Hello everybody. Welcome to Daily Zaps.

Here’s what we got for ya today:

• 🏆 Global prompt engineering championship

• 🙊 Scarlett Johansson vs. OpenAI

• 🤖 NIH wants to reduce medical chatbot hallucinations

• 🇪🇺 EU signs off on artificial intelligence law

Let’s get right into it!  

Sign Up | Advertise | Tools

WORLD NEWS

Global prompt engineering championship

Dubai's Global Prompt Engineering Championship, focused on advancing AI technology skills, concluded with winners in three categories: coding, literature, and art. Indian winners Ajay Cyril and Aditya Nair, and Austrian Megan Fowkes, each earned the title of Best Prompt Engineer and a share of the Dh1 million ($272,290) prize. The event, organized by the Dubai Future Foundation and Dubai's Centre for AI, showcased global talent and aimed to inspire AI adoption.

Minister Omar Al Olama emphasized the importance of prompt engineering skills for future workforce and quality of life enhancements. The competition highlighted the broad applicability of AI in various fields, demonstrating the crucial role of human creativity in utilizing AI tools effectively. The diverse skill set and structured thinking of participants stood out, showcasing AI's positive potential and the importance of understanding its capabilities and limitations.

LEGAL

Scarlett Johansson vs. OpenAI

Last week, OpenAI showcased a new voice for ChatGPT called "Sky," which many viewers, including Scarlett Johansson, found reminiscent of Johansson's AI character in the film "Her." Johansson, who had previously declined to voice ChatGPT, hired legal counsel and demanded an explanation from OpenAI.

In response, OpenAI paused Sky's use and clarified that Sky's voice belonged to a different actress. Legal experts suggest Johansson might pursue a "right of publicity" claim, as imitating someone's distinctive voice for commercial purposes can be problematic. OpenAI’s CEO, Sam Altman, made a post referencing "Her," which further complicates their defense. This incident highlights the ongoing legal and ethical challenges in AI development.

HEALTHCARE

NIH investing millions to reduce medical chatbot hallucinations

ARPA-H, part of the National Institutes of Health, is investing millions in reducing medical chatbot/LLM hallucinations—instances where chatbots provide incorrect or misleading information. This issue is a significant barrier to using medical chatbots for reliable health information. The CARE initiative, a 24-month program with a budget of $8 million, aims to develop technologies that evaluate chatbot outputs with the accuracy of human experts.

The initiative focuses on two key areas: improving prompt generation to reflect diverse stakeholder values and developing scalable technologies to detect factual inaccuracies. By enhancing the reliability of medical chatbots, ARPA-H seeks to make health information more accessible and accurate, particularly for underserved communities. The initiative's ultimate goal is to improve digital health communication and ensure inclusive, effective healthcare.

WORLD NEWS

EU signs off on artificial intelligence law

Ministers have approved a groundbreaking EU law that sets rules for using artificial intelligence in sensitive areas. AI systems in fields like law enforcement and employment must be transparent, accurate, secure, and use high-quality data. High-risk AI systems will need certification before being used in the EU, and certain systems, like China's social credit scoring, are banned. Real-time facial recognition is generally banned but allowed in specific law enforcement cases.

Lower-risk AI systems must disclose that their content is AI-generated. An "AI Office" within the European Commission will enforce the law, which will come into effect 20 days after publication, with most rules applying in two years.

🗣️ Feedback