One of the advancements I postulated in the near-future setting of Eidolon is a marked expansion and adoption of human-like robots, which are manufactured and programmed by a large (fictional) tech company named SystemOne. The name of their bots? Eidolons.
A lot of how I imagined Eidolons is based on the speed of advancement of existing tech, on how current tech is (or isn’t) adopted, and on the various pitfalls and ethical questions current tech has faced.
Let’s talk hardware first. Scientists are coming up with some amazing new materials. One of the most talked about is graphene, which is a thin layer of pure carbon. How thin? One atom thick. And graphene has a whole host of amazing properties with more still being discovered. The potential applications include major advancements in electronics and energy storage, leading not only to flexible, nearly transparent cell phones but also to powerful robot brains made of graphene circuitry and even artificial muscles made of graphene.
However, as this New Yorker article on graphene states, “The progress of a technology from the moment of discovery to transformative product is slow and meandering; the consensus among scientists is that it takes decades, even when things go well.”
Until the development of graphene’s practical uses speeds up, roboticists do have other materials and engineering methods to explore. Going back to building artificial muscles, material scientists at the University of Texas (Dallas) discovered that cheap, simple fishing wire can be “twisted into springy, energy-filled coils that are flexible [and] reusable” and “that operate with one hundred times the strength of a human muscle”.
Reading that last bit, you might think to yourself—eyes bulged—that a robot a hundred times stronger than a human sounds…well, pretty dangerous.
So let’s talk about software. The robot’s brain and the programs using it. This is what Lily, a freelance tech journalist and the protagonist of Eidolon, is all about. While better and better hardware that, with every iteration, more closely mimics human features and range of movement, the software moving a robot’s body is where some of the biggest technical barriers lie. And some of the biggest dangers, too.
As this video on “The Rise of Artificial Intelligence” from PBS Digital Studios says, “There are many things that are very easy for people to do and which have been very difficult to get computers to do. The main examples are vision, natural language—understanding and speaking—and manipulating objects.”
The video goes on to list examples of low-level AI already at work in our lives, such as Google Translate, ATMs that read handwritten checks, and the algorithms driving user-specific recommendations on sites like Amazon and YouTube. Going beyond these broad, possibly overly generous definitions of “intelligence”, getting an AI to the same level of intelligence as a human being has gone from attempting to write enough rules and code for defining the world to teaching an AI how to learn on its own and thus build itself.
One barrier to completing this major step still lies within the hardware. Computer scientists have built “machines that can calculate faster than the human brain and store more information”. But the human brain is still far more efficient at using its resources. “Even the biggest, fastest supercomputers in the world cannot match the overall processing power of the brain. And they are nowhere near as compact or energy efficient.”
Another major barrier, as the previous video explains, is figuring out whether the human neural cortex uses some sort of super-algorithm in order learn—to build itself through experience (i.e. interacting with the environment)—and whether computer scientists can emulate that super-algorithm in the same manner that airplanes emulate the physics of aerodynamics that allow birds to fly.
One way scientists could get at this super-algorithm would be to emulate a human brain and then study the emulation to understand the super-algorithm driving it. Easier said than done. Again, the computer hardware just isn’t there yet and obtaining the amount of data needed from scanning a live, human brain is a difficult problem.
In Eidolon, I explore one way a group of scientists chose to try to tackle that problem.
Besides the enormous and frequently discussed ethical questions as well as social and economic changes that robots with sufficiently robust AI would bring to the world, robots with hackable brains brings concerns of software security. And I’m not just talking about cyber criminals forcing malware into future AI.
What if the robot assistant you bought (or “hired”, depending on sentience) downloads an upgrade, and a bug in the upgrade prevents it from speaking anything but French? Or gives it the robot-equivalent of OCD? Or crashes it entirely, causing it to collapse “dead” on the floor?
But yes, let’s also talk about malware-infected or otherwise compromised machines. What if someone releases an auditory virus that infects any robot that hears it and causes them to murder every living creature in sight? What if your government forces a manufacturer to allow them a back-door into any unit’s software, allowing them to secretly spy on the humans who interact with it?
These are the kinds of questions and potential exploits that Lily concerns herself with in Eidolon, at least in her professional capacity. But she is also faced with the ethics of robots used for sex.
Yes, you may giggle if you like at the idea of sex-bots just as one might roll their eyes a little at the idea of killer robots akin to the Terminator, but the closer science gets to a world with intelligent robotic beings, the less people will snicker and the more they’ll ponder. Already, campaigns exist to limit the development of so-called “killer robots” and to call for a straight-out ban on developing sex-bots.
And we’re closer to that world than you think. Google has purchased several robotics companies as well as Artificial Intelligence start-up DeepMind. Forget the brain-dead Real Doll because a company called True Companion says it’s developing “the world’s first sex robot”. They’re already taking pre-orders for “Roxxxy” (ugh, that name), which is priced at $7,000.
I agree with this Gizmodo article about the Campaign Against Sex Robots wherein author Kaila Hale-Stern states, “Prohibition is seldom a fix.” As Hale-Stern points out, the campaign doesn’t take human sex workers and the difficulties they face into account nor the unmet needs of those with alternative sexualities or situations in which their “bodies or relationships…might benefit from a helping robot hand.”
While using the eponymous robots for sex in Eidolon isn’t illegal or prohibited, a lack of thoughtful regulation and company policy creates its own problems. And these problems are compounded by a society and culture that hasn’t grown up. Hale-Stern says at the end of her Gizmodo article that the “Campaign Against Robot Sex [sic] feels too much like an anti-porn brigade, blaming the product instead of the culture for wrongs committed.”
I'll talk more about how the intersection of tech, feminism, and porn fits into Eidolon in a future post. Thanks for reading!