The OpenAI reality show may have ended, but the industry’s attention remains on Sam Altman and other senior executives in the company. This is because OpenAI reportedly made a breakthrough in artificial intelligence that sources say panicked scientists, which ultimately led to Altman’s shock ouster.
Q* (pronounced “Q star”) is the codename for OpenAI’s ‘breakthrough’ project, and is believed to be the reason why Sam Altman was fired from OpenAI.
What is Q*?
First reported by Reuters, scientists at OpenAI believed the discovery of Q* marked a significant step towards the company’s search for ‘artificial general intelligence’, or AGI.
In the AI field, AGI is a hypothetical type of intelligence where a machine is able to perform tasks that humans or animals are capable of doing. For Q*, a source told Reuters it was able to “solve certain mathematical problems”.
“Though only performing math on the level of grade-school students, acing such tests made researchers very optimistic about Q*’s future success,” according to the report by the source to Reuters.
As AI models like ChatGPT and Bing Chat are so far mainly capable of tasks like writing, language translation, and planning, the ability to correctly solve mathematical problems is considered a milestone in AI development. In very simple terms, this is because there can only be one answer in maths equations, and arriving at said answer requires human-level logical reasoning.
In April 2023, Sam Altman himself wrote in a blog post describing AGI as essentially any machine that is “generally smarter than humans”. So far, there hasn’t been an AI model that is thought to possess AGI – until now, it seems.
Q* and AGI
After OpenAI’s scientists unearthed Q*, they panicked, and wrote a letter to Altman and Greg Brockman. In the letter, they warned both co-founders the potential dangers of Q* but did not specify the exact concerns.
OpenAI CTO Mira Murati, who was briefly interim CEO, was the one who allegedly told staffers that Q* was the reason for the plot to throw out Altman.
However, it is important to note here that Reuters was not able to review said letter, nor were they able to independently verify Q*’s AGI capabilities. So, there’s probably little to worry about… for now.