AI has yet again taken center stage at an event organized by Tech giant Google in May. Sundar Pichai and his team showcased the best of what Google has been developing since the last I/O meeting in 2022. I am most excited about the partnership between Google and Apple to create a security feature that is welcome news for Android and iOS users. I'll reveal what Google and Apple created in the course of the blog, so keep a sharp eye.
Google opened the event by showcasing added functionality to two of their most trusted software, Google Photos and Google Maps. The new 'Immersive view for routes' for Google Maps shows the driver the entire journey even before the engine starts. Google Maps will now show your location and destination and make a short video demonstration of the exact routes to take to get to your destination.
With the update to 'magic eraser,' you can now make any subject in an image take center stage. Snaps stuck in a corner can be centered while Google's AI works in the background to complete the rest of the image. Reading doesn't do it the justice it deserves. Check it in the video below.
PaLM 2 is Google's next-generation language model developed to excel at mathematical computations, writing and debugging code, translating between various languages, and advanced reasoning. According to Google, PaLM 2 was trained in over 100 spoken languages and 20 programming languages. It is the force behind Bard, Google's response to ChatGPT. Bard, powered by PaLM 2 has many functions, but one that will stand out to developers is its ability to convert code from one programming language to another. WOW!!
Google is focusing on applying PaLM 2 in health and cybersecurity with the models Med-PaLM and Sec-PaLM respectively. Med-PaLM 2 recently scored over 85% in a medical exam which means that currently, it is on par with some of the best medical minds around today. The best part of this is that PaLM 2 will only get better. The future is exciting.
We're not done with AI yet. Google Workspace has now received a Generative AI upgrade. Across Gmail, Sheet, and Docs, you can now generate an entire PowerPoint presentation based on a single prompt and it's crazy how good and accurate these are. We are in for great times ahead. Generative AI is getting added to search and Google Cloud as well. Google did this by adding a side panel to Workspace where you can run prompts and get real-time suggestions on whatever project you're working on. They call this 'sidekick'.
Generative AI integrates now with Google Search. It provides the most important answer to any given query and then provides a suggested question that the user may ask as a follow-up. It also gives suggestions on the next question the user might be interested in asking which brings the whole interaction into a conversational mode. This helps to limit misinformation on the web and reduce spam articles from malicious websites.
AI is being optimized for businesses too. With the introduction of Vertex AI, it is now easier to build Generative AI applications for businesses because Google has partnered with various companies to test out building applications custom-made for that company's everyday use. Vertex AI according to Google, helps to reduce workload in training personnel by about 80%. This is a piece of welcome news for startups who want to optimize and scale quickly but cannot afford the personnel to do so.
I don't know what project tailwind is supposed to be but I suppose I'll keep an eye on it to see what it becomes.
Responsible AI is one of the most outstanding news the world has received from the AI community because AI has been going nuts recently, especially with AI-generated images causing a stir now and then on Twitter because it is now more difficult to distinguish between what is real and what is not. Take the viral image of the Pope in sweaters for example. Someone needs to set an example and Google is taking up that mandate and I applaud them for that. To start with Google is adding two new ways to evaluate any image you come across in their search engine. First, they will bring the 'about image' feature which tells the user where and when an image that has little or no context first appeared. Secondly, images artificially generated with Google's AI will have metadata showing when and when it was created even on external search engines.
Find my device has so far worked for only Android smartphones but not anymore. Google will include smartwatches, headphones, and all other accessories to make sure that users do not lose their devices easily.
Now my favorite of all these is 'Unknown tracker alerts'. Remember I told you that Google and Apple worked together? Yes, stalkers for some time used air tags to track and stalk their victims until Apple released a patch to counter the issue by sounds when an unrecognized air tag follows you around. This did not protect those on Androids however and so to develop software to counter air tag stalking on Android, Google, and Apple developers came together to create this small but mighty product and also show us that security is more important than rivalry.
Other announcements were devices such as the Google Fold and Pixel 7a. More on those later