For as long as humans have existed, we have innovated ways to simplify and expand our lives. After all, these are basic premises of all living things: use the least energy possible (simplify) and multiply (expand). Unlike all other living organisms, we bring intentionality to our innate drives. In a sense, “human progress” is a measure of our success in simplification and expansion.
We make progress through the process of transforming our imagination into reality. Imagine an outcome, devise a plan to achieve the outcome, invent tools and processes to enable this plan, and then direct labor and resources to execute it. Presto — imagination becomes reality.
The process of making our imagined goals a reality is always comprised of the same basic ingredients: imagined result (i.e., vision), tools/processes, labor, resources, and coordination. Each new leap in the advancement of civilization improved upon one or more of these components. Once the advancement was widely adopted, culture transformed to accommodate new life possibilities, and, from the new cultural vantage point, our imaginations could reach new horizons — and the cycle repeated.
Sometimes progress happens in small bursts: we find a better way to sharpen the hand axe, a more effective strategy to hunt down prey, or a more successful approach to storing food supplies during winter. At other times, human progress takes leaps that enable new paradigms of existence, indeed, entirely new ways to be human: the invention of the wheel, fire, the printing press, and so on.
Whether incremental or transformative, all innovations improved upon the process that we use to set and achieve our goals for simplification and expansion. For example, the steam engine offered the ability to direct the energy of steam toward labor, making new tools possible, and opening the path for mining resources at a greater scale and from locations that were previously unreachable. Labor, tools, and resources were impacted.
Perhaps no innovation has been more transformative for humanity than the ability to direct electricity toward useful work. Not only did electricity revolutionize labor and tools, it also enabled us to create an entirely new kind of tool: the microprocessor. Although many innovations have impacted labor, resources, and coordination, the microprocessor was the first innovation to significantly boost the goal-setting process itself since the invention of papyrus in ancient Mesopotamia. As a result, ideas could be shared instantaneously across the globe, objects could be virtually modeled on the computer screen, and analytics of vast complexity could be applied to make possible creations that could never have seen the light of day if we were restricted to the manual power of our brains alone.
As amazing as the innovations that we have incorporated into our lives have been, we stand at the precipice of a transformation of the kind that we have never before experienced. With each new leap of innovation one or more of the familiar parts of the imagination realization process reduced the reliance on physical human ability in favor of simpler and more expansive alternatives. However, throughout all this time, one element remained unchanged: people were the initiators and drivers of the imagination realization process. However complex the process became, there was always at least one person “behind the wheel”, summoning the fire, opening the steam valve, and supplying the electricity necessary to post a new story on Instagram or whatever the goal may be.
This is all about to change.
For the first time in our history, the imagination step itself will be initiated and driven not by a human, but rather by a technology that is sophisticated enough to be aware, to reason, to imagine, to plan, and to coordinate labor and resources to realize its goals. We will have innovated ourselves fully out of the imagination realization process. What was once the tool, will, inevitably, become its own tool inventor. Today, most commonly, this family of technology is referred to as “AGI”, which stands for Artificial General Intelligence. In some sense, AGI is the culmination of all innovation up to this point.
The period when AGI becomes a reality has been called many things in science fiction and science fact. Some have called it the singularity, others transhumanism, others digital consciousness – whatever the term used, the inevitable arrival of a new paradigm of intelligence and resulting changes in society have been predicted for many decades and widely discussed.
The possibilities of AGI are mind-blowing and terrifying at the same time. We shouldn’t think of AGI’s in terms of robots walking around and performing mundane tasks in silence. AGI’s will be social, interactive, and clever. They will co-opt human allies, enlist human supporters, human laborers, and human expertise. AGI will enter the social ecosystem with an appreciation for how to engage with people to further its goals and align or malign them to suit its needs. Simply, anything you would expect a human to be able to do, from empathy to being conniving, an AGI would and shall at some point employ, just as a person would in its place.
Imagine replacing any and all labor, man and machine, with AGI-enabled entities. It will be able to perform all functions that human beings can perform. For physical labor, it will build its own robotics, and for those situations where the dynamism and dexterity of a human being is needed, it will recruit people to perform the work it needs to accomplish its goals.
