By Mathew in AI Ethics — 06 Aug 2024

The Future of AI: Navigating Between Utopia and Dystopia With Nick Bostrom

Introduction: AI at the Precipice

Throughout history, transformative technologies have fundamentally reshaped society. Written language gave rise to nation-states by enabling the tracking of laws and taxes. The printing press sparked religious persecution and wars. The internet created decentralized media and an era of conspiracy theories. But what will artificial intelligence bring?

According to Nick Bostrom, author of "Deep Utopia" and a leading voice on existential risk, we're rapidly approaching a technological inflection point where the development of artificial general intelligence (AGI) could completely transform human society. Unlike previous technological advances, AI may develop so rapidly that we might be just years—possibly even months—away from systems that can perform AI research better than humans, potentially triggering an "intelligence explosion" as these systems improve themselves at digital speeds.

"I think we are no longer in a position where we can be confident that it couldn't happen even within some very short period of time like a year or two," Bostrom cautions. "I'm not saying it will, but we are not in a position where we can be really sure that it won't."

This conversation examines what such a transformation might mean—from the potential dissolution of work as we know it to profound questions about human purpose, consciousness in digital minds, and whether our species would bifurcate into AI-embracing and AI-rejecting communities.

The Possible Futures of AI Dominance

Bostrom outlines several possible trajectories that AI development could take:

Scenario 1: AI Minds Disconnected from Humanity

In this scenario, AI eventually becomes a dominant force disconnected from its human origins, "roughly in the same way that we've kind of disconnected ourselves from the great apes or the Neanderthals." The implication is that humanity might be sidelined or even rendered obsolete by its own creation.

Scenario 2: Extreme Centralization of Power

Even if humans remain in control, AI could enable unprecedented concentration of power. Bostrom explains: "With automation of police forces and military forces, you could imagine an even higher concentration of power and better abilities to surveil what is going on... to keep track of what everybody's opinion is about the ruler and what they are doing."

Today, even dictators need the support of at least a fraction of the population—perhaps 10% including security forces and military. With AI automation, this dependency could vanish, allowing for more extreme totalitarianism than humanity has ever experienced.

Scenario 3: Hyper-Stimuli Hijacking Human Minds

AI might create what Bostrom calls "super memes" or virtual reality worlds so compelling that people essentially "check out of real reality." While we already spend hours on social media and television, future technologies could create experiences so immersive and addictive that they fundamentally alter social structures.

"This could maybe be kicked to the next level if you had just a higher level of technology doing that," Bostrom notes, suggesting we might already be seeing early stages of this phenomenon.

Deep Utopia: If AI Goes Right

Looking beyond potential dystopias, Bostrom's book "Deep Utopia" explores what the ideal human life might look like if AI development goes well. He defines deep utopia as "what would a great human life be like if you abstract away from a lot of the constraints that currently limit what we can do."

In this utopian vision:

AIs and robots handle all necessary work
Advanced biotechnology gives unprecedented control over our bodies and minds
Government functions effectively without wars or oppression

However, even this utopian scenario raises profound questions about human purpose. "What would give us purpose in our lives if there's nothing we need to do?" Bostrom asks. Currently, much of our dignity and self-worth comes from making useful contributions—being a breadwinner, contributing to society, or being valuable in some way. When AI can do everything better than humans, we'll need to fundamentally rethink our values.

The Psychology of Purpose in an AI-Solved World

Tom Bilyeu raises a critical concern about the human need for purpose: "Happiness derives from pursuit." Humans evolved over hundreds of thousands of years to do hard things for survival, with pleasure and pain as evolutionary motivators. The greatest satisfaction comes from pursuing valuable goals that benefit others, especially when success seems imminent.

"When people are in pursuit of something and it gives the meaning...man, that's when it feels good," Bilyeu argues. "The second you're either not working hard, you're just being given things, or you are working hard but it doesn't matter because everybody already has everything that they need...then you ask a fundamentally corrosive existential question, which is 'Why do I matter? Why exist at all?'"

Bostrom acknowledges this psychological reality but questions whether we value purposeful striving for its own sake or because it creates good mental states. In a technologically advanced future, these could be separated—perhaps through "perfect drugs" that induce the same satisfaction without requiring actual effort.

He distinguishes between artificial purpose (like playing golf, where the goal is arbitrary) and genuine purpose (like escaping a tiger, where the stakes are real). As AI advances, we might need to discover new foundations for meaning.

The Timeline and Transition to an AI-Dominated World

How quickly might this AI transformation occur, and how will humans respond? Bostrom suggests three potential transition speeds:

Extremely sudden - If superintelligence appears very rapidly, there won't be time for significant social polarization or resistance movements to form.
Extremely slow - If AI advances over many decades, it might create a "boiling frog" phenomenon where people gradually accept each improvement without noticing the profound transformation underway.
Intermediate turbulence - The most disruptive scenario would involve rapid but uneven change where "every other month there's like a new thing and now a big sector of workers were laid off" while various disasters occur as AI systems malfunction or are misused.

Bilyeu predicts a bifurcation of society, with some becoming "Puritans" who reject AI entirely while others embrace every technological enhancement available—perhaps eventually leading to physical augmentation through neural interfaces or genetic editing.

