How long will it take for artificial intelligence to reach human-level intelligence?
According to Adam D'Angelo, a member of the OpenAI board of directors, this milestone event could occur within "5 to 15 years."
D'Angelo, who is the CEO and co-founder of Quora, made this prediction during a recent event. He stated that the emergence of Artificial General Intelligence (AGI) will be "a very important change in the world when we get there."
His comments come on the heels of reports earlier this month that OpenAI has developed a method to track its progress in building AGI, sharing a new five-level classification system with its employees.
The company believes that its current AI has reached the first level, which involves conversational interaction with humans, and is progressing towards the second level, which involves solving problems as well as a human with a doctoral degree.
The next levels involve the ability to work on behalf of users for days, develop innovative products, and eventually take on organizational tasks in the fifth level AI system.
Last fall, OpenAI CEO Sam Altman and CTO Mira Murati stated that AGI will be achieved within the next 10 years.
"We believe that if you give people better tools, they will do amazing things. I think AGI will be the best tool humans have ever created," said Altman.
Recent reports from PYMNTS suggest that these efforts have sparked speculation in the business world about the potential for AI-driven commerce to rewrite global trade rules, assuming the technology lives up to expectations.
"OpenAI's pursuit of human-level reasoning ability is not just a technological marvel; it is a narrative that drives breakthroughs and inspires new possibilities in various fields," said Ghazenfer Mansoor, Founder and CEO of Technology Rivers, to PYMNTS. "In the business realm, AI can greatly change the way supply chains are managed, predict market trends with high accuracy, and provide highly personalized customer experiences on a large scale."
Earlier this year, OpenAI employees reportedly showcased an AI model capable of answering complex scientific and mathematical questions, with one model scoring over 90% on a championship-level math dataset. The company also recently demonstrated a project with novel human-like reasoning skills at an internal meeting.
"The algorithm works by creating multiple options, following a tree-like structure of possibilities, and then reasoning the results and selecting the best path," said Alexander De Ridder, CTO of SmythOS, to PYMNTS. "This is similar to how a chess player thinks through different moves before choosing one."
He suggests that OpenAI's innovation may involve "algorithmic breakthroughs in how to do this efficiently and at scale," possibly combining "autonomous network research and tool usage to achieve breakthroughs in reasoning."
