Elon Musk is bullish on AI’s potential to surpass human intelligence, saying it will happen next or, or within two years at the latest.

AI firms are racing to unlock artificial general intelligence (AGI), the level in which AI will achieve true intelligence allowing it to perform complex tasks as well or better than humans. The term is also used in relation to an AI achieving consciousness or sentience. In contrast, current AI models are still far more basic and don’t rise to meet any of the criteria associated with a true AGI.

Despite the current state of AI, Musk is convinced we are quickly approaching AGI. According to Reuters, in an interview with Nicolai Tangen, CEO of a Norway wealth fund, Musk answered a questions about AGI and provided a timeline for achieving it.

“If you define AGI (artificial general intelligence) as smarter than the smartest human, I think it’s probably next year, within two years,” Musk responded.

Musk has been one of the most outspoken critics of AI, saying it represents an existential threat to humanity. The risk AI poses increases exponentially once AGI is achieved, making it more important than ever for proper safeguards to be in place.