By Daniel Ng
“AI will be the best or worst thing ever for humanity.” - Elon Musk
Introduction
Throughout history, we have seen how technology serves as the vessel carrying the ambitions for automation into reality. As a natural consequence of advancing civilizations, the growth in knowledge and accessibility to resources open a new gateway of opportunities for developing technology. One of the general goals that technology seems to be embedded in is convenience or more specifically automation. For example, before airplanes, cars, and trains, the vast majority of humanity traveled by foot or on horse. This tedious process to get from point A to point B often took days or even weeks.
The invention of vehicles has allowed humans to travel long distances without stopping while limiting human input (effort). Artificial intelligence can make the process of getting from point A to B with even less effort through self-driving. This is only one of the many applications of AI that humans have developed throughout the past decade. It is only a matter of time before artificial intelligence will consume the transportation, manufacturing, healthcare, education, media, and customer service industries. Developing an understanding of AI may not only open up a world of interests and possibilities for you but also change the way you perceive most things.
Artificial intelligence (AI) is a type of technology that allows machines to replicate human intelligence through iterative processing, learning from experience, and algorithmic training. In essence, AI makes machines intelligent.
Concepts + Forms
There are four definitions of what an AI can do: (1) thinking humanly, thinking rationally, acting humanly, and acting rationally. Emerging from these definitions is the categorization of AI. (1) Artificial Narrow Intelligence, (2) Artificial intelligence, and (3) Artificial Super Intelligence.
Artificial Narrow Intelligence (ANI) is the most basic category of AI. This form of AI replicates human intelligence to perform singular tasks. An example of this is the AI that exists on most smartphones, like Siri.
Artificial General Intelligence (AGI) is a category of AI used for a diverse range of tasks. Its objective is extensive which is why it can improve itself by learning, similar to the capabilities of the human brain.
Artificial Super Intelligence (ASI) is the most superlative form out of the three, hence the label “super”. This type of AI is more sophisticated and advanced than any other as it can contemplate abstract concepts that transcend human capacity.
The three forms of AI listed above are defined similarly but are labeled differently across the world. Another definition for the terms above proposed by Michigan State University’s Arend Hintze in similar order is reactive machines, limited memory, and theory of mind.
Now that we know the different forms of AI, it begs a few questions about the actual creation of AI.
Programming Languages
AI is programmed on the basis of five major programming languages: C++, Java, Lisp, Python, and Prolog. C++ is loaded with many library functions and tools in programming that could allow AI systems to thrive. Java is a highly versatile language that can be implemented and traversed into other platforms. Lisp is efficient and fast as it has an automatic memory manager. Python is faster than Java or C++ as it not only has a variety of tools and libraries that assist you in coding but a testing ability for algorithms. Depending on the language used, breadth of complexity, and purpose of the AI, proficiency in various other mathematics-related subjects such as multi-variable calculus, matrix algebra, and statistics are required.
Subfields: AI Tech
When it comes to the characteristics of AI that make it so unique and powerful, the following five are what give AI its properties.
Neural networks that are interconnected allow machines to learn by processing data from external sources
Machine learning utilizes various neural networks and additional information to learn from without being programmed to make conclusions
Cognitive computing allows machines to simulate human-like processes by interpreting input such as images and diction
Natural language processing is the ability of a machine to comprehend, analyze and even reproduce human speech
Deep learning is a superior subset of machine learning that allows AI to process substantial amounts of information and identify patterns, allowing functions such as speech recognition in AI.
The application of these properties allows AI to replicate human processes such as learning, decision-making, interpreting, analyzing, and speaking more efficiently. This opens a highway of possibilities for technology. An example out of many is being able to use robots to work in scenarios that would otherwise be dangerous for humans.
What can it do? What it can’t do … yet.
Now that we get the gist of how AI works, you might be wondering what can they actually do, or what the limitations of such a powerful system are.
Can: Self-driving, voice assistant, digital marketing, produce digital art, writing assistant, etc.
Can’t: Generate long-form texts, compose songs, use common sense, adapt/improvise, etc.
Ethical implications
Unemployment - already a large issue at hand, AI might contribute to making it worse (except for more jobs for programmers)
Inequality - the wealth generated by machines, how to distribute
Humanity - the potential effects of replicating or imitating human behavior and interaction with human beings
Bias - some argue that although AI itself can be objective, the programmer has biases that can be programmed into AI
Security - powerful technology can be used for criminal purposes as well
Singularity - how do we stay in control
Robot rights - if robots become self-conscious or acquire other basic characteristics of humans, should they also have rights like humans? (they might have feelings and emotions)
Conclusion
Artificial intelligence algorithms, unlike passive machines that are only capable of predetermined or mechanical responses, are able to utilize real-time data to make well-considered decisions, similar to how a human brain detects patterns to learn and make decisions. However, artificial intelligence is not limited by cognitive capacity, ingrained cognitive biases, the ability to process complex information, and numerous other factors restricted by human biology. The potential of artificial intelligence to benefit humanity is revolutionary. Learning a bit about it or possibly viewing it as a career to pursue might be something to consider.
References
Schroer, A. (2022, September 19). What Is Artificial Intelligence (AI)? How Does AI Work? | Built In. Builtin.com. https://builtin.com/artificial-intelligence
admin. (2019, December 17). How Does AI Work. Digital Silk. https://www.digitalsilk.com/how-does-ai-work#Artificial-Intelligence-4-Concepts-of-Artificial-Intelligence
How Does AI Actually Work? (2021, August 9). Colorado State University Global. https://csuglobal.edu/blog/how-does-ai-actually-work
Artificial Intelligence (AI): What it is and why it matters. (2022). Sas.com. https://www.sas.com/en_us/insights/analytics/what-is-artificial-intelligence.html#technical
Kumari, R. (2021). 7 Differences between Artificial Intelligence and Human Intelligence | Analytics Steps. AnalyticsSteps. https://www.analyticssteps.com/blogs/7-differences-artificial-intelligence-ai-human-intelligence
phrasee. (2018, November 27). 6 things AI can do, and 5 things it can’t (yet) - Phrasee. Phrasee. https://phrasee.co/blog/6-things-ai-can-do-and-5-things-it-cant-yet/
Bossmann, J. (2016, October 21). Top 9 ethical issues in artificial intelligence. World Economic Forum. https://www.weforum.org/agenda/2016/10/top-10-ethical-issues-in-artificial-intelligence/
Comments