Demystify the explosive growth of artificial intelligence

IGOR OMILAEV | UNSPLASH

Artificial intelligence has been around, but is now expanding.

You may have used artificial intelligence without knowing.

Not many people realize they have interacted with AI in the past. However, the widespread application and use of artificial intelligence in human endeavor isn’t new. Before we talk about how long it has taken humans to get to the point where the artificial intelligence is a common name, I will not assume everyone knows what it is.

Now let’s unravel and demystify artificial intelligence as a concept.

As humans, we make decisions by tapping into our intelligence (information we keep in our memory, creative capability, adaptability, and ability to process our thoughts) to compare the situations with the stored previous experiences. When we are faced with a situation that is new to us, we ask from those who might have related experience. In a similar way, some computer systems have been trained to perform tasks typically requiring human intelligence.

When you interact with Siri in smartphones (right from when Apple’s iPhone 4 was launched in 2011) and Amazon's Alexa (a voice-controlled virtual assistant), security cameras with capabilities for facial recognition, motion detection and activity monitoring; smart thermostats and locks, as well as smart watches, you are using AI. Other applications include the use of autocorrect like Grammarly or applications that suggest spelling and words when you type on your phone, and when you use common office applications like Word.

The foundation of artificial intelligence can be traced back to the curiosity of 1940s and 1950s researchers who wanted to know whether machines could be trained to act like humans. In 1956, researchers like John McCarthy, Claude Shannon and others met at the Dartmouth Conference. It was at this conference that McCarthy coined the word artificial intelligence.

Though it was possible to program computer systems with rules and logic to carry out human-related tasks, there arose a limitation. How will machines understand human language? Remember it is a difficult task for most people to understand multiple languages. The question was not just how machines would understand but how would they understand multiple human languages.

The widespread application of artificial intelligence is now evident in commercial and personal transportation with self-driving vehicles equipped with adaptive cruise control that uses sensors to control vehicle speed and advanced driver assistance systems that helps lane navigation.

Medicine has improved due to the introduction of AI to diagnosis, imaging, and non-invasive operations. Military operations, air transportation and social media operations like chatbots are examples of artificial intelligence.

Maybe now you are curious about how it was possible for machines to learn and carry out these assistive functions. The breakthrough came in the 1990s and 2000s through a subfield of artificial intelligence known as machine learning.

Machine learning and a subset known as Deep Learning have made it possible for machines to learn and take action on their own. Machine learning involves the use of a large pool of data gathered over the years on specific areas.

The data is then used to train the computer to identify scenarios, make predictions and act without being explicitly commanded to do so. AI systems improve their actions as they receive more data about specific actions before they make decision because they are able to process and improve their capability to make more precise decision as data increases.

Computers are trained to recognize colors, shapes, tones and facial expressions using model data.

The data used in machine learning includes training data, features (attributes like age, and height), label (the target) algorithm (a set of instructions on how the machine is expected to handle data and process it for output) and evaluation. After training on large set of data, AI systems are able to make inferences from training data to make decisions on new scenarios.

The use of artificial intelligence is not without its associated problems. It is possible for humans to over-rely on AI tools, which can lead to reduced human expertise and skills, privacy issues and ethical concerns like negative use of deepfake technology.

Ayo Adenuga PhD is an information security/information technology consultant with experience in technology consulting, financial, and healthcare industries.

Comments