In today’s Quick Serve of AI news:
Reddit’s journey to the stock market could be greatly influenced by its partnerships with AI companies such as OpenAI. The company has highlighted the potential benefits of its data licensing agreements, which allow AI firms to train their models using Reddit’s vast amount of data, in its IPO prospectus. Although specific AI partners haven’t been disclosed, the prospectus reveals contracts worth $203.0 million. OpenAI, with close ties to Reddit through CEO Sam Altman, is a potential partner. Reddit believes its data can play a crucial role in improving AI models, offering real-time insights into various topics. This shift in strategy towards monetizing its data reflects a broader trend of content creators seeking compensation from AI firms.
Intel’s CEO Pat Gelsinger is steering the company towards a major comeback, focusing on AI and revamping its manufacturing capabilities. Gelsinger, optimistic about overcoming the company’s past struggles, plans to outperform rivals like TSMC with Intel’s updated manufacturing process. This includes a significant focus on AI, highlighted by Intel’s new “AI-era system foundry” business model. Microsoft has already shown interest in this new direction. Gelsinger discusses the importance of adapting Intel’s strategy to the evolving AI landscape and the potential of the company’s new technologies, such as ribbonFET and backside power, to change the chip-making game.
Stability AI‘s release of Stable Diffusion 3 (SD3) aims to solidify its position in the AI imagery field against rivals from OpenAI and Google. SD3 promises improved performance and versatility, supporting a wide range of hardware. The model’s advancements include a new architecture and techniques like “flow matching,” potentially setting it apart from the competition. Despite the technical complexities, Stability AI focuses on making SD3 accessible and safe, with enhancements to prevent misuse. This release illustrates the competitive landscape of AI-generated imagery and the evolving capabilities of these models.
Google’s new AI tool Gemini is facing backlash for generating historically inaccurate images of people. In response, Google has paused Gemini’s ability to create images of individuals. Critics have pointed out that the tool produced diverse depictions of the U.S. Founding Fathers and Nazi-era German soldiers, sparking controversies and discussions about AI and historical representation. Google is working on improving the technology to address these issues. The incident sheds light on the challenges of using AI in generating historical content and the broader implications for technology’s role in shaping historical narratives.
Samsung is set to bring its Galaxy AI features to a wider range of devices with an upcoming update. Initially introduced for the Galaxy S24 series, these features enhance user experience with functionalities like Google’s “Circle to Search” and Live Translate for voice and text. Additionally, features such as “Chat Assist” and “Generative Edit” will improve messaging and photo editing. Samsung’s move to expand the availability of these AI features underscores the growing integration of AI into consumer devices, aiming to provide users with more versatile and efficient technology.