Since I came from a creative background, I felt that I needed to understand what was going on in the world with artificial intelligence, so I decided to start learning on my own and share what I had learned so far. Let us get started.
What is AI?
Artificial intelligence, or AI, is technology that enables computers and machines to simulate human intelligence and problem-solving capabilities.
On its own or combined with other technologies (e.g., sensors, geolocation, robotics), AI can perform tasks that would otherwise require human intelligence or intervention. Digital assistants, GPS guidance, autonomous vehicles, and generative AI tools (like OpenAI’s Chat GPT) are just a few examples of AI in the daily news and our daily lives.
Demystifying AI
ANI — Artificial narrow intelligence (eg: smart speaker, self-driving car, web search, Ai in factories)
Generative AI — Generative artificial intelligence (eg: ChatGPT, Bard, Midjourney, DALL — E)
AGI — Artificial general intelligence (Do anything a human can do)
What is machine learning?
Machine learning is one AI tool that has been crucial to the field’s growth.
The most commonly used type of machine learning is a type of AI that learns A to B or input to output mappings, and this is called supervised learning.
Supervised learning is a machine learning technique that involves training a model to generate predictions using labeled data. The model is trained with input data and the matching correct output, and it learns to predict the output for new data based on the training. This makes it possible for the model to carry out operations like regression and classification.
To illustrate, imagine we aim to train a computer to identify various breeds of dogs. We expose it to a variety of dog images along with their respective breed names. Afterwards, the computer attempts to identify the typical characteristics linked to each breed.
Once the computer gains sufficient knowledge about different dog breeds, we can evaluate its proficiency by presenting it with unfamiliar dog images. Using what it has learned, the computer makes an effort to recognize the breed that is shown. A successful recognition indicates how well the computer has learned about dogs. In cases of incorrect guesses, we can provide more examples to improve the learning process.
What is data?
At their core, AI systems learn and make decisions by analyzing vast amounts of data. Based on the data they have been exposed to, these systems are intended to identify trends, draw conclusions, and produce forecasts or suggestions. AI would not function properly without data; it would not have the necessary basis.
Types of Data in AI
∘ Structured Data: Highly organized data with a clear format, often found in databases and spreadsheets.
∘ Unstructured Data: Data that lacks a predefined structure, such as text, images, and videos.
∘ Semi-structured Data: Falls in between structured and unstructured data, often represented in formats like JSON or XML.
∘ Temporal Data: Temporal data includes information collected over time, such as stock prices, weather measurements, or sensor readings.
∘ Spatial Data: Spatial data represents the physical location and characteristics of objects on Earth’s surface. It includes maps, satellite images, GPS coordinates, and geographic information system (GIS) data.
∘ Text Data: Text data comprises written or typed language, including articles, emails, social media posts, and more.
∘ Image Data: Image data consists of visual information captured in the form of pictures or photographs.
∘ Audio Data: Audio data includes sound recordings, such as speech, music, or environmental sounds.
Training AI Models
Data is used for training in order to create AI models that work. The AI system learns from past data during this process, seeing patterns and relationships. In natural language processing (NLP), for example, a model trained on a sizable corpus of text data can pick up on sentiment analysis, semantics, and grammar rules.
Real-Time Decision-Making
AI systems can confidently make decisions in real time when they have access to high quality data. In order to navigate and adapt to changing road conditions, self-driving cars continuously process data from sensors and cameras. Similarly, AI algorithms are used in finance to make split-second trading decisions by analyzing market data.
The Quality of Data Matters
The quality of data is critical in the field of artificial intelligence. Here, it is true what they say about “garbage in, garbage out.” Low-quality or biased data can lead to flawed AI models and inaccurate predictions. For AI systems to be reliable, the data they use needs to be clear, impartial, and representative.
Data Privacy and Ethics
Due to AI’s heavy reliance on data, ethical and privacy issues are brought up. Collecting, storing, and using data responsibly is essential. Laws such as GDPR are designed to safeguard individuals’ rights regarding their data and make companies responsible for how they handle data.
Acquiring Data
∘ Manual labeling
∘ From observing user behaviors
∘ Download from websites / partnerships
Data science vs. machine learning
Data science is a field that studies data and how to extract meaning from it. It extracts insights from both structured and unstructured data using a variety of techniques, algorithms, systems, and tools. This knowledge is applied to business, government, and other industries to drive profits, innovate products and services, build better infrastructure and public systems, and more.
Within artificial intelligence, machine learning is a specialized field devoted to creating models that, without explicit programming, use data to generate insights and forecasts.
Each field is good for different types of people. People who are interested in understanding data and deriving data insights from it can choose data science, while people who prefer creating models that improve performance using the data can opt for machine learning.
What machine learning can and cannot do
Machine learning tends to work well when you’re trying to learn a simple concept, such as something that you could do with less than a second of mental thought, and when there’s lots of data available. Machine learning tends to work poorly when you’re trying to learn a complex concept from small amounts of data. For example, they cannot learn from the probability that “rain clouds cause rain.”
KINOMOTO.MAG 