Conversational AI has been quickly evolving and revolutionising the way people engage with technology in recent years. Chatbots, virtual assistants, and messaging apps have become more common with our everyday lives and their popularity is growing. Recently, ChatGPT is one technology that has played a big part in this transformation. Here, Tom Watson shares some insights into ChatGPT and discusses global concerns around its future usage.
What is ChatGPT and how does it work?
For those who are unaware or haven’t heard of ChatGPT, it is an artificial intelligence-powered language processing tool that delivers human-like conversational interactions, responding with content-based answers depending on the nature of the user’s questions and requests. Surprisingly, ChatGPT doesn’t connect to the internet or retrieve information from its internal database. Instead of searching the internet, it provides responses based on data from a variety of sources, such as existing websites, articles, and professional publications.
ChatGPT runs on OpenAI’s language model architecture called ‘The Generative Pre-trained Transformer’ (GPT). The specific GPT used by ChatGPT is fine-tuned from a model in the GPT-3.5 series. However, it hasn’t taken long for OpenAi to develop GPT-4, which is a more advanced system, leveraging additional data to create more advanced and revolutionary results. ChatGPT’s language model was refined using both supervised and reinforcement learning and is distinguished by its use of Reinforcement Learning from Human Feedback (RLHF).
What can ChatGPT be used for?
Since its launch in December 2022, ChatGPT’s website has been visited by a staggering 1.6 billion people. ChatGPT has the ability to answer factual questions, write code, and solve complex equations. Aside from answering complex solutions, it has also been used by individuals to answer ‘simple’, meaningful queries, describe art in great detail, generate email responses and even write university calibre answers for students.
Equally, the AI has the ability to refuse to answer queries, for example, asking advice for robbing a bank leads to a response such as: “it is not within my programming or ethical guidelines to provide advice or assistance in illegal activities, such as robbing a bank. It is important to remember that such actions have serious consequences, both for oneself and for others. Instead, I encourage you to seek out legal and legitimate means of earning money and resolving any financial difficulties you may be facing”.
Looking ahead – what are the future concerns?
Due to the rapid growth and development of ChatGPT in recent months, an open letter was drafted calling for organisations to pause developments of AI which has been described as “an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control”. The letter has received nearly 30,000 signatures, with the likes of Elon Musk, Steve Wozniak and Yuval Noah Narari contributing their support and signatures.
The letter also came at the same time as a published report from Goldman Sachs, which suggested that nearly 300 million full-time jobs could be impacted by AI systems like ChatGPT, escalating additional concerns around their impact on the future economy.
Furthermore, countries such as Italy have banned ChatGPT over privacy concerns and have expressed that it is “not sufficiently regulated”. Whilst Italy is the only country to take such action, the UK and a collection of EU countries have raised concerns around the platform and its usage.
If you’re a technology professional working across Scotland, considering your options and seeking career advice; or a client seeking a contract/interim technology solution, please get in touch with Tom for a confidential discussion: [email protected]