OpenAI CEO Sam Altman is warning people to not get their hopes up too high for GPT-4. GPT-3, the first version of OpenAI’s powerful natural language processing system, was released in 2020 while its improved version, GPT-3.5, powers the ChatGPT chatbot. In a recent video interview with StrictlyVC, Altman was asked about the expectations that GPT-4 will be released in the first half of the year and responded “It’ll come out when it’s ready. People are begging to be disappointed.”
Misplaced expectations around any new product or version can lead to disappointment, and that’s exactly what Altman has warned against. While GPT-3 has been a revolutionary development in natural language processing and machine learning, everyone should proceed with caution when it comes to GPT-4. So far, there has been no confirmation on exactly when the new version will be available.
In the meantime, companies should focus on staying up to date with the current capabilities of GPT-3. This includes being aware of the system’s strengths (e.g. understanding natural language and learning from context) and weaknesses (e.g. not understanding concepts beyond its training data) and tailoring applications accordingly.
At the moment, the best thing to do is to remain patient. When GPT-4 is released, then it is time to start exploring its new and improved capabilities and perhaps even begin planning new applications of the technology. Until then, don’t get your hopes up too high and keep your eyes open to the latest GPT-3 advancements and new applications developed from it.