late people OpenAI Releasing is over GPT-4 of the new version of its Artificial Intelligence which constitutes the soul and central core of ChatGPT.
from his demonstration From the revisions, it became clear to us that this AI development represented a significant leap from what had been predicted.
But now Microsoft is pressing some warning signs, publishing the results of scientific research, where it concludes that GPT-4 is indeed showing signs of human intelligence.
It is a project that was previously seen as the stuff of science fiction and this is finally with this intelligent technology, with all the implications and dangers that this entails.
Under OpenAI’s own description, GPT-4 is now a much larger “multimodal (taking image and text input, and sending text output)” language model that, while less capable than humans in many real-world scenarios, presents a human. performance levels in various academic and professional benchmarks”.
But the conclusions of the research developed by Microsoft seem to find evidence that AI will go a little further than what its general description suggests.
GPT-4 and ChatGPT would show the true meanings of human-like general intelligence (AGI).
A group of Microsoft researchers published the ArchivesA new question titled Sparks of Artificial General Intelligence: Experiments with GPT-4which translates to Sparks of General Artificial Intelligence: first experiments with GPT-4.
In the said document, the team of scientists claim that GPT-4 has already shown the first signs of being an AGI, which would mean that it now has capabilities that are at the same level as humans or in some cases slightly above;
“GPT-4 performance is surprisingly close to human-level performance and often far exceeds previous models such as ChatGPT.
Given the breadth and depth of capabilities of GPT-4, we think it can reasonably be seen as the first (but still incomplete) version of what we think of as an Artificial General System (AGI).
We have shown that beyond language proficiency, GPT-4 can solve new and difficult tasks in mathematics, coding, vision, medicine, law, psychology, and more, using no special instruction.
These are some of the conclusions that the team of researchers came to after testing the capabilities of the GPT-IV. In that very document you dynamite part of the ideas. While the limitations of AI are listed in several sections.
While he has repeatedly emphasized that the claims about GPT-4 and its evolution into AGI do not mean that it can do something human or that it has unknown inner motivation, consciousness and goals.
Also, it will be worth noting the evolution of what is happening in this model.
