Our collection of expert insights

For years, Elon Musk has emphasized the significance of Dojo, Tesla’s AI supercomputer, as the foundation of the company’s AI ambitions. Recently, Musk announced plans to intensify efforts on Dojo, especially with Tesla’s robotaxi reveal now set for October. Dojo, a custom-built supercomputer, is designed to train Tesla’s “Full Self-Driving” (FSD) neural networks, essential for

Yum! Brands is expanding its “Voice AI” technology to Taco Bell drive-thrus, aiming to implement it at hundreds of locations by year-end. The technology, tested at over 100 US Taco Bell stores for about two years, is designed to ease staff workloads, enhance order accuracy, and reduce wait times, while driving profitability. Despite mixed results

Google has enhanced its Gemini feature on the Android lock screen, allowing users to ask general questions without unlocking their device. Previously, Gemini’s functionality was limited to controlling alarms, timers, media, and some phone features via Google Assistant. Users had to unlock their phones to ask simple questions like “what’s the weather.” Now, Gemini provides

A stylized use of sparse autoencoders reveals how models, like those recognizing Paris as the City of Light, process information. While interpretability researchers initially hoped for a direct mapping between activations and individual neurons, practical challenges arose due to neurons activating for multiple, unrelated features. Sparse autoencoders address this by isolating a small number of

In the past 18 months, Leonardo.Ai has revolutionized generative AI with its advanced platform, enabling users from hobbyists to professionals to create high-quality, customizable AI-generated images. A highlight was the launch of our foundational model, Phoenix, which set new standards for prompt accuracy and text-in-image integration. Now, joining forces with Canva, we aim to accelerate

YouTube Music has officially introduced its “sound search” feature, allowing users to identify songs by humming, singing, or playing a tune. This feature, accessible via the magnifying glass icon in the app, provides detailed song information and is now available on both Android and iOS platforms. Additionally, YouTube Music is testing an “AI-generated conversational radio”

AI Studio, powered by Llama 3.1, is now available in the US, enabling users to create, share, and discover AI characters. Accessible via ai.meta.com/ai-studio or Instagram, the platform offers customizable AI creation for diverse applications, from personalized cooking guides to meme generation. Creators can build AI extensions to engage fans and streamline interactions on Instagram,

Vimeo (NASDAQ: VMEO) has unveiled an AI-powered video translation solution, revolutionizing global communication for businesses. Utilizing generative AI, the platform translates video, audio, and captions into over 29 languages, preserving the original speakers’ voices and significantly cutting costs and time compared to traditional methods. This innovation facilitates easier video localization for global audiences, employee onboarding,