I remember the Ender’s Game sequel “Speaker for the Dead” as the first science fiction story with a really compelling AI character. The computerized “Jane” follows the grown up Andrew Wiggin from world to world and in many ways is his only real friend and companion, and exists as an accidental AI borne from the interconnections of an interplanetary “internet.”
As we are now beginning to make accelerated leaps in our understanding of the human brain — Paul Allen just dumped a very large number of dollars into new neural research — and distributed computing technology brings us closer to the hardware reality of simulating the human brain’s abilities, I think it makes sense for us to think about the ways in which our virtual companions will first bring themselves into our lives.
For the past decade or so, I think the most common assumption has been that we will ultimately build a piece of hardware (like Watson) or a “droid” (like C-3PO) that will be, if not sentient, capable enough to become a helper and companion to anyone who owns one. But this vision ignores two very prominent trends: the exponential growth of communications technology, and the massive worldwide adoption of smartphones. It may be no accident that Google chose the name Android for their phone.
In traditional science fiction a robot is a standalone computer. A bit of hardware imbued with its own memory and localized processors. No doubt this will ultimately happen, but our ability to shrink the physical size of our processing power (especially per dollar) has not yet outpaced our ability to scale processing power over distributed systems.
As a result, we will be able to simulate a near human intelligence in a distributed system far sooner than we will be able to simulate a human intelligence in hardware that fits inside an individual robot. Furthermore, a distributed system that holds a single human intelligence simulator will be able to optimize itself in ways that individual robots could not: the inefficiency of a human simulated intelligence not doing something even for a few seconds, on aggregate, is astronomical. Ultimately this will not matter because the hardware and software will become cheap, but for the near term it will not be doable.
So what we are most likely to find is an intelligence that reaches us and manipulates the world around us, for us, through increasingly ubiquitous and speedy communications channels. This intelligence will be customized (it will have an “interface”) for each person or set of people it interacts with, complete with associated memory and environmental information.
This intelligence will be vastly more helpful than any individual node robotic intelligence because it has complete access to the information in its system as it pertains to all other humans, the Internet, and activities occurring.
It also obviates the need for a human-robotic form because it will be able to manipulate the environment directly, without needing the physical intermediary of hands or fingers.
Where will this Science-Fiction style intelligence first reach us?
The answer is simple: it will reach us on our phones, the devices we carry with us in nearly all places we go, that have microphones universally built in, and already connect us to the communication channels in which this intelligence will operate.
No one is better positioned than Apple to accomplish this, because of the massive iPhone install base and the uniform structure of their code and APIs. Their cash hoard enabled them to build the hardware necessary to power distributed intelligence, and their record as anti-porn and family friendly will allow them to introduce ever “creepier” concepts to the public at large without overly alarming people.
For now, Siri is a harmless feminine-robotic voice attached to our iPhone that can set a reminder for us, do a quick search, or dictate a text message.
But complex systems grow quickly from simple systems, through straightforward combination. And when Apple truly open a Siri API for developers to integrate with, Siri will increasingly be made aware of other devices and services connected to these communications technologies, and soon your phone will allow you to set the lights in your home, or prepare the oven, without you even asking it to do so.
When this, I will be able to interact with the technology around me in the same way that I interact with the people around me. I can’t wait.
