Openai can only achieve general artificial ielligence (Agi) when its systems produce maximum profit.
Microsoft and Openai have agreed to provide a new and specific definition of public artificial ielligence (AGI) or Artificial General Ielligence. According to a report, Openai can only reach AGI when the system produced by it can produce $ 5 billion.
For Microsoft, this definition of Agi is of great importance. According to one of the clauses of the agreeme, when Openai reaches AGI, Microsoft will no longer be able to access the most powerful models of this artificial ielligence startup.
However, reports suggest that Openai is trying to remove the clause from the agreeme. In fact, if we wa to judge the new AGI definition, Openai is still far from it. This year, Openai suffered a $ 2 billion loss, while its revenue was $ 1.5 billion. Reports also suggest that Openai will not become a profitable eity by year 3.
A clear and precise definition of AGI was needed for a long time. When OpenAI released its most powerful series of O3 models, rumors were raised about the access to AGI due to its performance in evaluation criteria.
The O3 model in the Arc-Agi criterion earned approximately 2 % pois and surpassed human performance. Of course, the creator of this criterion, François Schwelle, wrote on the X platform: “I do not believe this is AGI, there are still simple tasks in Arc-Ange-1 that O3 cannot solve.”
“As I approach AGI, I think the term is less commonly used,” said the Altan Symposium in the FinREGLAB AI Symposium. “In fact, people use it in differe meanings.” He poied out that for some people, AGI, the extraordinary ielligence that surpasses humans Some may think that AGI will be smarter than all humans. Even when people believed that the O1 model shows the early signs of AGI.
However, for Microsoft, AGI means more money. Of course, Satya Nadella does not allow the poison to announce AGI with only one meal definition. In fact, it seems that their relationships are stressful. Microsoft is planning to reduce its dependence on Openai and add more iernal artificial ielligence and third -party models to Microsoft 365 Copilot.
This is due to Microsoft’s efforts to reduce costs and diversify artificial ielligence models. The company is also working on iegrating its models io 365 Copilot.
In addition, according to Dylan Patel, founder and senior analyst at Semianalysis, Microsoft may soon invest in Ahropic responsible for the developme of Claude models. Dylan Patel also posted a post on the X platform, saying, “This new financing round could raise the company’s value to $ 5 billion.”
