OpenAI CEO Sam Altman has declared that the artificial intelligence system ChatGPT will be much less lazy in the future. This announcement comes after reports surfaced last year of the underlying model, GPT-4, responding less fully or refusing to answer certain inquiries.
In December, OpenAI revealed that the performance of its large language model had deteriorated, leading to it becoming lazy following an update. Altman acknowledged on social media platform X that GPT-4 had experienced a slow start to its new year’s resolutions. Despite efforts to address the issue through updates, immediate improvements were not seen. However, Altman assured that the system is now expected to be significantly less lazy.
Large language models like ChatGPT have proven to excel in automating repetitive and mundane tasks. They are relied upon to provide answers ranging from simple to complex, with the expectation of useful responses.
Last year, reports began circulating that ChatGPT was displaying signs of laziness by providing incomplete responses or snippets instead of fully formed functions. Some speculated that this was a response to updates aimed at combating misuse and enforcing safeguards against illegal uses, while others believed it was an attempt to reduce the cost of running the model.
While the exact methods employed by OpenAI to address the laziness problem remain undisclosed, the company has released updates to the underlying model since the holiday season. These updates include new versions of the Turbo model, designed to enhance response speed without compromising quality.
OpenAI has introduced GPT-4-Turbo, which can complete tasks like code generation more thoroughly, and GPT-3.5-Turbo, which reduces the overall cost of task completion and query responses.
Despite speculation that the AI model had slowed down during the holiday season, OpenAI denied this, and the system did not immediately return to normal in January. Altman’s comment about ChatGPT having a slow start on its new year resolutions humorously referred to the updates still being implemented. Ultimately, AI responds to input based on pre-trained data and lacks the sentience to make independent choices.
In conclusion, Altman’s declaration indicates that ChatGPT will be more prompt and responsive in the future, addressing the previous issues of laziness. The updates implemented by OpenAI aim to enhance performance and efficiency, allowing ChatGPT to fulfill tasks more comprehensively and reducing overall costs. Users can anticipate a more agile and reliable AI system, catering to a range of queries and tasks efficiently.
