How to Think Computationally About AI, the Universe and Everything | Stephen Wolfram | TED #bigIdeas

Stephen Wolfram, in his enlightening TED talk, provides a profound vision of computation and its pivotal role in the future of AI.

He presents a unique theory that the universe is built upon discrete computational elements, akin to ‘atoms of space’.

Wolfram also introduces the concept of ‘ruliad’, a vast computational universe, and discusses the potential of AI to explore this space.

He emphasizes the importance of computational language, not just as a tool for humans, but also for AIs, and envisions a future where this language operationalizes everything we can think about, thereby bridging the gap between human understanding and computational reality.

AI has the potential to explore the ‘ruliad’ space.

However, the challenge lies in aligning this exploration with what humans and AIs understand.

Recent advancements in AI have been about creating systems that are closely aligned with human understanding.

Computational language is about something intellectually much bigger. It’s about taking everything we can think about and operationalizing it in computational terms. – Stephen Wolfram

A computational language has been developed to bridge the gap between human understanding and computational reality.

This language has the potential to operationalize everything we can think about in computational terms, thus granting us a sort of superpower to envision something in computational terms and bring it into existence.

Computational language is not only a tool for humans but also for AIs. An emerging workflow involves instructing an AI roughly what to do, and then it attempts to express that in computational language, which can then be executed to obtain a result.

The concept of computational irreducibility suggests that it’s not always possible to predict or explain what a system will do.

Instead, we must go through the same irreducible computational steps as the system itself to understand its behavior.

So computation isn’t just a possible formalization, it’s the ultimate one for our universe. – Stephen Wolfram

The use of AI presents a societal dilemma.

If we allow AI to reach its full computational potential, we won’t be able to predict its actions.

However, if we impose constraints to make AI more predictable, we risk limiting its capabilities.

AI could lead to a ‘promptocracy’ where people write prompts instead of voting.

While automation may seem to render humans redundant, history shows that it often opens up new opportunities and fields.

Defining what we want in clear terms is crucial as AI and automation can bring it to fruition.

The key to defining our wants effectively lies in computational language, which shifts the focus from mechanics to conceptualization.

Despite advancements in science and technology, the particulars of humans remain relevant.

Even our understanding of physics depends on how we humans sample the ‘ruliad’.