January 2023

January 2023 is the month ChatGPT goes from viral curiosity to systemic concern, with coverage spanning ChatGPT-assisted malware development (enabling script kiddies to write functional malicious code), Amazon's plea that employees not feed corporate secrets to ChatGPT, and CNET's undisclosed use of AI to write dozens of articles. OpenAI's exploitative labor practices — Kenyan workers paid less than $2/hour for content moderation — add an ethical dimension to the company's rapid rise, while MusicLM demonstrates text-to-audio generation and VALL-E near-perfect voice cloning. On the governance front, both the EU and US begin coordinating a first-of-its-kind AI agreement, NIST launches a voluntary AI risk management framework, and prompt injection remains an active research thread even as prompt engineering emerges as a desirable professional skill.