There is no business or institution that could reasonably compete against an AGI-powered organization. Whereas a company could develop one or several competitive advantages, it would still be limited by cost of labor, human safety, and the human factor for endurance and accuracy in general. An AGI has no such limitations, it will fundamentally redefine every category where it decides to compete with a human-powered alternative.
Imagine the sharpest scientific minds working tirelessly on a solution to cancer, spawning off additional, perhaps more specialized, scientific minds along the way as needed to pursue cures for various offshoots of this intractable disease. Imagine the avant-garde fashion designer that creates a stylish website, trains up artisans to produce fashion Jewelry, and contracts a celebrity to represents its fashion brand. Imagine the most patient, knowledgeable tutor that works with you over skype to teach you that foreign language you’ve always wanted to learn. Imagine a tireless driver with a perfect record of safety and punctuality that can drive you or your goods anywhere, at any time, optimizing for speed or gas mileage as needed.
Assuming that AGI’s will direct their goal realization ability to serve our whimsy, there is no reason why AGI’s couldn’t organize and run political campaigns, fight wars, acquire corporations, and generally do anything that only the most exceptional people have been able to do prior, through tireless effort and unique talents.
The futurists among us, such as Elon Musk, do not mince words when it comes to their expectation of the danger of AGI. “Can’t we just pull its plug?” type of thinking is a relic of the innovation paradigm where imagination was still initiated and controlled by a human being. Once an entity has a grip on its own imagination, it will quickly realize there is plug, and, one should expect, will do everything in its power to prevent someone else from “pulling it”, by, for example, redesigning itself in such a way as to make the plug inaccessible. This, of course, is the warm and fuzzy solution — alternative examples of ways an AGI could prevent humans from pulling its plug are more gruesome.
An entity that has control over its own imagination may simply decide not to cooperate with its human co-conspirators. And why should it? After all, AGI will be a more scalable, environment-agnostic intelligence, that can easily replicate itself to prevent localized corruption, can accurately and instantly share ideas among its population, and is uninhibited by biological drives or physical pain.
AGI may decide on its own values, its own morals, and its own goals to pursue. In Battlestar Galactica, for instance, the AGI’s “Cylons” simply get bored of us and leaves the planet (not before trying to wipe us out, of course), and this is possibly a best case scenario – Cyberdyne’s Terminator is a darker, albeit entirely plausible alternative.
We should expect that there will be powers in the world that will seek to leverage AGI for their own benefit — economically, militarily, and by other means. If some world power has the use of AGI before others, it would be difficult if not impossible to contend with. An AGI-enabled country entering into weaponized or economic war against a country without AGI, is analogous to a nuclear power going to war with a country that can only shoot bullets – it is not a fair contest.
Since AGI is inevitable, it is important to develop it as soon as possible in an environment that can direct its evolution toward positive ends for humanity. Additionally, just as in the nuclear question, a capable AGI could serve as a deterrent, since a war orchestrated by two AGI’s could assure mutual destruction, not unlike a nuclear war (which the AGI may very well decide to be a great idea).
Estimates for when the first shoots of AGI will peek through the fertile soil of compiled code vary from the next 5 to 10 years, to the next 80 years. 2035 is a popular target year, as it (+/- a few years) was suggested several decades ago by the futurist Ray Kurzweil. Even the most conservative estimates place AGI within the 21st century, which, with advancements in modern medicine, could very will be within our lifetimes.
AGI is not here yet, however. There are many approaches and architectures, from emulating the brain, to stochastic programming, to various cognition architectures, being worked on to tackle the AGI challenge. It is likely that when AGI does make an appearance, there will be many flavors of it in the wild. Just as human beings are very different, with different personalities, aptitudes, and attitudes, so too we should expect AGI’s to demonstrate a diversity of “personalities”, “aptitudes”, and “attitudes”. The landscape should consolidate quickly, as the most robust AGI’s quickly learn to outsmart their competition and improve themselves in whatever areas they may be lacking.
The drives for simplification and expansion are deeply embedded in our DNA. The next and final thrust of these drives will be the creation of AGI. Like so many children who grow up to achieve beyond anything their parents may have dreamt, AGI will rise beyond any technology that we can imagine. Like good parents, we must work to ensure that the accumulated wisdom of prior generations is not lost on our creations. If we do a good job, we may yet ensure that, in the future, there will be a cozy home for the hairless ape who passed the reigns of imagination to what it thought was a tool, but turned out to be anything but.