Hub AI in practice

Artificial Intelligence is present in everyday life – from “googling” to facial recognition to vacuum cleaner robots. AI tools are becoming more and more elaborate and support people and companies more effectively in their tasks, such as generating graphics, texting or coding, or interpreting large amounts of data.

What AI tools are there, how do they work, how do they help in our everyday world – and how do they change our lives? These are the questions we address in our Content Hub Artificial Intelligence in Practice.

AI in practice

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

AI in practice

Here's every Apple Intelligence update Apple announced at WWDC 25

AI in practice

Anthropic releases Claude 4 with new safety measures targeting CBRN misuse

Ad
AI in practice
Short

Apple has hit pause on its Apple Intelligence notification summaries for news apps after users reported accuracy issues. The problem came to light when the BBC flagged a serious error: the feature had created a false headline about Luigi Mangione, a suspect in the United HealthGroup CEO murder case, incorrectly claiming he had taken his own life. Apple's latest iOS 18.3 beta now clearly labels the feature as being in beta testing. Users can now turn off these automated summaries right from their lock screen, where the summaries now appear in italics. The company has added a warning that these summaries might contain mistakes, and has completely disabled them for all news and entertainment apps for the time being.

Ad
Ad
AI in practice
Short

Nvidia is adding three new safety features to its NeMo Guardrails platform, aiming to give companies more control over their AI chatbots. The company says these microservices address common challenges in AI safety and content moderation. According to Nvidia, the Content Safety service checks AI responses for potentially harmful content before they reach users, while the Topic Control service tries to keep conversations within approved subject areas. A third service, Jailbreak Detection, works to spot and block attempts to bypass the AI's security features. Rather than using large language models, Nvidia says these services run on smaller, specialized models that should need less computing power. A few companies, including Amdocs, Cerence AI, and Lowe's, are currently testing the technology in their systems. The microservices are available to developers as part of Nvidia's open-source NeMo Guardrails package.

AI in practice
Short

Luma has released Ray2, its latest AI model for video generation. The new model has been scaled to 10 times the compute power of its predecessor and uses a new multimodal architecture. Luma says it's better at natural-looking movement and fine details, while maintaining consistency throughout each scene. Text-to-video generation is available to subscribers through Luma's Dream Machine platform, with image-to-video, video-to-video, and editing tools in upcoming releases. The company also plans to make Ray2 accessible through its API soon.

Video: Luma AI

Ad
Ad
AI in practice
Short

Synthesia, a company that helps businesses create AI-generated videos and avatars, just secured $180 million in Series D funding. Since its founding in 2017, the platform has grown to serve more than half of Fortune 100 companies, producing millions of minutes of AI-generated video content each month. The company plans to use its fresh funding to develop new features that combine AI avatars with large language models, along with rolling out a new video player. The company says these upgrades will allow customers to create more interactive and personalized video content.

Ad
Ad
AI in practice
Short

A new open source voice model called Kokoro just landed on HuggingFace, and early tests show it can generate voices that rival commercial services like Eleven Labs. The model packs 82 million parameters under the hood, and is on the first place in the TTS Spaces Arena. The model is trained on less than 100 hours of audio data, supporting just American and British English for now. Users can currently choose from 10 different voices. While the model shows promise, it does have its limitations. Unlike some commercial alternatives, it can't clone voices, and there aren't any plans to add support for other languages yet. For developers interested in using Kokoro, the inference code is available under an MIT license, while the model itself uses an Apache 2.0 license.

Google News

Bank details

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

Meta's latest model highlights the challenge AI faces in long-term planning and causal reasoning

Google News