Google sees new AI threat

In a leaked document, Google says that both the company itself and ChatGPT builder OpenAI now face a new threat to AI development: the open source community. These big companies have been caught up without realizing it.
While the two companies kept an eye on each other, the third party with artificial intelligence went with it, the document warns. Open source developers work together as a community on the Internet, sharing everything they create. In this way thousands of developers can use the ideas of others and build on the work others have started. In this way, they can build artificial intelligence technology faster that is also more powerful than that large companies develop. Because the systems Google works on are large and complex, making it difficult to update them quickly. Plus, open source software is free, so why would people still want a paid version?
Finally, the document states “the inconvenient truth is that we are unable to win this arms race and neither is OpenAI”. The document warns that it's best to learn from it and try to collaborate with people who work outside of Google.

Secure, but slow

Google has repeatedly said it won't release its AI technology until it's responsible and secure, and that it's been relatively slow as a result. The report warns that other models are widely available without any restrictions.
Some of the open source work is based on previous research that companies have conducted. For example, the open source community now works with LLaMa, which was once made by Meta (Facebook), but was leaked in March and has since been renovated by the internet community.