How will ChatGPT change the future of the working world? 

Introduction to AI and GPT-3

AI or Artificial Intelligence is becoming increasingly commonplace. Once upon a time it was the thing of science fiction, one only has to look at the film 2001 a Space Odyssey to see why AI is quite a scary prospect for many. But in modern technology, AI is everywhere, from smartphones to the very word processor I am writing this article on. This AI has all been what we call weak AI, AI which is focussed on one specific or “narrow” task such as image processing or text analysis. In recent weeks however, a new AI has surfaced called Chat GPT. Chat GPT is based on the GPT-3 (3rd Generation Generative Pre-Trained transformer) neural network and marks a huge leap forward in AI technology. It’s creators: OpenAI, have been able to build numerous different models on top of it (Chat GPT being one of them), capable of writing complex English, producing code and even image generation. GPT-3 is therefore arguably the first strong AI we have ever seen, capable of numerous different tasks it is almost frighteningly good at anything you ask it. It does come with its limitations however which I will discuss in this article, along with how I believe GPT-3 can be used to have a positive impact on the way we work.

GPT-3 in more detail

GPT-3 is a neural network machine learning model, it achieves its impressive accuracy by having a truly enormous training set – 570 GB [1] of internet text data to be precise. The neural network itself has 175 billion internal parameters, which is a truly staggering number – for scale, the largest neural network before GPT-3 was Microsoft’s Turing NLG model which only had 10 billion parameters [2]. To produce these parameters, GPT-3 had to undergo what’s called “learning”. The learning method for GPT-3 was unsupervised learning, this means it was not specifically told what the correct answer is, rather, it learnt how to assess its own output based on the training data provided. This is what allowed GPT-3 to continually learn whilst it’s being used. The combination of unsupervised learning and such a vast training set are what allows GPT-3 to produce such human like text.

The effect of AI and GPT-3 on the working world

So, you may now be wondering quite what this has to do with the world of work. Well, GPT-3 and specifically Chat-GPT (text generation), Codex (code generation), and DALL.E (image generation) all show the possibility to replace humans in the work environment. Chat-GPT and Codex specifically can produce text and code faster than a human can. Throughout history, the value that humans have brought to businesses has been directly related to the knowledge that they bring, specifically when that knowledge or expertise would be costly or time consuming for someone else to obtain. Now however, AI technology gives everyone access to the collective knowledge of all humanity, enabling them to do tasks which used to be inaccessible to them. The question therefore stands:

Why would we pay an expert to do a job that technology can do for us? 

Because, as good as these systems are, they are not quite as good as a human. They still make small mistakes, code often has syntax or runtime errors, and it can sometimes misunderstand the meaning of questions or give responses which don’t have the same tone that a human would. So, as impressive as GPT-3 is, it’s not going to totally upturn the working world just yet. But that does leave a question:

Can we use AI to increase productivity and efficiency at work?

I believe that we can use AI tools such as Chat-GPT and Codex to revolutionise the way in which we work. Some example use cases which could potentially utilise AI include:

  • Summarising the key points of meetings
  • Planning documents
  • Finding errors/bugs in code
  • Brainstorming ideas
  • Creating social media posts

 

By viewing AI as a tool, rather than something that can replace us, we are able to improve our productivity, and allow us to make better use of our time at work, thus, bringing more value to our businesses and clients.

In conclusion, Chat GPT and other forms of natural language processing technology are rapidly changing the working world. As AI continues to advance, it is becoming increasingly important for organisations and workers to stay adaptable and continually learn in order to thrive in this ever-evolving landscape. The advancements made by Chat GPT in AI technology are opening up new possibilities for how we communicate and collaborate, and it is important to consider the potential impacts on the workforce as these technologies continue to develop. It is crucial for individuals and organizations to stay informed and prepared for the future of work. And, for those of you still sceptical about the quality of work produced by Chat GPT, this concluding paragraph was written solely by Chat GPT. Hopefully, now you can see these tools’ potential and are just as excited as we are here at NWT to start using this technology to bring about positive change in your professional life.

[1] Understanding GPT-3 In 5 Minutes | by Alberto Romero | Medium | Towards Data Science

[2] What is GPT-3? Everything You Need to Know (techtarget.com)

Together, Anexinet and NWT are uniquely positioned to help clients streamline their journey to the Cloud in the face of the pandemic by designing, building, automating and managing their workloads and applications on Enterprise-Cloud or Cloud-Hyperscalers, including AWS, Alibaba, Google Cloud, and Microsoft Azure. The strategic partnership has already helped one established financial institution unlock significant value by accelerating the development and delivery of effective, integrated Cloud-based solutions. Anexinet’s proven Kickstart process and comprehensive set of tools and services deliver an Agile, scalable Cloud-based environment that embraces traditional IT as well as Private, Public, and Managed Cloud. Migrating applications and business systems to the Cloud is a daunting task for even the most mature organization. As a result, a Cloud-adoption strategy and roadmap often means the difference between successful deployment and failure to launch. Anexinet helps organizations determine their ideal strategic approach.

 

How Anexinet aligns with our 5D framework