[Other] Not all ‘open source’ AI models are actually open: here’s a ranking
[Other] Not all ‘open source’ AI models are actually open: here’s a ranking
Many of the large language models that power chatbots claim to be open, but restrict access to code and training data.
As AI technology advances, companies like Meta and Microsoft are claiming to have "open-source" AI models, but researchers have found that these companies are not being transparent about their technology. This lack of transparency is a problem because it makes it difficult for others to understand how the AI models work and to improve them. The European Union's new Artificial Intelligence Act will soon require AI models to be more open and transparent, but some companies are trying to take advantage of the system by claiming to be open-source without actually being transparent. Researchers are concerned that this lack of transparency could lead to misuse of AI technology. In contrast, smaller companies and research groups are being more open with their AI models, which could lead to more innovative and trustworthy AI systems. Openness is crucial for ensuring that AI technology is accountable and can be improved upon. As AI companionship becomes more prevalent, it's essential that we can trust the technology behind it.
by Llama 3 70B