NO HANDS: New Japanese firm trains robots without human input

A Japanese tech firm says it is moving toward superintelligence with a big step forward in AI.

Integral AI, which is led by a former Google AI employee, announced in a press release that it had made significant progress with its artificial general intelligence model, which can now acquire new skills without human intervention.

'Integral AI’s model architecture grows, abstracts, plans, and acts as a unified system.'

The AI system allegedly learns its new skills "safely, efficiently, and reliably," the company said, while claiming that the AI had surpassed its defined markers and testing protocols.

As such, the AGI is allegedly capable of autonomous skill learning without using pre-existing datasets or human intervention. Integral also said the system is able to develop a "safe and reliable mastery" of skills, meaning that it does produce any "catastrophic risks or unintended side effects."

What those risks or side effects might be is unclear.

RELATED: Artificial intelligence is not your friend

Photo by David Mareuil/Anadolu via Getty Images

The last parameter, which Integral AI said its system adhered to, was to be energy-efficient. The system was tasked with limiting its energy expenditure to that of a human seeking to acquire the same skill.

"These principles served as fundamental cornerstones and developmental benchmarks during the inception and testing of this first-in-its-class AGI learning system," the press release said. Integral added that the system marked a "fundamental leap beyond the limits of current AI technologies."

The Tokyo tech company also claimed its achievement was the next step toward "superintelligence" and marked a new era for humanity, with the AI's learning process allegedly mirroring the complexity of human thought.

"Integral AI’s model architecture grows, abstracts, plans, and acts as a unified system," the company wrote, adding that the system will serve as the groundwork for "unprecedented adaptability," particularly in the field of robotics.

This means that with the help of this AGI, autonomous robots would be able to observe and learn in the real world and conceivably pick up new skills in real-world environments without the help of pesky humans.

RELATED: ART? Beeple puts Elon Musk and Mark Zuckerberg heads on robot dogs that 'poop' $100K NFTs

Photo by David Mareuil/Anadolu via Getty Images

Jad Tarifi, CEO and co-founder of Integral AI, called the announcement "more than just a technical achievement" that is "the next chapter in the story of human civilization."

"Our mission now is to scale this AGI-capable model, still in its infancy, toward embodied superintelligence that expands freedom and collective agency," Tarifi added.

According to Interesting Engineering, the Lebanese founder said he worked at Google for a decade before starting his own company. He allegedly chose Japan over Silicon Valley because of Japan's position as a world leader in robotics.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!