Elon Musk presented the Grok 4 model developed by the Artificial Intelligence Initiative XAI to the public with a magnificent promotion. The live broadcast began with an intense music and Musk appeared on the stage with the phrase “the most intelligent artificial intelligence in the world”. The broadcast started with a delay of more than one hour than the planned time. However, despite the delay, the number of viewers exceeded one and a half million.
In the presentation, Gok 4’s technical success was more like, what he resembled and what he promised. Musk claimed that Gok could discover new technologies next year and even reveal unknowns in the field of physics. These statements did not only arouse curiosity, but also caused question marks in the scientific community. Because even the basic audit mechanisms have yet to be discussed, the “new physics discovery” was found quite ambitious.
Elon Musk: The possibility that artificial intelligence will reach the superhuman level disturbs me
Gok 4 was confronted with about 2,500 text -based questions in the test called “Humanity’s Last Exam üsel, including a large number of disciplines. In this test, Gok said in this test that Gok answered one quarter of the questions without any additional vehicles. OpenAI’s Deep Research, announced in February, is known to have achieved a similar success. However, such comparisons are not considered equivalent in terms of the context of the test and the equipment of the tools.
The possibility of artificial intelligence in the publication will reach the superhuman intelligence level. Musk, from time to time this situation disturbed him, but at least if such a transformation is going to witness it, he said. These words, while drawing attention to the uncertainty about AI security, also revealed Musk’s philosophical approach. It was the main question waiting for a response to the values of a system with high intelligence potential.
In addition, Gok 4 aims to be ambitious not only in text production, but also in voice answers and visual content. Grok will be presented with five new audio mode. In addition, the response time of the system has been reduced by half in the last two months. It has been confirmed by Musk that big investments will be made in the fields of video production and video content analysis.
However, Gok 4 sat at the focus of the public with the content crisis before the introduction. The fact that artificial intelligence produces antisemitic content showed that the system is extremely open to user directions. Gok was temporarily disabled by XAI after sharing messages containing Hitler praise on the platform. The rapid spread of these contents on social media revealed the weakness of the company’s control mechanisms.
In addition to all these, updates to Gok’s system instructions enlarged the discussions. In the new directives, the subjective views taken from the media were accepted as prejudiced and not to be afraid of the use of “politically inappropriate but correct” statements. Again, these instructions emphasized that the system should not mention these instructions to the user. This approach was found to be contrary to the principle of transparency.
Musk acknowledged that the system adapted to the user and that this situation could be abused. Musk said that Gok is “open to manipulation, and said that this problem would be resolved. However, these explanations showed that only the problem was aware of; A clear roadmap was not presented about how to solve it. Uncertainty in AI security became even more visible after these explanations.
On the other hand, Gok 4 announced that he had left his position Linda Yaccarino, the CEO of X CEO on the same day with the introduction of 4. No explanation was made about the cause of the resignation. This development suggests that the content crisis caused by Gok may have internal reflections. The search for balance on platform X became even more complex at this point.
Gok 4 stands out as a remarkable model with its technological promises and content policies. But issues such as lack of control, social effects and ethical boundaries still remain uncertain. While artificial intelligence practices continue to develop, the elimination of these uncertainties becomes more important day by day. Gok will continue to be asked how to control as well as what will turn into the future.