AI in practice | The Decoder

Matthias Bastian

Mar 16, 2025

Baidu claims its Ernie X1 reasoning model matches Deepseek-R1 performance at half the price

The Chinese search and Internet company Baidu presents the latest version of its chatbot Ernie, which is supposed to deliver significantly better results than before. It also gets support for plugins.

Matthias Bastian

Mar 15, 2025

Apple's Siri delays reveal broader challenges of scaling generative AI

Apple's Siri fail shows the tech giants' AI scaling dilemma

Jonathan Kemper

Mar 15, 2025

17 new companies join top 50 consumer AI apps in latest a16z rankings

An Andreessen Horowitz analysis shows that creative tools for image, video, and music creation dominate the generative AI market. 52% of the companies analyzed focus on content creation or editing. Text-based applications such as ChatGPT continue to lead the market. ChatGPT's competitors, such as Anthropic's Claude and Perplexity, have recently gained some ground and are showing faster growth. However, they still lag far behind in total traffic. Video generator Luma Dream Machine has quickly become a top AI site with 22.6 million visits. Music generator Suno has also seen significant growth.

Jonathan Kemper

Mar 14, 2025

Update

Sesame releases CSM-1B AI voice generator as open source

Audio waveform visualization with yellow and green amplitudes showing a dialogue between two parties "Human" and "Sesame".

Matthias Bastian

Mar 13, 2025

Google opens up its AI Deep Research feature to all users for free "a few times a month"

Matthias Bastian

Mar 13, 2025

Xbox's AI Copilot to provide real-time gameplay guidance through natural conversation

Xbox introduces AI-powered gaming co-pilot as a digital gaming advisor

Matthias Bastian

Mar 13, 2025

Google adds search history integration to personalize Gemini AI

Google is adding a personalized function to its AI assistant Gemini: with users' permission, Gemini analyses their Google search history to provide more relevant answers. The new function uses Google's experimental AI model Gemini 2.0 Flash Thinking. Users can activate personalization, deactivate it at any time and track which data has been used. Following the launch for Gemini subscribers on the web, Google plans to extend personalization to Google Photos and YouTube soon to make Gemini a comprehensive personal assistant.

Matthias Bastian

Mar 13, 2025

AI expert Andrej Karpathy envisions a web where 99.9% of content is optimized for AI, not humans

Maximilian Schreiner

Mar 13, 2025

Update

OpenAI launches in Europe and more countries

OpenAI introduces Operator, an AI agent that can operate web browsers autonomously. The service, which is initially only available to US ChatGPT Pro users, marks a further step towards autonomous AI assistants.

Matthias Bastian

Mar 12, 2025

Google Deepmind unveils new AI models for robotic control

Google Deepmind has introduced Gemini Robotics On-Device, a version of its robotics model that runs directly on robot hardware, removing the need for a cloud connection. By processing everything locally, robots can operate in environments with unreliable or no internet access.