If you've read my other blogs, you must know I love to start off with some history. Recently, I've started taking a liking for two very new subjects. History and Economics. I've come to the realisation that one cannot appreciate the present without knowing the past. With History, you understand the decisions our ancestors took decided today's way of life, and that has helped me answer a lot of questions I have about the present. And with Economics, you understand how to prepare for the future.

Coming back to the topic of AI, ultimately AI and Machine Learning are built on the foundations of mathematical concepts. It's just that with today's computing power, we are able to use huge amounts of data to derive complex mathematical models that are nearly impossible to represent in traditional mathematical notations. So my point being, the foundations of AI has been laid almost a century ago by mathematicians. Let me share with you some key developments in this field which helped us reach where we are today with AI tools.

1. 1950s-1970s: The first mathematical models of artificial neurons were developed.
2. 1970s-1990s: Early "rule based" systems, and later the introduction of the concept of backpropagation
3. 1990s-2010s: Rise of machine learning driven by statistical models, data driven applications, text and image processing
4. 2010s-2020s: This is the current era where AI boomed, driven by the ready availability and abundance of data and extreme developments in computing technology.

I'm sure there won't be many engineers who hasn't heard of the Moore's law. But for the sake of my occasional non-tech readers, let me briefly describe, what I would call, the single most important principle that drives technology forward.

Moore's law began as an observation made by Gordon Moore in 1965, that the number of transistors on a microchip doubles approximately every two years, while the cost per transistor decreases. Now you may think, why I would give so much credit to a rather simple "observation" made decades ago. Moore's law is so fascinating, because of it's accuracy even after 50+ years. If you look at the below chart of transistors per microchip, it is still very much following a trend someone predicted half a century ago. And, it's this trend in growth of computer chips that made it possible for humans to build supercomputers capable of training models like GPT and llama.

Yes, and no. The way I see it, AI is going to eliminate a lot of incompetent people from the market. The whole concept of GenAI is to repeat things which it has seen previously, in the context of the current prompt. This approach is only good enough to do trivial, redundant tasks which has already been done by many other people. So, my dear friend, if you are incompetent, or don't keep youself upskilled, then within then next 4 years (2029) you can probably be replaced by AI, or someone who knows how to use it to his advantage.

On the other hand, AI is an invaluable tool for inquisitive people who try to keep themselves upskilled. GenAI tools have made learning easier than never before. It's like having your personal tutor in your pocket at all times, who can feed you high quality data in simple language and answer all your questions. And that is exactly what AI is good at. It is good at teaching you theory from the books in simple language. You just need to know how to ask. Like the bible says "Ask, and it shall be given to you".

The only way to prevent AI from mastering you is to master it first. Understand what AI can do for you, and use it to your advantage. Here are some tools you can start off with for free :

1. ChatGPT: Nowadays I learn new things by going to ChatGPT and asking it questions on the topic. Once I have a basic understanding, I browse the web and find articles to read. I have found this method really effective.
2. ClaudeAI:As a software engineer I often come across the need for quick prototyping. You can give screenshots to Claude and ask it to build your app.
3. Gemini: Not as good as ChatGPT in my experience, but when ChatGPT runs out of free prompts, I go to Gemini.