Author HubMaximilian Schreiner

Max is the managing editor of THE DECODER, bringing his background in philosophy to explore questions of consciousness and whether machines truly think or just pretend to.
AI research

New Othello experiment supports the world model hypothesis for large language models

AI in practice

ChatGPT might be draining your brain, MIT warns - what ‘cognitive debt’ means for you

AI in practice

Anthropic releases Claude 4 with new safety measures targeting CBRN misuse

Ad
AI in practice
Short

ChatGPT has expanded its video and screen sharing features to users in the EU, Switzerland, Iceland, Norway, and Liechtenstein. The update, which first launched in the US last December, brings real-time visual interaction to the platform. Plus and Pro users with Advanced Voice Mode can now show ChatGPT their surroundings through video or share their screens, eliminating the need to upload photos or type out detailed descriptions. To use these features, users can tap the voice icon in the mobile app and select either the video icon for calls or the three-dot menu for screen sharing. Additionally, ChatGPT now offers more personalization options, letting users adjust the AI's tone, writing style, and response length.

AI in practice
Short

Nvidia has rolled out DLSS 4 alongside updates to its Broadcast app, timing the release with its new RTX 50 series graphics cards. The latest version of the AI upscaling technology uses a new transformer model to enhance image quality. Players can now update to the latest DLSS version in 75 supported games using DLSS 4 Overrides. The update lets users take advantage of the most recent improvements without waiting for individual game patches. The company has also updated Nvidia Broadcast, an app that taps into RTX graphics cards' AI capabilities to clean up audio and video in real time during video calls and livestreams. The app now includes a studio recording mode that mimics professional recording environments, plus video noise reduction and a virtual key light feature that uses AI to brighten faces and reduce shadows. Rounding out the updates, Nvidia has refined its RTX Video Super Resolution tool, which uses AI to improve the quality of online videos.

AI in practice
Short

OpenAI has rolled out an updated version of GPT-4o in ChatGPT, extending its knowledge base through June 2024. The refresh aims to provide more current and contextual responses across topics. The model's visual capabilities have also gotten an upgrade. It can now better understand spatial relationships in images and make stronger connections between diagrams, graphics, and text. This enhancement should lead to better results when handling visual tasks. Testing shows the updated model performs better on several fronts, including math, science, programming challenges, and general language understanding, according to OpenAI. Users might also notice it's more likely to mirror their emoji usage during conversations. OpenAI has also added a "Use more intelligence" button to the chat interface, prompting the model to take extra time processing before responding. While these changes are available in ChatGPT, the updated model hasn't yet appeared in the API documentation.

Ad
Ad
Ad
Ad
AI in practice
Short

A new open source AI music tool called YuE lets anyone turn lyrics into songs, offering a free alternative to commercial services like Suno and Udio. According to developer Ruibin Yuan, the system can create songs up to five minutes long in multiple languages and musical styles. Under the hood, YuE uses two separate AI models - one handles speech and music generation, while the other manages additional song elements. Users can customize their creations by choosing the musical style, mood, and type of voice, and the system can replicate various singing techniques. The project is available now on HuggingFace, with several demo examples posted on GitHub. Yuan says a research paper detailing the technical aspects of YuE is currently in development.

Ad
Ad
Google News

Bank details

IBAN: DE87 1203 0000 1086 0070 75
Account holder: DEEP CONTENT GbR
Purpose: Support THE DECODER
AI research

"Cat attack" on reasoning model shows how important context engineering is

AI research

Apple's claims about large reasoning models face fresh scrutiny from a new study

AI in practice

Cloudflare CEO Matthew Prince sees trouble ahead for the open web

Google News