Large language models: fast proliferation and budding international competition
The improvements in the capabilities of Large Language Models (LLMs) in recent years now lead to a proliferation of their development globally. The strategic implications of LLMs are beginning to come into focus.
LLMs are increasingly being viewed as potential tools of disinformation due to their incredible ability to mimic human language use. Creative human prompting enables actors to bypass the guardrails implemented within major models like OpenAI’s ChatGPT. The realm of international competition broadly, moreover, is seeing an uptick in states wishing to develop their own indigenous LLMs as a ‘point of national pride.’ Great powers like the United States are utilizing export controls on advanced semiconductors and relevant hardware required to produce LLMs and other forms of AI against China as this competition heats up.