OpenAI plans to release a new open-source AI model in response to increasing competition in the open-source large language model (LLM) space, according to a report by The Information. While this development is exciting, the model may not be as advanced as OpenAI’s proprietary model, GPT, and may not directly compete with it.
The report also states that OpenAI’s $27 billion private valuation relies on a future where the most advanced AI for commercial purposes remains proprietary rather than open source. This shift towards accessible AI development is influenced by pressure from Meta, OpenAI’s rival, which released several open-source models in February. As more developers choose free models, OpenAI’s move signifies a significant shift towards more democratic AI development.
Since the launch of ChatGPT, an AI chatbot that enables human-like conversations, there has been a surge of interest in generative AI. Microsoft’s substantial investment in OpenAI, the developer of ChatGPT, has intensified competition with Google, which recently showcased numerous AI advancements at its i/o conference.
The long-term performance comparison between open-source and proprietary models will be intriguing. However, discussions among AI proponents and critics now focus on broader concerns, such as misinformation and security.
OpenAI CEO Sam Altman recently testified before a U.S. Senate panel, addressing the risks and limitations associated with AI. Altman emphasized that a “great threshold” would be a model capable of persuading or manipulating a person’s beliefs. He also stated that companies should have the freedom to choose whether their data is used for AI training, although public web material would be considered fair game.
OpenAI has previously released open-source models like Point-E, Whisper, Jukebox, and CLIP. The performance of its new open-source model compared to its competitors remains to be seen.
