Google Maps is adding AI features for travel planning. Databricks acquired Lilac and released dbrx, an AI model that outperforms Llama 2, Grok, and Mixtral on benchmarks.
Google Maps is rolling out some awesome updates this summer with a special focus on artificial intelligence to make your travel planning smarter and more personalized. The AI in Google Maps now helps you quickly understand what makes a place special by analyzing photos and reviews from the Maps community. For example, if you're looking at a restaurant, AI can identify the dishes being shown and provide you with key information like the name of the dish and its price, popularity and even if it's vegetarian or vegan. This is super helpful when you're in a new city trying out local cuisines. Besides this, Google Maps is introducing features to help you find and organize recommendations more easily. In select cities in the US and Canada, you can access curated lists of places to visit, like the hottest new restaurants or hidden gems sourced from trusted sites and the Google Maps community. You can save these lists for later or share them with friends. To tailor your travel experience even further, Google Maps allows you to create your own lists of places you want to visit or have already visited. The new updates give you tools to customize these lists by changing the order of places or linking to your social media reviews, making your recommendations more personal and useful. The combination of AI powered insights with the enhanced features for discovering expert curated lists and personalizing your own ensures that Google Maps is more useful than ever. And with a fresh new design including a cleaner home screen and new pin colors, navigating maps and planning your summer adventures is now easier and more enjoyable than ever. Alright now Databricks, a big company that works with data, obviously just bought another company called Lilac which is located in Boston and does research to help people understand and work with data better. They didn't tell us how much they paid for it, but Lilac is going to join Databricks and bring their technology and team to Databricks platform which was once known as a data lakehouse. This means that people using Databricks can now have an easier time making their data better for creating apps that use big language models. Why is this interesting? Because Databricks is really pushing to be the go to place for everything related to generative AI. They also recently put money into another AI company, Mistral, which is doing pretty good in the generative AI field and Lilac is going to help make dealing with data a lot easier. Now Databricks has created a new AI tool called dbrx and it's making waves because it's doing better than other similar tools like llama2grok and Mixtral. Wired magazine reports that this new tool from Databricks isn't just outperforming others from big names like Xai and Mistral. But it's also really close to matching OpenAI's GPT4 in understanding language, solving math problems and coding. Databricks also made sure to check how safe their new AI is by having people evaluate it. This is important because there's a growing worry about the risks of sharing powerful AI tools with everyone. Now. Databricks bought another AI company, Mosaic AI before, and shifted towards focusing more on AI. They want to use the data they have to help people make AI applications safely. They have also shared some AI models for people to use for different business needs. However, having good data is so super important for these AI efforts. Teams need to make sure their data is top notch for training and testing their AI models. This includes making sure the models aren't biased or making mistakes. This is where Lilac comes in. It helps teams work with unstructured data, which is data that's not neatly organized, and make it better and ready for AI use. Lilac was started by two people who used to work at Google, Daniel Smilkoff and Nikhil Thorat. In 2023, they made a tool that's easy to use and powered by AI to help understand and change unstructured text data. With Lilac data, scientists and AI researchers can organize their data, search through it, find and remove personal info or duplicates, and get it ready for AI models. The people at Databricks wrote in a blog post that Lilux tech will be added to their Mosaic AI tools. This means developers will have an easier time making their data ready for custom AI systems. They haven't shared all the details on how this will work, but the goal is the same to make it simpler for teams to check and prepare their data for AI use. Lilac said that joining databricks means businesses will have more control and insight into their unstructured data. This will help create better AI products for users. They think this move will let more developers use generative AI easily. By buying Lilac, Databricks is moving closer to offering everything needed to create great AI apps with their own data. This includes using AI models from other companies and tools tools to play with these models, use them in apps or tweak them with their own data for specific needs. Snowflake, another big name in data, is also trying to do something similar by offering services to help customers build apps with powerful AI models. 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.
