MENLO PARK (ANP) – Facebook parent company Meta and technology group IBM are establishing an alliance with more than fifty companies and organizations that will share knowledge about artificial intelligence (AI). In doing so, they hope to stimulate innovation and quickly detect and limit risks associated with new technologies.
The group, which also includes universities, wants to increase the number of AI models that are available for free to researchers. The idea is that by having many people work with these types of models at the same time, knowledge about them will increase faster.
Benefit and safety
Meta previously released large language models on which AI chatbots run. These are models, or LLMs for short, that learn to predict what logical sentences are based on large amounts of text. “We think it’s better if AI is developed in an open way. More people can benefit from it, create innovative products and work on safety,” said Meta director Nick Clegg.
OpenAI does not participate in the partnership. The company behind ChatGPT caused a stir last year with its advanced AI programs that create texts or images themselves. The recent turmoil at the company, where CEO Sam Altman was fired and returned shortly afterwards, fueled discussions about the level of transparency needed in the development of new AI technologies.