Artificial Intelligence, an idea to integrate human intelligence into machines to assist us in daily activities and sometimes do specialized operations using the precision that humans may be incapable of achieving due to human error rate, has seen a new dawn after the inception of Large LLMs (Large Language Models), which gave machines the capability to understand and actually learn the language which it was trained upon.
OpenAI, which is an American artificial intelligence laboratory, launched ChatGPT on 30th November 2022 and took the world by storm. People who never bat an eye towards the possibilities of AI are now interacting with an AI bot themselves in the comfort of their homes, amazed by its capabilities. It didn’t take much time as many people started using it to draft their emails, create content and sometimes even write code or even debug it.
Now, the same company that brought us the ChaptGPT (GPT-3.5), has introduced GPT-4 with even more capabilities than its predecessor. GPT-4, like its predecessor, can process text, but better, GPT-4 has the ability to solve complex or nuanced instructions reliably and creatively as compared to ChatGPT, and not only that, it also accepts images as input, which changes everything, we can insert a graph or worksheets and get a good explanation regarding it.
(P.S. Can explain the humor within memes as well! So next time when a new meme is in trend, just ask GPT-4)
But is that all? Far from it, it has the capability to even simulate the wordings in a specific manner. Do you feel ChatGPT was too monotonous with a similar tone every single time? Do you long for some good ol’ Shakspearean dialects, worry not. You can instruct GPT-4 to respond in a certain manner according to your liking and it will do so.
The collaboration of OpenAI and Microsoft has now given Microsoft the edge (not the browser) of using the most advanced AI to date to be integrated into various Microsoft products such as Office 365, and Bing AI. But the fruits are not just enjoyed by Microsoft themselves, it is also used by the business messaging platform Intercom to interact and provide support to the customers, and gain insight and help resolve them all from a click of a button, which will enable the bot to learn about the past interactions and similarly replicate it with the future customers.
ChatGPT had some concerns regarding the reliability of the answers provided and even answering questions that could be let to create some dangerous substances (Looking at you all the aspiring Walter White’s out there), GPT-4 has increased its security and reliability to not respond to such questions or limit the response for the safety of the users by bringing in the team of experts from highly specialized fields to do the validation of the responses and ban certain keywords. But it is still not perfect, there might still be some holes to exploit by the users, hence it is advisable to use such products responsibly. It can be accessible via ChatGPT plus subscription or by purchasing APIs from OpenAI as well.
Mr. Kushagra Makharia