OpenAI releases update to fix ChatGPT 'laziness'

OpenAI announced last Thursday (25) that corrected ChatGPT’s “laziness” problem reported by many users in late 2023. The solution arrives through an update to the generative artificial intelligence (AI) platform, serving subscribers of the GPT-4 Turbo version first.

Between November and December 2023, many complaints began to emerge about the chatbot takes time to complete certain tasks requested by users. At the time, the startup recognized that ChatGPT was lazy and would work on a fix.

Some users have come across lazy ChatGPT in recent weeks.Source: Unsplash

According to the developer, the problem was being caused by unpredictable behavior of technology and lack of updates. As a result, generative AI began to perform below average in some cases, and may not fully answer all the questions asked.

With this latest update, the trend is that the bot use fewer shortcuts and reduce the number of failures while trying to respond to requests, as OpenAI explained. The update includes other improvements that help make the platform more intelligent and accurate when interacting with people.

ChatGPT problem may continue

At this point, lazy ChatGPT will no longer be a problem only for those using the GPT-4 Turbo model, as previously reported. Subscribers of the GPT-4 version and free users of GPT-3.5 will not yet have access to the bug-fixing update.

Therefore, it is possible that those who use these last two models still continue to notice problems related to the bot's lazy responses. Although OpenAI has not confirmed, It is speculated that the update will also be made available for other versions soon.

Bringing a more up-to-date knowledge base, GPT-4 Turbo presents greater agility to carry out complex tasks, which can be completed with just a text prompt. The model should launch new functions for subscribers soon.