MIT Technology Review got to try out Astra in a closed-door live demo last week. It was a stunning experience, but there's a gulf between polished promo and live demo. Astra uses Gemini 2.0's built-in agent framework to answer questions and carry out tasks via text, speech, image, and video, calling up existing Google apps like Search, Maps, and Lens when it needs to. "It's merging together some of the most powerful information retrieval systems of our time," says Bibo Xu, product manager for Astra. Gemini 2.0 and Astra are joined by Mariner, a new agent built on top of Gemini that can browse the web for you; Jules, a new Gemini-powered coding assistant; and Gemini for Games, an experimental assistant that you can chat to and ask for tips as you play video games.

Technology company Kognitiv Corporation ("Kognitiv") delivers advanced loyalty management, data activation and partner collaboration solutions to enable data-driven personalization and optimization of the customer journey across its innovative, omnichannel SaaS platform. Next week, Kognitiv is excited to officially launch a new AI-driven predictive customer insights and activation tool, Kognitiv Pulse, a solution that monitors the health of your customer base, predicts future behaviour and identifies risks and opportunities while enabling you to act quickly on these insights. Recommended AI: The Future of AI Is Here. Now Let's Make It Ethical Anchored on Kognitiv's proprietary SmartJourney customer lifecycle methodology, Kognitiv Pulse is powered by proprietary artificial intelligence and machine learning software. It enables marketers to drive efficiencies by identifying – at an individual level – which customers to engage proactively and helps them focus on activities that grow their bottom line.

Meta AI has built and unveiled BlenderBot 3, a 175 billion-parameter chatbot that it has made publicly available, complete with model weights, code, datasets, and model cards. "BlenderBot 3 delivers superior performance because it's built from Meta AI's publicly available OPT-175B language model -- approximately 58 times the size of BlenderBot 2," said Meta in an announcement on Friday. Unlike its predecessor, BlenderBot 3 can search the Internet to chat about almost any topic. Moreover, it can learn and improve its skills and safety through natural conversations and feedback from people in the real world. In contrast, most datasets are typically collected through research studies that "can't reflect the diversity of the real world", claims Meta.

Last year, OpenAI announced Codex, a model for efficient programming with the aid of Artificial Intelligence (AI). One of the videos uploaded to the OpenAI YouTube channel showed a live demo that was hard to believe even when seen with one's own eyes. With just a few lines of commands, it was possible to create a whole game in JavaScript. The level of the commands seemed somewhat high, but with Codex you can see that it is immediately able to implement the code and run the game. In this way, Codex is a model that helps people write code much more efficiently than they could on their own.

If you haven't read the first series, please read that and come back here. Let's continue from where we left of last time. I'm really impressed by the speed at which the transcription is generated to match the speed of the game. There can be a lot more done to make sure it's even more faster or we could also make it real-time or at least near real-time. With that said below is the Live Demo that you can watch to see how the real-time transcription is working and at what speed we get the transcripts from the Deepgram servers.

GNY's Machine Learning engine went head-to-head with the U.S. Energy Information Administration and outperformed it in predicting energy demand for California. The GNY team are building a larger vision for how GNY can support the work of scientists and advocates fighting for a sustainable and green planet long-term.

Thanks to the SAP devX external conferences program I had the opportunity to attend virtually the BlackHat USA conference. Together with DefCon, BlackHat is one the most advanced and exciting conference on offensive and defensive security that a security expert should attend. As explained in my previous article about BlackHat Europe 2019 the conference is proposing three main tracks: Briefings (advanced hacking and research presentations), Arsenal (live demo of open source security tools) and Sponsored sessions (presentation made by security companies). I made my selection among these different tracks according to my interest topics (Social Engineering, Machine Learning Security, Threat Intelligence, Code scanning, vulnerability management, ..) and here is a non exhaustive list from my personal selection of interesting topics: MLaaS (Machine Learning as a Service) platforms, are cloud based commercial services proposing pre-trained machine learning models and prediction functionalities deployed in a powerful cloud computing infrastructure and accessible to any user through an API. The goal is to benefit from a descent computing power hosted by a cloud provider but also take benefit from pre-trained models to make directly some predictions on the data.

Last week, we had Machine Learning 101: Hype or Reality. In this meetup we will do a deep dive into the Machine Learning features of Watson Studio - IBM's Data Science Platform. We will demonstrate in form of a live demo how data can be accessed & explored and how a Machine Learning Model can be build & deployed. For building and deployment we will be using several analytical tools: Jupyter notebook, SPSS Flows, Auto AI, Watson Machine Learning. By using Webex you are free to register using an alias, to mute yourself and to switch on/off the camera option at your disposal.

Since introducing its Einstein AI platform a few years ago, Salesforce has built AI into more and more of its tools. At this year's Dreamforce conference, for instance, the CRM giant announced new tools for customizing voice assistants and for incorporating AI into contact centers. The new capabilities showcase how Salesforce is progressively making work easier for its customers -- albeit in incremental steps. To give Dreamforce attendees a more forward-looking glimpse into its product capabilities, the Salesforce Research team demonstrated some of its breakthroughs in areas like conversational AI and natural language generation. Their research is focused on building an AI-driven world so far only found in sci-fi, said Salesforce Chief Scientist Dr. Richard Socher.

Computer vision at the intelligent edge is real and it is here! Access sample code to get started and watch a live demo on the Channel 9 IoT Show for inspiration. Thanks to a collaboration between NVIDIA and Microsoft, the NVIDIA Metropolis video analytics application framework, which runs on EGX, is now optimized to work with Microsoft Azure IoT Edge. The NVIDIA Metropolis framework includes the NVIDIA DeepStream software developer kit. With Azure IoT Edge and NVIDIA DeepStream, you can take a small, inexpensive NVIDIA Jetson Nano Developer Kit and analyze HD video streams in real-time.