The setup of Pascal’s Wager, as this argument is generally known, is quite similar in form to Newcomb’s paradox. The glass box with the visible $1000 bill is your ordinary life on earth: you know it exists, and is yours to spend. The opaque box is your eternal reward. It might be empty, or it might be filled with a vast reward far beyond the one in the glass box. You will discover which one is the case only when you die and the box is opened. Do you take the glass box with the known, but finite reward, or the opaque box that could have nothing or everything inside it?
The thing about the Basilisk that makes it so scary is its combination of vast power with certain both human and mechanical weaknesses. It is designed by human beings to be the greatest and most benevolent force in the universe, but all we can gift it is our best guess at an ultimate rational moral standard, utilitarianism, the greatest good for the greatest number. And as a machine, it administrates this implacably, and entirely without mercy. Roko’s Basilisk is scary because it is simultaneously our parent and our child.
Another possible strategy for fending off the robot apocalypse is to ask if there are characteristically human traits or characteristics that are humanity-preserving, and if so, can those be passed along to our machines? What is it that has given us our identity as a species, all these years, and that, if we lose, we run the risk of losing everything?
Given how likely killer robots are, and how clearly the paths we are currently embarked on lead to that eventuality, can this destiny be averted? Acceptance of the unstoppable inevitability of progress is the motivation behind yet another approach to artificial intelligence called “Friendly AI.” It starts with the assumption that runaway technological progress is inevitable, that some one among the many teams around the world working on artificial intelligence will soon succeed, and that disastrous robotapocalypse is the far most likely result. Given that, the belief of the Friendly AI camp is that it is absolutely essential that we ensure the first artificial superintelligence is “friendly,” meaning that it has the best interests of humanity at heart, and is willing and able to protect us from its nastier cousins.
Three short episodes (on Sartre, Nietzsche, and Machiavelli) by Mark Linsenmayer of a new potential podcast for the PEL network. We’d like your feedback, and even more importantly, the feedback of your friends for whom the long-form PEL discussions are or would be just TOO MUCH.
In 1989, Star Trek: The Next Generation, the second major iteration of the durable televised Star Trek science fiction franchise, introduced a terrifying new villain called “the Borg.” An unhallowed melding of a humanlike life form with cybernetic technology, the individual members of the Borg were born, raised, lived, and presumably died entirely surrounded by technological innovations. There was no such thing as “natural childbirth” for them, they were cloned mechanically, nurtured in artificial wombs, and raised to maturity in pods. An implacable collective intelligence, they mercilessly converted any creatures they encountered into extensions of themselves, cannibalizing their planets for raw materials, and sucking other intelligent lifeforms into the inescapable machine.
Marie Kondo, the Japanese organizing consultant devoted to uncluttering our households. Socrates, the ancient Greek philosopher devoted to uncluttering our minds. Can their methods of tidying up help us live happier lives?
If you plot the graph of technological progress, it looks exponential. It is long and nearly horizontal extending into the past, it curves rapidly upward in the present, and many people expect that it will be nearly vertical at some point in the near future. The question is, what happens then? The technological singularity is the idea that at some point, perhaps even in the next few decades, computing power will essentially become infinite.
It is possible, given that we still understand so little of the brain, that it has evolved in such a way that it does bridge the gap between the subatomic world and the macroscopic world? Perhaps the free will of the quark is transmitted up through the intermediary of the brain and into the otherwise deterministic macroscopic world. But if this is true, does it preclude the possibility of a truly living simulation? Are the human beings inside the computer doomed to be dead, deterministic automata, lacking the quantum free will of the real ones?
The paired opposite to reductionism is called emergentism, and in recent years it has begun to gain an increasing number of advocates. In summary, it means that the whole is more than the sum of its parts. Unexpected behaviors and properties can emerge, even from simple well-understood parts, at high enough levels of organization… Some of the ways emergentists have proposed creating artificial intelligence include building or simulating artificial neural nets, or using quantum computers, which take advantage of wave-particle duality and superimposition to perform fuzzy logic. Others reject the entire idea of shortcuts to emulating human intelligence, in favor of simply duplicating the entire fine structure of the human brain in virtual form –something not possible today, but perhaps in the future.
At root, Bostrom’s argument hinges around a single controversial question. Is it possible to truly create or simulate a person? Is there any point, with any level of technology, no matter how advanced, that this becomes possible?
We left off last week with the question of how much weight we should give to Nick Bostrom’s argument that we are not only possibly simulated, but likely to be so. This argument, or at least our representation of it, rests on two key claims: first, that our descendants will be able to create people just like ourselves; and second, that they will create a lot of them. The argument is compelling only in the case that both are true.
Although there have been attempts at creating true simulations of intelligence, machines that can learn and respond appropriately to unbounded input, they have not, as of the time of this writing, progressed significantly far in the way of believably duplicating human interactions (although they have mastered tasks as diverse as as playing chess, competing on the television game show Jeopardy, and identifying other robots as robots). Are these major steps on the pathway, or deceptive dead ends? Could technology ever improve to the point where it could convincingly simulate, not you perhaps, but other people, in all their deep, multifaceted, and endlessly surprising soulfulness? Is true artificial intelligence, to the point that computers could believably create people, actually achievable?
The technological ability to emulate a convincing world is plausible in the not-so-distant future. We additionally know that the motivation to create one already exists, given the huge popularity of video games, and the amount of money and effort put into making them. A big difference, however, between a current-day video game and this potential game of tomorrow, is that the player of a current game knows she is playing a game. Could we really be in a game and not know it?
We all have a solipsistic experience nightly, when we sleep and dream. Each night we inhabit a universe which seems to us, convincingly at the time, to have a wealth of external people and places in it. But all of those people and places are created inside our brains solely for the benefit of the dreamer. In the modern world, however, we can place an additional, familiar experience of a solipsistic reality next to that of the dream: the single-player video game.
One of the first things people discovered when modern computing became a reality is that it’s relatively easy to simulate laws of physics, representing aspects of the real world. This theoretically enables an approach to simulation that builds an entire universe from basic building blocks. A quark could be a tiny bit of fundamental matter (whatever that might be), but it could just as easily be a rule programmed into a computer—or perhaps even a coherent thought in the mind of an all-powerful intellect.
In the year 1999 CE, just on the cusp of a new millennium, the then Wachowski Brothers released “The Matrix,” one of the most influential, imitated, and widely discussed movies of its times. It was only four years later, in 2003 CE, that philosopher Nick Bostrom of Oxford University introduced an argument that it is not only possible we are living inside a computer simulation, it is actually significantly likely. Although it may have sounded like a high-concept science-fiction thriller, the argument drew upon well-established lines of logic and a widely held series of assumptions.
Did Nick Bostrom, professor of philosophy at Oxford University, provide the first convincing modern proof of the probable existence of God? At first glance it seems more than unlikely. Bostrom—best known for his notorious theory that the world exists only on a giant computer—isn’t a notably religious man. What’s more, philosophers and theologians have argued for thousands of years whether God exists; whether the existence of God can be proven; and whether demonstrating proof of God’s existence is something we should even try to pursue. Despite all this, in the year 2003, when Bostrom published a new theory detailing the strong probability that God does in fact exist, nobody noticed (except David Pearce).
Randolph Bourne died 100 years ago this December at the age of 32. While his legacy lives on, to properly pay homage to his work we can recover the spirit of his prophetic philosophy, which has too often been overlooked or misapprehended.
Chaucer’s philosophical exploration of human nature takes a dark turn in “The Pardoner’s Tale,” a greedy man’s proud confession of his own corruption.