1. Gemini Gets a New Scrollable Feed Interface
Google is currently experimenting with a significant user interface overhaul for its AI assistant, Gemini. This new update introduces a dynamic, scrollable feed to replace the conventional chat bubble layout, offering users a more intuitive and visually engaging experience. This change aims to simplify content consumption and interaction, making Gemini more accessible and user-friendly for a wider audience.
Read full story
2. Sora Gets New Copyright Controls and Revenue Sharing
Sora, the text-to-video AI model, has implemented new copyright controls. Moving forward, content owners will need to explicitly opt-in for their intellectual property (IP) to be utilised in the generation of AI-powered videos. This significant update also introduces revenue-sharing options, allowing content creators to potentially benefit financially when their licensed works are used by Sora. This move aims to address growing concerns around AI ethics and creator rights, offering more control and compensation to artists and copyright holders.
Read full story
3. OpenAI Eyes Wallet: Acquires Roi’s CEO for Finance Integration
OpenAI has acquired the CEO of Roi, a strategic move aimed at integrating deep financial expertise into its operations. This acquisition is crucial for OpenAI as it plans to develop more sophisticated and customised consumer applications. The company’s adaptive product strategy will benefit significantly from this new leadership, enabling the creation of tailored financial tools and services within its AI ecosystem. This step highlights OpenAI’s ambition to venture into personal finance and expand its influence beyond traditional AI applications.
Read full story
4. AI Engineering Crash Course: Build Applications Effectively
This comprehensive crash course offers essential knowledge for anyone looking to build robust AI applications. It covers critical topics such as selecting the right AI models, evaluating their performance accurately, fine-tuning them for optimal results, and mastering dataset engineering techniques. This course is ideal for developers and engineers aiming to enhance their skills in practical AI application development.
Read full story
5. NVIDIA Experts Share Key LLM Security Tips
NVIDIA’s AI security experts have provided crucial advice for securing Large Language Models (LLMs). Based on vulnerabilities they discovered, their recommendations include implementing sandboxing techniques and strict access controls. These measures are essential to protect LLMs from potential exploits and ensure their secure operation in various applications, enhancing overall system integrity.
Read full story
6. Doctor-Approved ChatGPT Tips for Health Communication
Doctors suggest using ChatGPT to understand complex medical terminology and enhance communication between patients and healthcare providers. However, experts strongly advise against using AI tools like ChatGPT for self-diagnosis or making treatment decisions. It should be seen as an assistive tool for information and clarity, not a diagnostic one.
Read full story
7. AI Voices Fool 58% of People, Study Reveals
A recent study highlights a concerning trend: over half of the population, specifically 58%, cannot distinguish between voices generated by Artificial Intelligence and those of real humans. This research underscores significant concerns regarding the widespread adoption of AI voice technology and the potential misuse through deepfake audio. Experts are now calling for greater public awareness and the development of robust detection mechanisms to combat the risks associated with increasingly sophisticated AI-generated content.
Read full story
8. AI Bypasses DNA Security, Creates Undetectable Toxic Sequences
New research indicates that Artificial Intelligence (AI) can now circumvent existing biosecurity measures by rewriting toxic DNA sequences. This AI capability allows it to create genetic variants that deliberately avoid detection during standard screenings. This raises significant concerns regarding biosecurity protocols and the potential for malicious use of such technology in developing undetectable harmful biological agents. Experts are now evaluating how to fortify defences against this evolving threat.
Read full story
