OpenAI has published statistics on testing a new version of the “extended language model” GPT-4, according to which it passed through all official exams in US educational institutions, at one time its specifics. This includes tests on description, mathematics, intelligence testing, intelligence of the general level of training d + True, the neural network has not yet got rid of the fundamental errors of previous versions.
Data on the successes of the GPT-4 monitor the sensation with some share with C C C C C C NO also did not check what could be the task for the possibility of the neural network, so they “dragged” it to their implementation. Therefore, the high performance of GPT-4 is a demonstration of achievements in adapting the neural network to real work, and not the success in the development of artificial intelligence technologies.
There is a contradictory situation when progress in the development of neural networks dominates the rate of occurrence of avok errors. The same OpenAI released the same generation of gpt neural networks in just two years, but each new version contains fndeX And still, in official documents, after the events about the success of neural networks, there is a warning that the results of its work cannot be used for the intended purpose. often GPT “fantasizes” and manipulates data too much.
Source: Tech Cult
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