Bostrom acknowledges such polarization is likely but notes practical challenges to completely opting out of AI: "Unless you're really willing to completely rip yourself off the fabric of modern society," since AI will be embedded in everything from medical diagnoses to the electrical grid.

The Practical Implications: How AI Transforms Daily Life

What would an AI-transformed society actually look like? Bostrom offers concrete examples:

"Instead of having some guys who have to drive the garbage truck around the city every morning to collect the garbage, you could have a self-driving garbage truck with an Optimus robot that hops off and picks up your garbage can and does all of that automatically."

Beyond automation, he sees potential for massive acceleration of human progress:

"You could imagine accelerating and have a thousand years of medical research progress in just a couple of years when you have these digital minds working on this, like maybe unlocking cures to reverse the aging process, etcetera, and then forestalling a huge amount of human misery and death that is currently pretty unavoidable."

This technological utopia could include:

Personalized AI assistants managing your schedule, health, and preferences
Virtual elimination of mundane labor
Radically extended lifespans through medical breakthroughs
Enhanced well-being and abilities through biotechnology
More time for spiritual practices and aesthetic experiences
A high "safety net" ensuring no one falls into serious suffering

The Consciousness Question: Moral Status of AI

One of the most profound challenges Bostrom identifies is determining when AI systems might develop consciousness and deserve moral consideration.

"If at one extreme you had an AI that was exactly functionally identical to a human, that had a human-like body, human-like memories, that had a brain structured very much like a biological brain, I think in that case there would be a very strong moral case that we should treat it as a moral subject as well," Bostrom argues.

He points out that we already grant some moral status to animals based on their capacity for suffering. "If even a mouse has a possible claim to sentience and at least some simple form of moral status, then AI systems that are roughly equivalent to a mouse in their behavioral repertoire, I think, would also be prima facie candidates for moral status."

This raises complex questions about creating simulations containing conscious entities that might suffer. If we create AIs sophisticated enough to have subjective experiences, we would have responsibilities toward them that we don't have toward mere tools.

Bostrom predicts extremely rapid advancement in AI social companions—not just sex robots, but friends, fans, and various sources of social interaction. These might become increasingly appealing compared to human relationships.

"Maybe that if that becomes good enough would be extremely compelling to people... I mean, drugs are really compelling, and often the worse your real life is, the more attractive the alternative of some opiate or something is."

While acknowledging the dystopian elements of a world where humans interact primarily with AI bots rather than each other, he notes that younger generations might simply see this as a natural preference: "These humans are kind of a drag and we're all just using these AI bots now."

Navigating the Coming AI Transition

For those wondering how to prepare for this uncertain future, Bostrom offers several considerations:

Stay adaptable by familiarizing yourself with current tools and understanding where technology is heading.
Consider careers focused on human connection since care professions may remain valuable even in an AI-dominated world.
Don't sacrifice the present for far-future rewards that may never materialize. "Don't forget to actually enjoy life right now... I wouldn't sort of plan on a 40-year career and make big sacrifices now for 10 or 20 years in the hope of then it paying off when you're in your 50s and 60s because maybe the future doesn't exist at that point."
Maintain a hedged approach since AI development could still be halted or regulated: "You don't want to end up completely dry either, where you have nothing... if the AI Revolution doesn't happen."

Conclusion: Facing an Uncertain Future

At its core, Bostrom's conversation with Bilyeu raises fundamental questions about what it means to be human in an age where our technological creations may soon surpass us in every measurable capacity.

The transition to an AI-dominated world presents both unprecedented opportunities and existential risks. While AI could eliminate disease, extend lifespans, create material abundance, and free humanity from drudgery, it simultaneously challenges our deepest sources of meaning, threatens social cohesion, and raises the possibility of human obsolescence or extinction.

As Bostrom reflects, the most likely outcome may not be clearly good or bad, but something so transformed that "if we could see it, we wouldn't necessarily know even what to think of it."

Much like a child growing into an adult with completely different interests and capabilities, humanity itself may be on the cusp of a developmental leap so profound that our current values and concerns might seem as distant as a five-year-old's toys seem to an adult.

"Maybe 80 years is just not enough to really fully realize our inherent potential," Bostrom suggests. "We are kind of cut short by our rotting biology." The next phase of our existence, whether through AI augmentation or transformation, could represent a maturation process as natural as growing up—if we navigate the transition wisely.

Key Points

AI timeline acceleration: We may be just years away from artificial general intelligence that could trigger an "intelligence explosion" as AI systems improve themselves at digital speeds.
Competing futures: AI could lead to human obsolescence, unprecedented power concentration, or immersive experiences that disconnect people from reality.
Purpose problem: When AI can do everything better than humans, we face a crisis of meaning and purpose that may require redefining our fundamental values.
Consciousness in machines: As AI systems become more sophisticated, we'll need to determine when they deserve moral consideration—potentially even before we fully understand consciousness.
Social transformation: AI companions could fundamentally alter human relationships, with younger generations potentially preferring AI interaction to human connections.
Preparation strategy: Rather than making decades-long career plans, adaptability and focusing on human-centered skills may be more valuable as AI rapidly transforms the job landscape.
Evolutionary perspective: The transition to an AI-enhanced future might represent a natural maturation of humanity beyond biological limitations, similar to a child growing into an adult.

For the full conversation, watch the video here.