When you hear artificial intelligence (AI), does your mind think of the infamous Terminator movie series or another sci-fi about AI takeover or dominance? With movies like the Matrix series, "Space Jam," "Tron: Legacy," "A-X-L," "I Am Mother," and even "Avengers: Age of Ultron," it’s no wonder that when we hear the term AI, our mind drifts toward Hollywood movies.
However, Hollywood does a hatchet job portraying AI. As of now, AI is used as a fancy word for a computer that can perform pattern recognition. Basically, it is trained to use the information and distinguish patterns from it. This may sound incredibly complicated, but we interact with AI all the time. Smartphones, smartwatches, even smartTVs and smartblinds, typically have some sort of in-built AI.
For us to be able to talk about AI and an impending AI takeover, it’s important that we discuss the different types of AI, starting with the one we interact with the most often.
Artificial narrow intelligence, known as ANI or narrow AI, is the most common type. It is an AI that is created to focus on one type of function or skill set, this would include AI’s such as Apple’s Siri, Samsung’s Bixby, iRobot’s Roomba, Facebook’s (now named as Meta) targeted news feed, or Google’s prudently named Google Assistant.
Artificial general intelligence, or AGI in short, is the next step in the AI chain. It is an AI that can actually "think" on a level similar to what humans can do. The AGI would have fine motor skills (coordination), natural language processing and understanding (can understand and communicate), as well as problem-solving. We are still in the process of taking the leap from ANI to AGI, and many experts debate when this jump happens.
Artificial superintelligence, or ASI, is the "Terminator" level of AI. It refers to the AI that can outperform humans in all aspects. This is the AI that we fear, the one we saw in the Terminator movies, and cannot beat.
If we survive long enough to continue down the path we are heading, we might be seeing AI take over the world within our lifetime. Since the early days of AI in the 1950s, we have been creating better and more sophisticated AIs, and as long as companies and governments see potential in AI, this trend toward more sophisticated AI will continue at an exponential rate. It was only 11 years ago when Siri (now Apple’s voice assistant) was created. Just last year AI flawlessly defeated a human fighter pilot in a simulated combat. It is gradually becoming smarter and the field of AI is flooded with more and more researchers studying new AI with more and more companies funding new research. Some experts say there is a 50% chance before 2040, but almost all experts agree that we will see AGI by the end of this century.
When we think about AI’s impending destruction of humanity, scenarios of crazy robots shooting at innocent humans fills our mind, but that’s not going to be the destruction caused by AI.
We need to be scared of the economic impacts. Why? Because AI can do jobs that humans can, much more effectively and efficiently, significantly disrupting the economy. A recent AI by Defense Advanced Research Project Agency (DARPA) ended up beating a seasoned United States Air Force pilot in five matches-flawlessly. Even then, many people believe jobs like computer scientists and doctors are here to stay.
U.S.-based software developer Github’s Copilot project, an AI coding system, writes lines of code in whatever language your heart desires (within reason of course) with just a simple description; John Hopkins University created Smart Tissue Autonomous Robot (STAR) that successfully did, "one of the most intricate and delicate tasks in surgery," and still managed to perform better than most human doctors doing the same procedure. Even if we don’t "evolve" ANI to AGI, we still have a lot to fear. These robots don’t need health care, sleep" or breaks, they can work 24/7, they’re easily replaceable and they’re becoming significantly more cost-effective at what they do.
We like to live in a fantasy world where Terminator level AI is centuries away and that we don’t have to worry about major AI disruptions, but even one of the most simple types of AI will be able to replace a significant portion of an ever-increasing workforce. For many, it sounds scary to think it is likely that AI in the next decade will replace most of the workforce on earth, but it’s coming, even though it might not be this decade. We need to start preparing for it. Now.