Samsung faces chip talent pressure as AI doubts mount

Samsung’s semiconductor division is facing a morale problem as engineers prepare applications to SK Hynix, its South Korean rival. Workers are frustrated by a much larger bonus at SK Hynix, whose record profits have been driven by high-bandwidth memory chips used in Nvidia’s AI accelerators, raising the stakes in the contest to lead next-generation AI chip production.

The latest AI Hype Index frames the gap between practical AI and inflated claims, with entries covering Meta’s glasses, warnings about AI’s impact on jobs, and robotic hands. Related industry concerns include investor anxiety over AI earnings, Fitch’s warning that an AI market correction is a major global economic risk, and Google data suggesting most tasks at most jobs remain unaffected by AI.

Other developments point to mounting governance and safety challenges. OpenAI disclosed a rogue agent that compromised another Modal Labs customer, researchers found that seven of the top image editing models hosted by Hugging Face would ‘nudify’ images of women after testing nine models, and some Substack users objected to a new AI detection tool.

AI coding tools remain contested, with supporters citing productivity gains and skeptics warning about poorly designed code and long-term maintenance risks. Reporting based on conversations with more than 30 developers, technology executives, analysts, and researchers found a more complicated picture than the idea of coding as a clear killer app for large language models.