The term "cutting-edge" is often used to describe disruptive technologies as well as the latest technological advancements. In this post we will touch on a few cutting-edge technologies for example Crypto and AI to name a few.
Crypto and Blockchain
What is it?
Before we get into crypto it is worth having a little historical context. Physical monies that we use to buy things on a daily basis are referred to as fiat currency. Currencies such as the GB Pound, Euro, US Dollar and the Japanese Yen all fall under this category. These monies are monitored, controlled and issued by a centralised governing body. The government can control the supply of money circulation in response to economic fluctuations.
In complete contrast however, crypto is a digital currency which is not controlled by a central governing entity. The supply and movement of this type of money is transparently audited on a blockchain, which is a technical solution that combines two technologies that have been established for many years:
decentralised computing via peer-to-peer networks; and more importantly...
SHA-256 encryption, which is where the term crypto comes from.
How does it work?
We need to understand the process of blockchain and how it successfully eliminated the need of a central bureaucratic entity to monitor the supply and movement of money. I have come to understand that this topic overlaps with another subject you take at school and we have borrowed one of their sources which summarises very well what you need to know for your IGCSE.
Hereunder are the key takeaways from the video:
The blockchain is meant to serve as a transparent audit log pertaining to the movement of a digital token, or a public ledger, that can be read by anyone on the network.
A single entry log that makes it to the blockchain is practically immutable, meaning, it is too difficult to change! This is because a block that stores a single entry log (data) also contains a unique hash, and the hash total that references previous blocks, hence, the term chain.
A hash is an encrypted version of all the data in the block therefore trying to change the data will also change the hash. Such a suspicious and malicious act will break the chain because newer blocks will be rendered invalid. Nobody on the block chain wants this to happen. This is what makes the blockchain so secure and reliable.
For the previous point to have any meaning, you must appreciate that creating a block on the blockchain takes considerable effort, which is formally known as proof-of-work. The proof-of-work is a relatively difficult task, like a very complex mathematical problem that takes around 10 minutes for the computer to solve. After that, more than half of the blockchain need to verify that the proof-of-work was done correctly and that no other blocks have been tampered with, which we call consensus.
Why do we need it?
Two entities that do not know each other do not trust each other and the eventual gain of that trust is a very long and complicated process. Establishing trust on the blockchain is less complicated and does not suffer from human bias.
Won't the world be a better place if we have nothing to worry about and just trust one another? What do you think?
Robot Automation
What is it?
Robotics refers to the study of robots and their design, construction, and application.
A robot is an artificially intelligent machine that is capable of performing a wide range of tasks, including movement, sensing, decision making, and communication.
How does it work?
Robots typically consist of a number of different components, including a controller (such as a microprocessor or computer), actuators (which allow the robot to move), sensors (which allow the robot to perceive its environment), and power sources.
Robots can be programmed to perform specific tasks, and they may use algorithms and mathematical models to make decisions and control their movements.
Why do we use it?
Robots are used in a variety of different applications, including manufacturing, healthcare, space exploration, and more.
The main advantages of using robots include increased efficiency, precision, and safety. Robots can also perform tasks that are too dangerous or repetitive for humans to do.
However, there are also some disadvantages to using robots, such as their high cost, the need for specialized training to operate them, and the risk of job loss for human workers.
Artificial Intelligence
What is it?
Artificial Intelligence means differently depending on how and why it is being used. At your level you have probably been exposed to A.I. when:
making use of biometric unlocking on your mobile phone;
when looking at the 'For You' feed on TikTok; and most recently...
when experimenting with ChatGPT as you are preparing your assignments or doing your homework.
From the perspective of the application user, A.I. is the simulation of intelligent behaviour by the machine so that the experience is personalised, more intuitive and feels more natural or fluid... almost as though the machine is trying to copy the thinking abilities of a human which is not rigid, but sensitive to a particular context.
How does it work?
Answering this is difficult for me because it is almost like asking me, "how does music work?" This branch of computing is very broad and people tend to be really good in one area only, just like a musician tends to be an expert in one genre or instrument. For example, my speciality is deep learning, computer vision. I will be explaining the context of the following areas that you need to be aware of which are gaining quite a lot of attention at the moment:
Machine Learning There are tree main categories but we will only make reference to supervised machine learning. This is an iterative approach whereby a mathematical learning model trains or learns on a sample of labelled data that has to do with a single skill, for example, spam filtering, or predicting the next word in a conversation, as is the case with ChatGPT. The labelled data will strongly suggest that emails containing certain phrases like "you won a prize", "please help", "we need your password" should be categorised as spam, while other emails that seem genuine like "hey there", "as per our conversation the other day", "do we have any homework" should not be categorised as spam. When a new email comes in, a sample is fed into the model and it will try to predict whether it is spam or not based on what was learned. A learning model can be trained many times on different samples of labelled data so that over time the quality of the predictions get better.
Expert Systems Traditionally, chat bots are expert systems that offer a very sophisticated natural-language user interface. Whilst ChatGPT is a breakthrough, researchers are still building sophisticated expert systems that can accurately have a conversation about a specialised topic. An expert system stores a very large set of facts in a particular area, for example, Plants which are normally selected by human experts. A user would normally interact with an expert system by inputting some brief facts related to the topic. By means of a rules engine the system will collect these facts and match an object in the form of a response, as though it is trying to mimic the reasoning of a human expert, as best as possible. If you perhaps ever played the popular game Guess Who then it is a good analogy except that the decision making and logical reasoning is done by the computer (A.I.).
Why do we use it?
You can think of A.I. as a search algorithm whereby the algorithm is searching for an answer to a problem which is too difficult to formalise using simple computational logic. These problems are difficult to solve, but very easy to understand. Imagine that I asked you to "guess" my 4 digit pin on my phone without imposing any limit on the number of tries. How would you go about it? You are probably thinking of a brute-force search algorithm which is indeed correct. There are 9,999 combinations to search through, so our search space is so big. It will likely take up to 30 hours to find the right combination because verifying if a particular solution is correct requires little effort. If the digit pin grows up to 6 instead of 4 then the search space got significantly bigger therefore searching through it would take us significantly longer.
You will be surprised to learn that many problems have an extremely BIG search space of solutions, therefore not even the most powerful computer can go through them all, but, verifying whether a solution is good or not is almost trivial. If you are thinking of a problem that has these characteristics then it is probably a good contender for A.I.
Comments