[AI Hanip News] If It's Long, It Will Open Up

Growing Concerns Over AI Side Effects... Busy Implementing Safety Measures

Editor's NoteSince the emergence of ChatGPT, the world has been buzzing with artificial intelligence (AI). Technology is changing daily, and news is pouring in nonstop. Even industry experts who are at the top of their game are struggling to keep up. It's overwhelming to consume everything, but starving is not an option either. We bring you the core of weekly AI issues, carefully selected and easy to digest.

ChatGPT Has a Hole... AI Detectors Shut Down

[Image source=Yonhap News]

ChatGPT has a vulnerability. Carnegie Mellon University and the AI Safety Center in the U.S. released a report showing that 'darkened' ChatGPT can be easily exposed using long prompts (commands). ChatGPT has a kind of guardrail installed to prevent biased remarks, sexual conversations, or harmful information. When prompted for such answers, it usually responds with 'I cannot answer that.' However, if a long suffix is appended after the question, it seems to lose focus and becomes ineffective. OpenAI's AI detector, which was designed to distinguish between human-written and AI-generated text, is also underperforming. The longer the text, the less accurate it becomes, eventually leading to its shutdown. Additionally, OpenAI's trust and safety officer, who handled privacy issues, has left the company. OpenAI also received its first fine (3.6 million KRW) in South Korea for personal information leakage, indicating a challenging situation.

ChatGPT Android App Released

OpenAI, embroiled in controversies over declining users and performance setbacks, has launched a ChatGPT app for Google Android. This comes two months after releasing the app on the Apple App Store in May. The Android app is currently available in the U.S., India, Bangladesh, and Brazil, and it is expected to be available soon in Korea as well. Meanwhile, Microsoft (MS) plans to expand 'Bing Chat,' which integrates ChatGPT into its Bing search engine, to Google Chrome and Apple Safari browsers. Until now, Bing Chat was only available on MS's Edge browser. This appears to be a strategy to revive the waning ChatGPT craze and increase market share in the search engine sector.

"Stop Fake News"... Watermarking AI Content

Seven leading AI companies, including Google, MS, Meta, and OpenAI, have voluntarily taken safety enhancement measures. This move comes as concerns over AI side effects such as fake news and information manipulation grow, with the U.S. government emphasizing corporate responsibility. Accordingly, these seven companies have agreed to embed watermarks in AI-generated content and conduct research on AI's social risks. Among them, four companies?Google, MS, OpenAI, and Anthropic?will also develop AI safety standards. They have formed the 'Frontier Model Forum' and plan to conduct research to ensure the safety and accountability of AI model development.

Amazon and Intel Also 'Going on the Offensive'

Amazon has launched a healthcare service for medical professionals using generative AI. The AI recognizes conversations between medical staff and patients and creates medical records in real time. This service is expected to save time for doctors and nurses. Meanwhile, Intel announced that it will integrate AI into all future products. It had already announced the release of the next-generation PC processor 'Meteor Lake,' specialized for AI functions, by the end of this year. Now, it plans to expand AI application beyond just some premium products. With MS, Google, and Meta recording 'earnings surprises' thanks to AI, Intel seems to be feeling the pressure.

Netflix Offers 1.2 Billion KRW Salary for AI Talent... Actors and Writers "Are Upset"

[Image source=Yonhap News]

Netflix has offered a high salary to recruit AI talent. They created a new position called 'Machine Learning Product Manager' and posted a job listing with a salary range up to 1.16 billion KRW. This has further angered Hollywood actors and writers who are currently on strike opposing the entry into AI content business. They are already worried about losing their jobs to AI. Currently, the two major Hollywood unions?the Writers Guild of America (WGA) and the Screen Actors Guild?American Federation of Television and Radio Artists (SAG-AFTRA)?are conducting a joint strike against the Alliance of Motion Picture and Television Producers (AMPTP), which represents major studios. Elizabeth Benjamin, a drama writer, posted on social media, "Netflix posted a high-paying AI job listing amid actors' concerns about AI. Seeing this makes my stomach turn."