Many of us reading may have heard about AI. This could be through friends and family but also social media or your own research. However, there is one particular AI that is taking the world by storm. From young teens on TikTok to old grandparents on Facebook, everyone is fascinated by ChatGPT.
Now there is a chance that you may have never heard about this so let me introduce it to you. ChatGPT is an AI model developed by OpenAI. They are also known for producing Dall-E which is another AI that can create and develop any picture you want using a prompt. ChatGPT works when the user enters a prompt on what it wants the AI to do and in my case, it works 99% of the time. You can ask it to do anything from writing essays, emails and computer code to having a conversation with you as a baby.
I, myself get confused and ask “How is this AI able to do anything and get it right every time?” Well, I did some research and this is what I learned: ChatGPT is a type of neural network that is essentially a computer system based on the human brain and nervous system. More specifically, the type of neural network they use is based on transformer architecture which is modelled on the concept of self-attention and works by weighing up the significance of different words in the input to generate the output. In more simple terms, the architecture looks at the relationship between words and seeing which ones are important and how they are related to each other to generate an output that makes sense. Afterwards, the AI is fed a large amount of text data through unsupervised learning where the data is unlabelled and the AI has to learn from it without any guidance. This allows the AI to learn about the relationship between a colossal amount of words and phrases which allows the AI to generate an output that makes sense from most prompts. This makes the AI very knowledgeable and can provide an answer from any context. For me, this can most likely be another alternative to Google as it can provide you with the right answers 99% of the time which is easy to understand and can explain complex topics with ease.
This technology is being used in various ways in today’s time. Some examples include AI chatbots which act as virtual assistants by giving out information and taking questions from customers on behalf of firms. This has many benefits such as eliminating extra costs in terms of employee wages as this AI removes the need for human’s working as an assistant. Also, help can be provided 24/7 on demand without a delay to millions of people. Another great example is in healthcare where ChatGPT can analyse data from patient’s test results and symptoms to help doctors diagnose people and provide treatment recommendations. This can be very useful in developing countries where doctors may not be as skilled to identify and diagnose every health problem and provide the correct treatment. ChatGPT is also popular for content creation in terms of generating news articles and social media posts where it can generate information relating to the topic and try to maximise engagement. Other main examples include language translation, personal assistants, education, creative writing and many more. However, these are only a really small amount of uses of ChatGPT. The uses are unlimited if you use your imagination and be creative.
ChatGPT regularly gets updates that make it even more reliable and smart such as the new GPT-4 model which can pass some of the hardest college entrance exams with a high mark. In the future, AI will just be getting better and better which will give us more uses and allow us to solve a larger number of complex solutions. For example, ChatGPT could be heavily personalised which can help us greatly in our life such as by making recommendations for us. These recommendations could be in the form of anything from what movie we should watch, to what clothes we should wear, where we should eat, et cetera. This could help us make decisions in our life which will save us a great amount of time. Another way ChatGPT could be applied to benefit a vast amount of people in the future would be by helping to solve mental health problems. This could be done by diagnosing patients with their conversations with AI and providing recommendations for treatment such as therapy and medication. Some other future applications include fraud detection by analysing suspicious behaviour, shopping assistants which inform customers which goods they should buy, financial planning in terms of one’s financial situation and much more. These applications and some others could be very rewarding in terms of saving thousands of lives. However, there is still a lot of time for us to reach a point where these will work effortlessly without any problems.
However, ChatGPT and other AI can pose a lot of new problems currently and in the future. One main argument is that AI will lead to a loss of jobs; Each firm has a profit incentive and one of the best ways to do that is having the least amount of employees as possible as this will heavily reduce the total costs of the business. Having AI do the work will remove the need for employees as they will be able to complete the same jobs in a quicker amount of time in most cases. One example I can think of is software engineering. Nowadays, ChatGPT can code anything you want in any coding language. You can type a prompt asking ‘Create a Python script which takes any input and tells you the amount of characters it has’ and it will do that for you in less than a minute. ChatGPT can code complex scenarios which essentially removes the need for software engineers. The problem with this is that many people will lose their jobs and their degrees and skills will become worthless.
Another big problem that ChatGPT has is cheating in school. AI has become a global phenomenon among students around the world where they use it to do their homework and other projects but also in exams. Students can ask ChatGPT to write an essay for them and the AI can write one which can get the student full marks in no time saving students hours of work and more time to relax. This can have many repercussions. Students may not learn as much as they should and may fall behind on knowledge. Also, if students aren’t learning skills by themselves and are instead asking ChatGPT, they won’t be able to do specific tasks to a good enough standard without any help.
In conclusion, ChatGPT can be very useful in our daily lives now and help us with particular problems and in the future, it can probably change our lives in a good way and be very beneficial. However, new problems are being created by it which can be harmful to society as a whole which means we should have some control over AI so it doesn’t get too powerful even though some may argue it is smarter than mankind already. There is no need to worry too much though as humans will always find solutions to their problems as they have since their existence.