Debunking the Artificial Intelligence SkyNet Fallacies
There is a fear among many people that AI Systems will become so much smarter than humans, take all of our jobs, and humans may even lose their dominance at the top of the food chain.
Their worry seems to originate from experts who are asking themselves the same question— e.g. Elon Musk, Max Tegmark, Sam Harris, and Bill Gates.
Underlying the scenario of an AI world takeover are five misconceptions that are not based in evidence.
These claims could be true in the future, but there is no evidence today to support these five misconceptions.
The misconceptions are:
- Artificial intelligence is already getting smarter than us, at an exponential rate.
- AIs can gain general-purpose intelligence, like our own.
- Human intelligence can be built on silicon.
- Intelligence can be expanded without limit.
- Once we have synthesized super-intelligence, it will be able to solve most of our problems.
The facts and evidence support the following:
- Intelligence is not a single dimension, so “smarter than humans” is a meaningless concept.
- Humans do not have general purpose minds, and neither will AIs.
- Emulation of human thinking in other media will have to be constrained by cost.
- Dimensions of intelligence are not infinite.
- Intelligences are only one factor in being human and we are more than just intelligence.
Debunking Misconception #1
The misconception that intelligence is a single dimension is flawed.
Every species alive today is equally as evolved as human. Our evolution exists alongside cockroaches, bears, ferns, foxes, and bacteria. Every species on earth has undergone an unbroken chain of three billion years of successful reproduction, which means that bacteria and cockroaches today are as highly evolved as humans. There is no ladder of evolution from crude to advanced; we all just evolved.
Intelligence is not a single dimension. It is a complex of many types and modes of cognition, each one a continuum.
Zoology is full of remarkable differences in how animals think. There is no measurement, no single metric of intelligence.There are many metrics for many different types of cognition.
Each species has both a unique and similar set of cognitions These suits of intelligences vary between individuals and between species.
For example, a squirrel can remember the exact location of several thousand acorns for years. Squirrels exceed humans in this one area of cognition. In other words, there are many specific feats of cognition in the animal kingdom that are superior to humans.
AI systems already exceed humans on certain dimensions. Some of these modes are things we can do, but AI can do much better, such as probability or math calculations.
Others are type of thinking humans can’t do but AI excels at — for example, memorize every word on six billion web pages.
AI will invent new modes of cognition that don’t exist in humans and don’t exist anywhere in biology.
Debunking Misconception #2
There is a misconception that AI will takeover the world.
The evidence suggests that AIs most likely won’t be superhuman. AIs will have many hundreds of new extra-human modes of thinking, most different from humans and none that will be general purpose.
Think of this as a universe of finite intelligences, working in unfamiliar dimensions, exceeding our thinking in many of them, working together with us in time to solve existing problems and create new problems.
Debunking Misconception #3
The belief that intelligence has an infinite scale is flawed.
There is no other physical dimension in the universe that is infinite, as far as science has discovered today.
This misconception is based on the belief that the limit of intelligence is significantly beyond what humans are capable of.
We imagine an “intelligence explosion,” however it is much more like a scattering of new varieties of intelligence that are brought into existence.
Accelerating AI technology will most likely not be super-human, but extra-human-- outside of our human experience, but not necessarily “above” it.
Debunking Misconception #4
The misconception that humans have general purpose intelligence is not accurate.
If we view intelligence as providing a large possibility space, there is no general-purpose state.
Human intelligence is not in a central position, with other specialized intelligence revolving around it.
Human intelligence is a very, very specific type of intelligence that has evolved over many millions of years to enable our species to survive on this planet.
AI is already omni-present on our planet and will continue to proliferate, deepen, diversify, and augment.
By the end of the century, AI will touch and remake everything in our lives.
Humans will co-exist with AIs. Each will have a place in our world.
AI and humans will both have unique intelligences; just like all lifeforms do today.