No slowdown! Human-level ChatGPT with GPT-5 may come out this year

OpenAI, the firm behind ChatGPT, announced the GPT-4 language model a few weeks ago. For now, the new model used by ChatGPT Plus subscribers has a much more advanced structure than the free version GPT-3.5. However, OpenAI is already working on the next major software upgrades for ChatGPT, and GPT-5 is said to be out this winter.

OpenAI can achieve AGI

If a reported report on GPT-5 capabilities is true, OpenAI is about to reach a huge milestone for ChatGPT. According to the reports, with the GPT-5, ChatGPT will reach the level of artificial general intelligence (AGI) and become indistinguishable from a human. According to developer Siqi Chen, OpenAI aims to complete the training process of GPT-5 and reach AGI this December. Additionally, GPT-4.5 is said to be available in September or October as an interim release.

This is incredibly exciting and scary at the same time as we don’t understand what AGI is capable of. On the bright side, the GPT-5 can greatly speed up processes, lighten workloads and save time with its productivity and advanced structure. After all, since this is a technology, serious effects may occur in the abuse part. Disinformation is one of them, and any social media is not prepared for it.

Although there is no full definition of AGI yet, the general opinion is that when artificial intelligence reaches AGI, it will have human-level comprehension and task understanding. So we can talk about reasoning. In other words, artificial intelligence can automatically perform something else necessary for the given task and give you the final result. We humans often do this in our daily life without realizing it. Therefore, chatting with ChatGPT after a GPT-5 upgrade can feel like talking to a human.

This is all speculation for now, as we don’t have an AGI model to compare ChatGPT to. High accuracy is what is expected of the GPT-5, whether it achieves AGI or not. On the other hand, there seems to be a need for an AGI definition. Human-level or not, the GPT-4 is already capable of incredible things. We can assume that the GPT-5 will be much more advanced than that, but it is clear that there is still some way to go for the AGI we see in the movies. Until then, regulators need to put artificial intelligence on a healthy footing.