For the future of science, join the open source LLM camp!
The free ChatGPT is very cool to use, but the biggest disadvantage of this closed-source language model is that it is not open source, and the outside world cannot understand the training data behind it and whether it will leak user privacy. Jointly open source a series of alpaca models such as LLaMA.
Recently, an article was published in the Nature Worldview column. Arthur Spirling, a professor of political and data science at New York University, called on everyone to use more open source models. The experimental results can be reproduced and are also in line with academic ethics.
Why do I recommend everyone to use open source generative AI models?
Researchers should avoid the temptation of proprietary models and develop transparent large language models to ensure reproducibility.
Arthur Spilling
It seems like every day a new large language model (LLM) is released, and its creators and scholars alike draw breathless commentary on its remarkable ability to respond to human cues. It can fix code! It can write a letter of recommendation! It can summarize an article!
From my perspective, as a political and data scientist who uses and teaches such models, academics should be wary. The most widely touted LLMs are proprietary and closed: run by companies that don't expose their underlying models for independent inspection or validation, so researchers and the public don't know which files the model was trained on.
The rush to incorporate such artificial intelligence (AI) models into research is a problem. Their use threatens hard-earned progress in research ethics and the reproducibility of results.
Instead, researchers need to collaborate to develop open-source LLMs that are transparent and independent of corporate favors.
Proprietary models are really convenient and work out of the box. But one must invest in open source LLMs by helping to build them and use them for research. I'm optimistic that they will be widely adoptedÿ