Welcome to ValiantCEO Magazine, your gateway to exclusive insights from the world’s most innovative leaders. In this captivating interview, we sit down with Jon Gillham, the driving force behind Originality.AI, a pioneering venture at the forefront of artificial intelligence and content detection.
Jon Gillham’s journey is nothing short of remarkable. After the successful sale of a content marketing agency, he embarked on a new adventure by founding Originality.AI. This cutting-edge platform harnesses the power of a modified BERT model, a Google-developed AI system, to transform content verification.
With a dedicated team of machine learning engineers, researchers, developers, and marketers, Jon is on a mission to provide unparalleled transparency when it comes to assessing the authenticity of written content.
Join us as we explore the challenges and triumphs of AI integration in business, the future of AI in the industry, and Jon Gillham’s vision for Originality.AI as it aims to become the industry leader in AI detection by 2023. This is a conversation that promises to unveil the limitless potential of AI innovation.
Check out more interviews with entrepreneurs here.
WOULD YOU LIKE TO GET FEATURED?
All interviews are 100% FREE OF CHARGE
Table of Contents
We are thrilled to have you join us today, welcome to ValiantCEO Magazine’s exclusive interview! Let’s start off with a little introduction. Tell our readers a bit about yourself and your company.
Jon Gillham: My name is Jon Gillham – I am the founder of Originality.AI. After selling a content marketing agency he had previously founded.
I launched Originality.AI in November 2022, after selling a content marketing agency I had previously co-founded. Originality.AI was started as a way to be able to detect GPT-3 generated content (before ChatGPT had launched).
Having been one of the earliest adopters of generative AI content for SEO purposes at scale through our agency, I understood the wave that was coming which Chat GPT and GPT-4 have fully unleashed.
Additionally I believed that there was a need for a modern plagiarism checking solution that didn’t suck! One that provided scan history, detection scores, shareable results, team access etc etc… what you would expect from a modern Plagiarism Checking solution.
The team now includes experienced Machine Learning Engineers and Researchers, Developers and Marketers all on a mission to try and provide transparency on the true originality of any writing.
Can you share with us your journey towards integrating AI into your business operations?
Jon Gillham: Originality.AI is built using AI. The tool’s core is based on a modified version of the BERT model. Also known as Bidirectional Encoder Representations from Transformers, BERT is a Google-developed artificial intelligence (AI) system for processing natural language.
We used a modified version of the BERT model for classification. We tested on multiple model architectures and concluded that with the same capacity, the discriminate model with having a more flexible architecture than the generative model (e.g bidirectional) makes it more powerful in detection.
Another thing we also learned is that the larger the model used, the more accurate and robust the prediction will be. The usage model is large enough but still meets the response time, thanks to the most modern techniques in deployment.
What specific areas of your business have been most impacted by AI, and how?
Jon Gillham: Originality.AI is in a unique position in that it’s both built using AI and was built in response to AI.
The idea was launched based on the rapid growth of generative AI content creation – we saw a need to address this for content creators and publishers by offering a tool that could accurately detect whether or not a piece of content was generated by AI vs human-written.
So, in a sense, you could say that EVERYTHING about our business has been impacted by AI (and continue to be)!
What are the biggest obstacles you’ve faced in implementing AI, and how did you overcome them?
Jon Gillham: To generate text, there are many different sampling techniques, like temperature, Top-K, and nucleus sampling. We expect more advanced generation strategies (such as nucleus sampling) could make detection more difficult than generations produced via Top-K truncation.
Using a mixed dataset from different sampling techniques also helps to increase the accuracy of the model. In our training data generation, we try to generate in different ways and on many different models so that the data is diversified and the model understands the types of text generated by the AI in a better way.
One of the challenges of the detection model is detecting difficult short texts, although it is very short texts, we have also improved the accuracy of the model in these cases by using our different techniques. “after using different methods and testing, our model…” can improve accuracy by more than 15% compared to the previous model.
Therefore, the current model has improved accuracy on much shorter texts, but according to our experiments, with lengths of 50 tokens or longer, the model has acceptable reliability.
What advice would you give to other CEOs looking to integrate AI into their business?
Jon Gillham: Building a business and/or tool on AI vs integrating AI into you business are two very different things. Originality.AI is both built on and integrates AI, so our model is unique. However, for anyone looking to integrate AI into their business, here is my advice:
- Test before you dive in: there are so many AI tools and capabilities on the market currently. Make sure to do your research and test out different models, tools, processes, etc before incorporating them fully into your business and procedures. See which ones fit best, how they can be incorporated, and whether or not they will make you more efficient/effective.
- Don’t incorporate AI for the sake of it: just because everyone is doing it, doesn’t mean you have to. Find gaps and opportunities in your procedures and see where there is room for AI to be incorporated – and ONLY do so when it makes things more efficient. Remember, AI is a great tool to make you, your team, and your procedures work faster, better, etc. It is not a replacement for someone. A great pair of shoes can make you a faster runner, but you still need to put them on and run.
- Track metrics: you’ll never know whether or not incorporating AI into you procedures is making your team more efficient/effective without tracking the progress and results. Make sure to track everything and review constantly to see where improvements are being made, and where they can be made even further. If it’s not aiding your business, don’t be afraid to ditch it and try someone else.
How do you see AI evolving in your industry over the next 5 years?
Jon Gillham: AI is being incorporated into almost every faucet of our lives and will most likely continue over the next 5 years and beyond. We’ll see new tech come to fruition; most businesses will be using AI in one way or another as a tool to make their teams more efficient; and we’ll also see a lot of rules and regulations come into play from governing bodies and large brands to ensure it’s used ethically and fairly.
Personally, I don’t’ see AI being as “scary” as a lot of people make it out to be – I see it being used as an extremely useful tool in people’s everyday lives (and businesses). We’ll definitely see a lot of disruption (in a good way) over the next 5 years.
What does “success” in 2023 mean to you? It could be on a personal or business level, please share your vision.
Jon Gillham: Simple: Originality.AI becoming the leader in AI detection
Jerome Knyszewski, VIP Contributor to ValiantCEO and the host of this interview would like to thank Jon Gillham for taking the time to do this interview and share his knowledge and experience with our readers.
If you would like to get in touch with Jon Gillham or his company, you can do it through his – Linkedin Page
Disclaimer: The ValiantCEO Community welcomes voices from many spheres on our open platform. We publish pieces as written by outside contributors with a wide range of opinions, which don’t necessarily reflect our own. Community stories are not commissioned by our editorial team and must meet our guidelines prior to being published.