Mastering CS50 Machine Learning RT Hidden
Published 4/22/2024
AI For Everyone by DeepLearning.AI taught by Andrew Ng, Founder, DeepLearning.AI & Co-Founder, Coursera
The IBM 701 computer was playing chess while Arthur Samuel was playing checkers (actually, the computer was playing checkers too, and eventually became an AI Marion Tinsley).
Week two was an introduction to machine learning algorithms and how to build a machine learning project. Speech recognition is a prime example, as we often use “Hey Siri, Alexa, Google, etc.”
Unsurprisingly, the first step in this process is to collect as much data as possible. To build a speech recognition device, one would need several recordings of different people saying the activation word, such as “Alexa.” In addition, it is also essential to collect other words like “hello” so the model can be trained.
When it comes to language and communication for systems such as speech recognition, several factors must be considered. One of which is the variation of spoken accents across different regions.
Depending on one’s location and cultural background, the accent and dialect can vary greatly, sometimes making communication more challenging. By being aware of these differences and adapting one’s research, one should expand the data collection process to include as many regional dialects as possible. This can help ensure the effective and successful operation of these systems.
We don’t know what we don’t know, and as we develop systems over time, we will continue to learn the model’s needs.
This process aims to create a set of methods used by an AI system to perform various tasks. Creating these algorithms enables machine learning to learn.
Am I saying that we are actually training AI systems to teach themselves? Do we even need humans anymore? Is this where the robots take over!?!
…maybe a little bit, but in a way that streamlines productivity for small businesses or gives us all a break.
Think about the most monotonous tasks that you do for work. Maybe you spend hours entering data from emails into an Excel spreadsheet. What if a solution allowed you to push a button, make some eggs and potatoes, and come back to find the work was not only finished but done well?
What would you do with your time? Go to Cabo? Take a nap? Lay in the grass on a sunny day sipping an iced tea? You could do it all!
These systems are not being created to replace jobs or replace people, but instead to work alongside people to make our lives significantly more manageable.
Though this AI technology is relatively new, we are already helping companies find comfort in a reality that offers time off, time with loved ones, or time to be productive elsewhere.
AI is allowing dreams to come true!