Experts Explain What AI Superintelligence Means And How Close We Are

TABLE OF CONTENTS

1. What Is AI Superintelligence (ASI)?
2. How Close Are We To AI Superintelligence?
3. How Could AI Go From AGI To Superintelligence?
4. What Could Superintelligent AI Unlock?
5. What Are The Risks Of AI Intelligence Surpassing That Of Humans?
6. Can AI Superintelligence Be Developed Safely?
7. Frequently Asked Questions (FAQs)

The term superintelligence is stirring up confusion, skepticism and angst. Lately, it has been pushed as a new branding for artificial intelligence. The White House directed all federal departments and agencies to use the term super intelligence versus AI, in all documents and communications. However the term has been used for some time to describe a higher level of AI, exceeding human capabilities, that is not quite here yet. For many observers across the industry, AI superintelligence is a yet-to-emerge technology that will reshape our world and businesses in ways we cannot yet anticipate, both with great promises and great risk.

What Is AI Superintelligence (ASI)?

There is no agreed-on definition of what AI superintelligence means, what it will look like or when it will arrive. It is purported to represent an upcoming phase of AI, either stepping up from or competing with numerous terms being bandied about such as “artificial general intelligence (AGI),” “artificial superintelligence” or “hyperintelligence.”

AGI is seen as the most likely progression up from today’s AI. This takes the form of machines that will be capable of supporting all aspects of knowledge across any and all domains. AI superintelligence, the next step up, could be seen as AGI on steroids — vastly accelerating human capabilities beyond what current or planned models can deliver, while constantly improving themselves through continuous learning, free of human intervention.