laitimes

The AI big model GPT-4 has become stupid, can there be no need to panic?

author:Look up at the starry sky and explore science and technology
The AI big model GPT-4 has become stupid, can there be no need to panic?

GPT-4 (Generative Pre-trained Transformer 4) is a large-scale, multi-modal artificial intelligence model developed by OpenAI, which has excellent performance and predictive capabilities to help people better understand and process natural language data. GPT-4 excels in people's fears and concerns about AI technology.

Musk and other technology bigwigs and AI experts jointly issued an open letter, calling on all artificial intelligence laboratories to immediately suspend the training of artificial intelligence systems more powerful than GPT-4 for at least half a year. This is because today's artificial intelligence systems can already compete with humans in their ability to think independently. This call has attracted widespread attention and triggered a rethinking of artificial intelligence technology.

Recently, a large number of users have reported that the quality of answers to large models has deteriorated recently, thinking that it has become dumb. There has been a lot of user feedback that the quality of answers to large models has deteriorated recently, especially in terms of program generation, and GPT-4 currently generates code that often has errors. So why do so many people feel it is performing poorly?

First of all, it may be due to psychological effects. Some insiders believe that it is not that the performance of GPT has deteriorated, but that everyone has passed the initial surprise period and their expectations for it have become higher. We may have high expectations for GPT-4's performance, so when it performs slightly less than satisfactory in some way, we feel like it's dumb.

In addition, there is a speculation that with the plugin enabled, the plugin's extra prompt may be a kind of "pollution" to the problem to be solved. That is, when we use GPT-4 to solve a problem, if some plugins are enabled at the same time, they may interfere with the GPT-4 generation process, causing the quality of the answers it generates to deteriorate.

Of course, these are just some of the possible reasons, and we need more actual data to verify these guesses. However, whether GPT-4's performance really deteriorates or not, we should maintain an objective attitude, not rely too much on it, and always pay attention to its limitations and shortcomings. After all, GPT-4 is just a tool, and its performance is ultimately determined by our own intelligence and ability.

The AI big model GPT-4 has become stupid, can there be no need to panic?
The AI big model GPT-4 has become stupid, can there be no need to panic?
The AI big model GPT-4 has become stupid, can there be no need to panic?
The AI big model GPT-4 has become stupid, can there be no need to panic?