Tech
The ChatGPT Disruption: How One AI Shifted the Foundations of an Entire Field
ChatGPT disruption
It might be challenging to ask scientists to recognize a paradigm change, particularly in real time. After all, decades may pass before genuinely revolutionary advances in knowledge become apparent. However, you don’t have to use the P-word to admit that one specific area has evolved: natural language processing, or NLP. A great deal.
Making the erratic nature of human language (the “natural” portion) manageable for computers (the “processing” part) is the aim of natural language processing. NLP, a 1940s-era fusion of science and engineering, gives Siri a brain, Stephen Hawking a voice, and social media firms another tool to target us with advertisements. Large language models, a technology that NLP helped create but whose rapid development and revolutionary potential nonetheless caught many in the industry completely off guard, also emerged there.
In other words: The term “large language model” was never used in Quanta’s 2019 report on BERT, a then-groundbreaking NLP system. Just five and a half years later, LLMs are all over the place, sparking discussion, upheaval, and discovery in each scientific community they come into contact with. However, natural language processing was the first one they touched, for better, worse, and everything in between. How did those who were directly affected feel about that impact?
To convey that tale, Quanta spoke with 19 NLP researchers, both past and present. Experts, students, tenured professors, and startup entrepreneurs all share a sequence of events that altered their lives, including epiphanies, joyful experiences, and at least one “existential crisis.” as well as ours.