Author – KiteEdge – kiteedge.co.uk
James Flavin, CEO, KiteEdge
In the iconic supernatural TV series, ‘The X-Files”, the opening titles of every episode features the tagline, ‘The Truth Is Out There’. As the growth in information and knowledge spirals ever-faster (one estimate suggests 3.5 quintillion bytes of data are now produced every day), it often feels that identifying and locating trustworthy content quickly and consistently becomes increasingly difficult. Recently, ChatGPT, a large language model using artificial intelligence, has been making a lot of headlines due to its ability to answer almost any question in a human-like manner (which it does exceptionally well). It has been lauded as a silver bullet which will deliver a step-change in search solutions and a potential category killer for the likes of Google, Yahoo and Baidu. But can you trust it?
The development of technology has had a significant impact on our ability to determine the truth. Historically, information was primarily shared through word-of-mouth or written records. These forms of communication were susceptible to inaccuracies and biases. Today, we have access to a vast amount of information through the internet and other forms of technology. However, this also means that it can be challenging to determine the credibility of information and the context in which it was presented.
Artificial intelligence and machine learning have the potential to revolutionise the way we determine the truth. For example, Natural Language Processing (NLP) can be used to analyse large amounts of text data and extract insights. This can be used to detect patterns of deception or misinformation. Additionally, AI-powered fact-checking systems can be used to verify the accuracy of information. However, it is important to note that these systems are only as good as the data they are trained on. If the data is biased or inaccurate, the system will likely produce biased or inaccurate results.
ChatGPT is unquestionably impressive but it is, however, not capable of determining the ‘truth’ as it is purely generating text based on the input provided to it. Chat GPT uses statistical patterns in the text rather than connecting words with actual meaning – in essence, it makes a guess at what the answer should be. It is important to remember that the text it generates may not be true and should be fact-checked before being used, particularly in any business decision-making context. The creators of ChatGPT openly admit the tool has shortcomings – in a blog post from November 2022 they noted that the system sometimes comes up with “nonsensical” answers because, when it comes to training the AI, “there’s currently no source of truth”. There is a danger, therefore, that systems powered by ChatGPT are trusted by those using them (due to the convincing human-like output), without there being a sufficient process to check that the results produced are accurate.
KiteEdge is a UK-based company that is building a platform which allows users to easily find relevant insight from within the knowledge/data that their organisation already owns. Unlike ChatGPT, which creates answers based on unknown training datasets, KiteEdge takes people straight to insight that they already have access to but which may not be easy to locate within fileshares and data stores. Importantly, the highlighted insights are from data sources that the company, or individual, already trusts – rather than being synthesised from an unknown information repository by a language model, like ChatGPT, meaning the information can be relied upon.
As AI-powered systems like ChatGPT are only as good as the data they are trained on, it is important to be mindful of their limitations and to use them in conjunction with other forms of fact-checking and critical thinking. This includes being transparent about the data and algorithms used to train these systems, as well as regularly re-evaluating and updating them to ensure they continue to produce accurate results. Ultimately, it is important to remember that technology is not a guarantee for determining the truth. It is always crucial to consider the original source and context in which information is presented and being reassured those sources can be accepted as true and trustworthy.