A.I. and Blog Writing – Assistive not Creative Yet!
By Khurram Sultan, Senior Developer @ Authority Partners
Artificial intelligence, and the array of technology that comes with it, is fast becoming an integral part of a new set of assistive tools and applications. At Authority Partners, our consultant Khurram Sultan is working on ways of integrating A.I. technology into the innovative business solutions our 500+ global team offer our international clients and partners. In this blog he discusses some of the key aspects of A.I. and machine learning and how they are being used with the assistive blog writing technologies that are now increasingly available across the internet.
Artificial intelligence (A.I.) has been with us in one form or another for over six decades now. The fundamental principle is simple – it is all about making a machine learn and infer. Learning happens through training: engineers use massive amounts of existing data to teach a computer how to perceive and synthesize patterns. They then teach the computer to generate new data based on reasoned and logical inferences that are drawn from those patterns. But principle and practice are two very different things!
Moore’s Law and innovative research have made this technology faster, smaller and better at doing a pretty good job at generating those inferences. Still, it is very far from perfect – particularly when the outputs involve creative and personal types of expression like literature and art. Self-driving cars and blog writing applications both use A.I. technology, but we’ve come much closer to perfecting the mechanical and highly rules-based task of navigating a car on a highway or a city street than generating an award-winning novel written by an A.I. bot.
Assistive. But (Very) Not Perfect
Though the Nobel Prize in Literature may not be within the grasp of existing machine learning or natural language generation technology, writing basic work emails to colleagues or creating outlines for a blog post are things the technology can do well. The key word when thinking about A.I. in this context is “assistive.”
These technologies can help an inexperienced or a second-language writer get started with the writing process by creating outlines or helping them generate short and relatively simple pieces of text – and some of these short excerpts are even passing the Turing test. But as soon as pieces get longer, more specific and more complex, the machines get into trouble. The auto-regressive, deep learning algorithms that most writing applications use have very particular limits.
They might help a user through initial writer’s block or quickly generate an outline, but they typically require crystal clear instructions (since most online applications will pull research material from wherever it’s available on the web). They also aren’t particularly good with language structure or voice and tone – meaning that users need to be able to restructure, edit and proofread material extremely carefully before publishing or presenting their documents publicly.
GPT-3 Natural Language Processing
A catalyst for the recent surge in online writing technology has been the development of GPT-3. The Generative Pre-Trained Transformer 3 (GPT-3) uses deep learning to produce patterned language and is part of a trend in natural language processing (NLP). Developed by Open A.I. in 2020, the technology offers pre-trained language representation that uses a neural network based on a deep learning model. This makes it exceptionally good at generating human-like speech and text.
The New York Times has described the technology as ground breaking and able to “revolutionize how we search and use information in the next few years.” The sophistication of the A.I. system means that it will steadily improve its ability to create structured and coherent language-based documents. And though Open AI is the only company with access to the underlying technology, the GPT-3 API is available to any user willing to pay for a license.
The sudden explosion in online writing applications made possible by GPT-3 is a testament to the incredible value of the market for A.I.-based technologies. The size of the A.I. market has been valued at $328 billion in 2021 and is projected to grow to over $1.3 trillion USD by 2029.
NLP is only one slice of the A.I. technological pie. In healthcare, manufacturing, sales and marketing, in addition to the obvious applications in hardware, software and services, A.I. will become an integral element of any IT solution. Amidst a raft of as yet undiscovered possibilities, and in addition to the assistive capabilities in writing emails and blogs, A.I. technology will improve accuracy, efficiency and forecasting capabilities, driving market growth.
A.I. at Authority Partners
At Authority Partners we are investigating new and innovative ways to include A.I. technology in the suite of solutions we make available to our clients. We are dedicating resources to develop new areas of in-house expertise to create and find better ways to integrate assistive and generative A.I. solutions and harness the full power of the technology.
A.I. represents the next significant step in technological advancement and our global team of consultants are well placed to maximize its potential. A.I. goes to the very heart of our commitment to delivering excellence and innovation. Harnessing innovative technology and integrating it into cutting-edge, actionable business solutions to boost market opportunities for the companies and institutions we work with is what sets us apart.
For 25 years, we at Authority Partners have been providing IT solutions that have helped our clients become leaders in their industries. We have been partners in their growth, creating software solutions that deliver a clear competitive advantage. We are partners who help deliver business outcomes.
We are Authority Partners.
Interested in getting in touch? Fill out this form and a member of our team will reach out.