, you can either upload a picture of the product you're thinking of, or see it placed in different scenes created by the AI. This way, you can get a better idea of how the product looks in various settings, helping you decide whether it's exactly what you need. Business Insider shared this news, mentioning that the tool is all about helping shoppers find and buy things more quickly by giving them a visual taste of the products in action. Okay, next YouTube has introduced a new feature that lets video creators tell viewers if their content has AI generated or fake elements. When uploading, a video creators will see a checkbox that to mark if their video includes stuff like fake scenes, altered real footage, or deep fake voices. This is for videos that look real but might show something that didn't actually happen, like a made up disaster or a person saying things they never said. However, YouTube won't ask for this label on videos with obvious special effects or animations. This move follows YouTube's announcement from last November about its plan to handle AI generated content, dividing the rules into stricter ones for music and and more lenient ones for other content. YouTube also mentioned it's working on ways to catch AI fakes more effectively, but for now they're relying on video creators to be honest about their content. Alright now. Sean Hollister, a seasoned editor from the verge, delves into Ubisoft's latest venture with AI powered video game NPCs, highlighting the collaboration with InWorld and Nvidia. For three months, curiosity loomed over which developer would innovate with this tech, and Ubisoft, known for embracing new technologies, stepped up. Hollister experienced Ubisoft's experimental demo featuring Neo NPCs, engaging with them in scenarios that test emotional and strategic responses. Despite recognizing these NPCs as prototypes and acknowledging their limitations, such as scripted responses and latency issues, he sees potential in this technology. Ubisoft aims to advance in real time, emotion, contextual awareness, and collaborative decision making. Though Hollister notes the experience didn't significantly surpass previous encounters with Nvidia's AI NPCs. Nonetheless, Ubisoft's attempt to blend narrative with interactive AI dialogue offers a glimpse into the evolving relationship between players and game characters, suggesting a future where game worlds react more intelligently to player actions. Okay, next There is a new AI system developed by researchers at Duke University that can predict breast cancer up to five years in advance by analyzing differences between a person's left and right breasts. It is called Asim Mirai and it stands out for its transparency, allowing doctors to see how it calculates risk, which could lead to early detection, fewer unnecessary tests and reduced healthcare costs. Unlike previous AI models that worked like a black box with unclear decision making processes, this model's workings are open and understandable. This innovation marks a significant leap in the use of AI in healthcare, offering a more accessible and trustworthy tool for medical professionals. By focusing on breast tissue differences, Asimiri provides a simpler yet effective approach to predicting breast cancer risk, potentially revolutionizing how doctors screen for and manage this disease. Okay, finally, in a significant move to strengthen their AI capabilities, Microsoft has welcomed Mustafa Suleiman and Karen Simonian to spearhead a new division, Microsoft AI, with a focus on enhancing Copilot and other consumer AI initiatives. Suleyman, co founder of DeepMind and Inflection, has been appointed EVP and CEO of Microsoft AI and will be part of the senior leadership team, directly reporting to Nadella. Simonian, also a co founder of Inflection and a leading AI researcher, will join as chief scientist reporting to Suleiman. Strategic Edition aims to drive forward Microsoft's AI research and product development, with several talented individuals from Inflection joining the team. The update reiterates Microsoft's commitment to AI innovation, leveraging its partnership with OpenAI and continuing to build a comprehensive AI ecosystem. Alright, don't forget to hit that subscribe button for more updates. Thanks for tuning in and we'll catch you in the next one.










