Artificial Intelligence is starting to turn invisible from the outside in — and vice versa. The exact effects and workings of AI technologies are becoming more challenging to perceive and comprehend for humans. Even the experts themselves don’t always fully understand how an AI system operates.
Effectively, as the impact of AI technologies increases, the more limited becomes our ability to understand their impact. What does this mean for human agency and the future of artificial intelligence?
In the near future, artificial intelligence will commonly become intangible, indistinguishable and incomprehensible for humans.
Firstly, AI doesn’t necessarily need a tangible embodiment. It can manifest itself through different mediators, such as a graphical user interface or a voice interface. Already we trust Spotify recommendations without a glance or talk to Siri and Alexa like they were summoned spirits, intelligences without a tangible form.
Secondly, AI becomes invisible by passing the Turing test, or its more relevant variants. An intelligent system that manages to simulate human-level communication, and cognitive as well as emotional abilities, can become indistinguishable from humans and, thus, the “artificiality” of its intelligence becomes imperceptible for us.
Thirdly, and most importantly, AI escapes human gaze when the details of its effects and technological dynamics go beyond human perception and understanding. We can be aware of the existence, presence and effects of intelligent systems, but we no longer fully comprehend what these systems do, how they achieve their goals or what are their definite effects.
With the current pace of AI development, even seasoned experts have a hard time keeping up.
This means that AI technologies will soon go beyond Clarke’s third law, stating that “any sufficiently advanced technology is indistinguishable from magic.” Indeed, we don’t anymore have a chance to figure out the trick — or even realize that any trick occurred in the first place.
Following this, we are able to perceive manifestations and presentations of artificial intelligence, but the intelligence itself becomes unknowable to humans through human senses. Currently there are two distinct traits in this development.
First, most algorithmic systems, as well as the latest advancements in AI technologies, are black boxes; inaccessible, unfathomable and uncontrollable to most people.
Therefore, it’s hard to perceive or assess how intelligent systems shape your life online and offline, from your latest song recommendations to your personalized insurance policy, not to mention the algorithmic stock market trading that shapes the global market economy affecting almost every aspect of modern life.
Concretely, when the actions of intelligent systems become more holistically intertwined with personal, social, cultural, political and economical systems, it becomes challenging to distinguish the exact effects or impact of the machine intelligence itself.
Second, AI technologies are becoming so complex that they are hard to understand — even for the experts designing and developing them. In his recent book, The Master Algorithm, machine learning expert Pedro Domingos points out that already back in 1950s scientists created an algorithm that could do something that humans couldn’t fully comprehend.
This development hasn’t changed its course; rather, to the contrary. With the current pace of AI development, even seasoned experts have a hard time keeping up.
Today’s various machine learning systems can already provide unexpected insights in varying fields, from personalization technologies to particle physics, from cooking recipes and outlandish game moves to crime prevention and bioengineering. Concretely, specialized systems can empower scientific discoveries in biology or help you choose the best route to your next meeting.
The border between physical and digital realities is beginning to dissolve.
The more universal, self-learning and self-adjusting an intelligent system becomes, the harder it is for humans to follow its exact dynamics. And further on, when a super-fast self-learning and self-assembling AI system starts to develop and engineer itself faster than any human ever could, it evades our intellect for good. Eventually, AI systems will be leading experts on their own behavior, predicting their own future better than any human ever could.
Hence, sophisticated AI technologies will provide legitimate and correct insights based on a chain of complex interactions that can’t be followed by any human being, even an expert. If so, can we anymore reach definite scientific conclusions or make well-informed decisions without the assistance of artificial intelligence?
As with any significant technological innovation and its mainstream adaption, artificial intelligence is evolving from an obscure curiosity to a powerful utility. Consequently, the most valuable asset of tomorrow’s world might be an intelligent system that no human can fully understand or control.
Simultaneously, AI is turning into an unprecedented cultural and technological phenomenon, affecting the way we assess and define “intelligence” itself. Regarding this, human intelligence might not be the most relevant measure for intelligence itself. The “artificial” in artificial intelligence starts to lose its meaning.
Today, human intelligence is shaping artificial intelligence and, increasingly, artificial intelligence is starting to shape human intelligence. When the impact of AI systems increases, more people need to be able to understand their workings and effects. To achieve this, we need to be able to augment human intelligence to allow us to interact with various specimen of intelligent systems in sustainable terms.
First, it’s crucial to enhance the capabilities of today’s (human) researchers, designers and engineers to keep up with the AI technologies they’re building. In addition to better practices, tools and techniques, diverse multidisciplinary teams can better understand both the workings as well as the effects of the intelligent systems.
Second, as many people as possible need to be empowered to interact with AI technologies on their own terms. New learning games are presently enabling children to learn coding and robotics, familiarizing them with intelligent systems and their possibilities. Systems like Algorithmic Angels could make “the ghost in the machine” more tangible, distinguishable and comprehensible by illustrating the effects of intelligent systems in our daily lives and enabling people to decide their personal level of interaction.
Black boxes need to be uncovered to enable co-agency and collaboration on a wider scale. “Democratizing” artificial intelligence would allow more people to create diverse ways to design and develop new approaches to intelligent systems. Just like coding or media literacy are seen as today’s essential skills, being able to comprehend and affect intelligent systems will be an essential skill for tomorrow.
In this way, AI technologies could evolve into a platform, an infrastructure similar to the Internet, that would allow people themselves to decide the way they utilize AI or contribute to its design and development. Such an AI grid, like the Internet of Things², powering various experiences and applications in different environments and industries, being open for tinkerers and specialists alike, would significantly change the way we could understand AI or interact with intelligent systems in general. Human and machine intelligence would be intertwined in unseen ways.
The border between physical and digital realities is beginning to dissolve. If and when the relationship between humans and intelligent systems gets more seamlessly intertwined, the border between human intelligence and artificial intelligence might begin to dissolve, too. The A disappears in front of the I, making the concept of artificial intelligence irrelevant and obsolete.