print("Hello AI")
Artificial Intelligence (AI) is an area of computer science focused on enabling computers and machines to perform human-like tasks and simulate human behavior.
In a blunt sentence, AI aims to make computers equal to or greater than humans.
We can classify AI into different types based on their capabilities, functionalities, and theoretical versus realized implementations.
Based on the capabilities we have:
Narrow AI: This is realized AI, which encompasses all the AI currently available. Also known as weak AI, it is called 'narrow' because it specializes in specific types of tasks. For example, ChatGPT specializes in general knowledge as a chatbot, and Spot the Robot from Boston Dynamics specializes in motion and control.
Artificial General Intelligence (AGI): This is theoretical AI, also known as strong AI, which is envisioned to specialize in a wide range of tasks and make autonomous decisions. For instance, if AGI (Artificial General Intelligence) needs to learn a new task, it would be able to figure it out by itself. This represents the next frontier for AI, with companies like Google, through DeepMind, and Tesla, with their humanoid project, already making significant strides in this direction.
Super AI: This class of AI, also known as Strong AI, remains theoretical. In addition to a wide range of specializations more than AGI, this type of AI would be able to have emotions, beliefs, and desires of their own. This is supposed to be self-aware, conscious, and sentient. This is supposed to be the final stage of AI where they become equal to or better than humans.
Based on functionalities we have:
Reactive Machine AI: This is realized AI and under Narrow AI. The systems are designed to perform specific tasks using statistics. An example is IBM's Deep Blue which played chess.
Limited Memory AI: This is also realized AI and under Narrow AI. This form of AI can recall past events, predict outcomes, and monitor specific situations and it can improve performance. An example of this is ChatGPT.
Theory of Mind AI: This is theoretical AI and is under AGI. This form of AI would have the ability to understand emotions, feelings, and human interactions.
Self-Aware AI: This is theoretical AI and under Super AI. It would be able to understand itself and its conditions. And make its own set of emotions and beliefs. To be sentient, self-aware.
So now you can see why some people are scared of AI. I would be too when AI becomes sentient, that'll be some Terminator moves lol. But for now, that's a long shot or not... we'll see.
Just over a year ago a Google engineer claimed the AI technology Google is working on has become sentient, you can check him here. I am more scared of what AI could be used for in the wrong hands, it could be like George Orwell's 1984 Big Brother.
Enough of doomsday stories, let's see what our opinion will be by the end of this series.
AI to me is an assistant, that's the same thing Jack Dalhgreen the Project and Design manager at Nvidia said to WSJ in an interview recently.
Viewing AI as an assistant reveals its potential as a tool for leverage rather than a threat. Need a robot for high-risk tasks in your mining company? AI. Want to practice for an interview with ChatGPT? AI. The possibilities are vast. While not everyone can afford a personal human assistant, AI can provide similar benefits, offering opportunities that were once out of reach.
The history of AI dates back centuries, it's been through incredible incremental growth to bring us up to this point. While I won't talk about the history of AI in this series, I'll leave a video for you to check out here.
The future of AI is still yet to be seen. Here is an interesting, hour-long documentary by 60 Minutes of CBS about the current state and future of AI where frontiers in the game like Dr. Kai-Fu Lee of Sinovation Ventures and Sundar Pichai of Google tell us about what is and what to expect.
AI is used in a lot of industries and more applications are found in new industries daily. It helps us do things better, more efficiently, safer, and at scale.
We can see this with precision robots in industries like medicine, manufacturing, and construction, we can see it in marketing with social media platforms, we can see it with our smartphone assistants, chatbots like ChatGPT and the list goes on and on.
If your job is something you think AI can do, it probably will, eventually. Maybe even better than you. So it's better to know where you should use AI to upend and augment your job as appropriate instead of fighting it. It's a losing battle.
AI has several subsets or categories, and new ones are being created as I write this. For this series, we'll group them into 5 subsets:
Machine Learning (ML)
Computer Vision (CV)
Speech Recognition (ASR)
Natural Language Processing (NLP)
Motion and Control
These subsets often overlap with each other as we'll see later in the series.
You can check the series mind map here to see the breakdown.
If you have taken some of these courses in school like me, still read along I'll refresh your memory, and many things have changed too. We can also discuss further in the comments.
We'll take them one after the other. I'll keep things simple.
Now to be as good or better than humans the computer should be able to first learn, so up next; Introduction to Machine Learning.