😳Blueprint for Military AI?!

Seoul hosts international summit to develop blueprint for responsible military AI use

Key Points 

• The proposed framework aims to align with international law and prevent autonomous weapons from making life-and-death decisions without human oversight.

• Although the summit seeks to develop a framework, any agreement reached is unlikely to be legally binding. 

☕News - South Korea held an international summit on Monday aimed at creating guidelines for the responsible use of artificial intelligence (AI) in military applications.

Representatives from over 90 countries, including the United States and China, are attending the two-day event in Seoul, marking the second time such a gathering has taken place. At the first summit in Amsterdam last year, the United States, China, and other nations supported a modest "call to action."

🕵🏻‍♂️What's going to happen? South Korean Foreign Minister Cho Tae-yul stated that the discussions would focus on ensuring compliance with international law through legal reviews and developing mechanisms to prevent autonomous weapons from making life-and-death decisions without proper human oversight.

The Seoul summit aims to establish a basic framework for the responsible use of AI in the military by setting minimum safety standards. The blueprint is expected to incorporate principles from NATO, the U.S., and other countries. 

It's unclear how many nations will endorse the document on Tuesday. It's also important to note that while the document seeks to develop a framework, any agreement reached is unlikely to be legally binding. 

Microsoft trusts diversity to address AI bias

Key Points 

• Microsoft views increased diversity in AI development teams as essential for reducing bias in technology.

• While diverse teams alone can't fully resolve bias in AI due to inherent issues in training data, integrating D&I principles into AI models from the start can help the technology better support these values.

📜Context of the news - At the start of 2023, Microsoft faced a PR crisis while trying to showcase its advancements in AI after investing heavily in OpenAI. The company integrated an AI-powered chatbot into its Bing search engine, becoming one of the first major tech firms to incorporate AI into its core products.

However, the rollout quickly hit trouble as users began sharing interactions with Bing that included troubling remarks and racial slurs, drawing global attention. In response, Microsoft promptly restricted the AI’s functionality and later replaced the Bing chatbot with Copilot. 

Despite these setbacks, Microsoft remains committed to using AI to advance equity and representation with appropriate safeguards.

💡News -  Microsoft believes that increasing diversity and inclusion within the teams developing AI technology is key to addressing bias in the systems.

Lindsay-Rae McIntyre, Microsoft's chief diversity officer, asserts that for AI systems, including large language models like ChatGPT, to achieve empathy, relevance, and accuracy, they must be developed by a diverse group of developers, engineers, and researchers.

🤔Can diversity really fix AI bias, though? While diversifying the teams behind AI development is a step forward, it doesn't guarantee a complete solution. Large language models like Copilot, ChatGPT, and Gemini are trained on extensive internet data, which can contain inherent biases that are hard to eliminate. Still, Microsoft believes that integrating diversity and inclusion principles into AI models from the beginning can help the technology support these values effectively.