Following the news that a student made a successful appeal against a parking ticket by using a letter written by the ChatGPT chatbot, we look at how it happened, how to get the best results from ChatGPT, and we look at some of the recent concerns about the advances in AI technology.
£60 Parking Ticket
22-year-old York student Millie Houlton received a £60 parking notice recently from York City Council.
Although she strongly believed that the fine (for parking in her own street) had been issued in error because she had a permit to do so, she turned to the ChatGPT chatbot for help in writing a letter of appeal against the parking ticket by giving it the instructions “please help me write a letter to the council, they gave me a parking ticket”.
Miss Houlton has been reported as saying about ChatGPT’s response to her instructions that: “I put in all my details about where and when it happened, why it was wrong and my reference for the fine and it came back with this perfectly formed personalised response within minutes.”
After sending off ChatGPT’s appeal letter, she was surprised and delighted to receive a response from the council telling her that the parking fine had been withdrawn.
Help From ChatGPT
This story shows one of the many valuable uses of an advanced conversational chatbot. ChatGPT can assist a person in writing responses to many different types official or legal letters and notices by providing relevant information and language that is appropriate for the situation. If you’d like to use ChatGPT in this way, here are tips to help you get the best possible letter from it:
– Provide context: Start by providing ChatGPT with the relevant details about the letter or notice you received. This could include the sender, the date of the letter, and the nature of the request or issue.
– Provide the text of the letter or notice: You can copy and paste the text of the letter or notice into the chat with ChatGPT. This will allow it to analyse the language used and provide suggestions for how to respond.
– Provide any relevant information: If there is any information that you think is relevant to the situation, be sure to provide it to ChatGPT. This could include documentation, receipts, or other evidence.
– Use ChatGPT’s suggestions: Based on the information you provide, ChatGPT can generate a response that is appropriate for the situation. It can suggest language to use and provide guidance on the tone and style of the response.
– Edit and refine: Once ChatGPT has provided a response, you can edit and refine it to make sure that it accurately reflects your position and addresses the issues at hand. You may need to make adjustments to the wording or tone to ensure that your response is appropriate for the situation.
It is worth remembering however, that however plausible ChatGTP’s output appears to be, it can have some shortcomings.
As an AI language model, ChatGPT is programmed to generate responses based on the patterns and information it has learned from the large corpus of data it was trained on. Like any machine learning model therefore, ChatGPT makes mistakes, produces incorrect or irrelevant information, and even generates and shares misinformation or disinformation, especially if it is fed with biased or unreliable data.
It is also important to remember that ChatGPT does not have the ability to verify the accuracy of the information it generates or assess the reliability of its sources. The chatbot has been designed and trained to mimic human language, but obviously does not have consciousness, intention, or moral values. This means that although it is not capable of intentionally lying or deceiving (as far as we know), it can nevertheless produce misleading or false information (especially if the input data is biased or incomplete).
There are also concerns that ChatGPT may not be compliant with GDPR’s privacy guidelines, and this has led to ChatGPT being banned in Italy. The Italian data watchdog’s concerns centre around:
– Whether there is a legal basis for the mass collection and storage of personal data to be used for ‘training’ the chatbot’s algorithms.
– Whether the inability to verify the age of ChatGPT users could result in minors being exposed to unsuitable answers compared to their degree of development and awareness.
What To Do
To avoid sharing a ChatGTP output (e.g. an important letter) that isn’t accurate, users should evaluate, fact-check, and cross-check the information with other reliable sources before using the response for any critical or important decision-making.
Also, to minimise the risk of misinformation or disinformation, it is recommended to use ChatGPT in conjunction with basic human critical thinking, human judgment, alongside other reliable sources of information, plus to try and ensure that the input data is accurate, complete, and comes from a variety of reputable sources.
To minimise the risk of privacy issues, ChatGPT users should not share personal, private, and identifiable information with the chatbot.
What Does This Mean For Your Business?
It’s still early days for the mass usage of powerful conversational chatbots – ChatGTP was only introduced in November, but search engines like Bing have already incorporated OpenAI’s chatbot technology, Google Bard is about to be introduced and Copilot is being integrated with Windows 365. Having access to the incredible capabilities and versatility of chatbots like ChatGPT has revealed many possibilities, e.g. getting parking tickets revoked (chatbots can pass legal exams).
Businesses have found that ChatGPT has saved them time, money, and helped them avoid the drudgery of some work.
Others, however, such as the 1,000 AI experts (including Elon Musk) who signed an open letter calling on all AI development to be paused for six months, are concerned that conversational chatbot technology is going too fast and could be exposing people, businesses and even humanity to existential risk. These concerns have, no doubt, been heightened with the news that OpenAI’s GPT-5 is scheduled to complete training this December and it is so powerful that OpenAI expects to achieve Artificial General Intelligence (AGI) whereby its chatbot could perform any intellectual task that a human being can!
There are also concerns that as much as chatbots can help with work, they could also threaten jobs or be used for malicious purposes, so while although the UK government has no plans for a UK AI regulator, it does plan to regulate artificial intelligence with new guidelines on “responsible use”. That said, it also recognises the contribution to the economy of AI.
The next few months look like being an exciting (or terrifying) and undoubtedly disruptive period as more businesses discover just how many unexpected ways that ChatGPT and other chatbots can be used to help them to add value and reduce costs.