By now, we all should have heard about ChatGPT. This AI language platform has taken the world by storm since the beginning of the year, and as such, has received praise, amazement as well as criticism for its approach towards content creation. As more and more people are becoming aware of its capabilities, ChatGPT has opened up some rather interesting discussions around AI technology, with a particular focus on the ethical stance we should apply when individuals use AI tools to create ‘original’ content. In this article I want to breakdown ChatGPT from a number of angles and explore the way it is used.
What is ChatGPT?
As stated on their platform, ChatGPT is ‘a language model trained to produce text.’ In its active form, it is not connected to the internet and does not draw information directly from a connection. The text it produces is a result of a trained model built on ‘vast amounts of data from the internet, written by humans.’ Unlike some other AI language creators, ChatGPT is impressively convincing when it comes to adding a ‘human’ element to its words. Within this context, trying to determine whether a piece of work has been written by AI or not can be difficult.
Creating Content on ChatGPT
As an example, below are two short paragraphs that try to answer the same question ‘How do I stay safe online?’ One is written by myself, and the other is written by ChatGPT – see if you can determine which one is which.
- Staying safe online requires being aware of potential risks and taking precautions to protect yourself. Some key tips include using strong and unique passwords, enabling two-factor authentication, avoiding clicking on suspicious links or downloading attachments from unknown sources, keeping your software and devices up-to-date with security patches, and being cautious when sharing personal information online. It's also important to use antivirus software and a virtual private network (VPN) when browsing the web to further enhance your online security.
- In order to stay safe online, individuals must encompass a wide variety of understandings, behaviours and practices to support them when they go online. As well as cyber security initiatives to protect our devices such as antivirus software and strong passwords, users will also need to build their own understanding around how they can safely navigate online spaces, without falling victim to online harm. This requires critical thinking, awareness of what online harm can look like and knowledge around how to safely respond to and report harmful content. Different situations require different responses though, so training and education must be applied across specific areas within online safety and cyber security.
Just for clarity, ChatGPT wrote the first paragraph in under 10 seconds! As you can see, this answer went into a few more specifics than I did and directed its focus more towards cyber-security than actual online safety. But credit where credit is due, their response is not wrong, and from a ‘security’ point of view, this is pretty good information to take away.
My response on the other hand is broader and tries to highlight that online safety and cyber security are separate entities. I deliberately chose this question as it cannot be answered properly in a single paragraph, and I thought it would be interesting to see which direction ChatGPT would take it. Its answer wasn’t wrong, but it missed out some fundamental points.
How Accurate is ChatGPT?
Despite being rather impressive in how it creates its information, ChatGPT themselves have stated that ‘it can occasionally produce incorrect answers’ or create content that is ‘inaccurate, untruthful, and otherwise misleading at times.’ To address this, ChatGPT has implemented a feedback tool for users to highlight incorrect responses.
Although this may seem like a significant downside to how the AI works – I personally feel as though this can be seen as a positive element in how users engage with the content. It is rather appropriate that those who use AI tools will still need to apply vigilance and understanding of the content in order to produce quality work. This layer of uncertainty can encourage users to take more diligence by applying appropriate research and fact checking to the content. Needless to say, ChatGPT can produce words at an impressive rate, but there’s still a clear reliance on users to apply critical thought towards its output.
Harmful Online Content
More concerningly, the platform has also raised that content ‘may also occasionally produce harmful instructions or biased content.’ As ChatGPT was launched in November 2022, it is still young, and as such, we would hope that their approach towards harmful online content will improve as they continue to update. But as ChatGPT is now a worldly recognised system, there is a need to understand how harmful content may be experienced.
Much like we’ve already said, critical thinking is an essential behaviour when using online spaces such as these. As such, we might approach ChatGPT as if it were a search engine without necessarily having the appropriate filters in place. In the same way that we educate around why it is not good behaviour to google something inappropriate, the same must be applied here. As well, acknowledging that content may be presented as biased or inaccurate can encourage users to re-evaluate and take specific care when extracting information.
ChatGPT at School
One of the largest controversies surrounding ChatGPT has been around its impact on schools. In particular, the issue of students using AI tools to produce written homework and other assignments. With concerns around cheating and debates around whether students can effectively apply knowledge when using AI tools, schools have yet to be formally guided by Ofsted and the Department for Education. There have been instances of banning ChatGPT on school systems altogether, as well as encouraging more vigilance from staff, but overall, the correct approach is still very much up for discussion. Needless to say though, ChatGPT may be adding an extra layer of consideration for schools right now.
While it seems too early to determine whether ChatGPT is causing any significant damage to schools, it is clearly looking to be a ‘game-changer’. There have been some very healthy arguments for why schools should be embracing AI tools to assist with their teaching strategies. Much like calculators and search engines though, AI tools could be seen as an assistive technology that supports work rather than hinders it. Whichever approach you decide to take, acknowledging the benefits as well as the shortcomings can still help you present a balanced outlook within your community. Remember that putting a guard up and banning its use may result in the opposite effect.
As a piece of technology, I can safely say that ChatGPT is incredibly impressive. It really is a milestone in how AI can be used to consume content. Its capabilities are often quite astounding in how easily it creates content off the back of specific commands, and I would recommend anyone to give it a try.
Despite this, we must acknowledge that there is still a lot of work to be done. While I have stated that its shortcomings in accuracy can be considered a positive in keeping a stable connection with the user, I feel as though its approach towards harmful content, still requires attention. In its current form, users will need to apply vigilance and critical thought towards ChatGPT if they hope to use it in a safe and positive way.
I feel as though we are getting a clear look into the future of how we search for and approach content. ChatGPT may just be the start of how we use AI tools within our education as well as our professional lives, for better and for worse. As such, we must embrace it, examine it, refine it and see how we can safely introduce it into our lives. To fully understand what impact this technology could have on the world, I asked ChatGPT how it will change the way we search for and consume content online. Below was the response:
ChatGPT has the potential to significantly change the way we search for and consume content online in several ways:
Conversational search: ChatGPT can help evolve search engines from keyword-based queries to more natural, conversational interactions. Users can ask questions in plain language and receive more contextually relevant and detailed responses. This will make searching for information more intuitive and user-friendly.
Personalised content curation: ChatGPT can understand user preferences, interests, and needs, helping curate and recommend content tailored to individual users. It can deliver more accurate and relevant results, ensuring a more personalized browsing experience.
Enhanced language translation: ChatGPT's ability to understand and generate human-like text in multiple languages can revolutionize content consumption across language barriers. As a result, users will have access to a broader range of content, regardless of their native language.
Improved accessibility: With ChatGPT's natural language processing capabilities, it can help make the internet more accessible for people with disabilities, such as visually impaired users, by providing voice-based search options and converting text to speech.
Context-aware assistance: ChatGPT can integrate with various applications, websites, and devices, providing context-aware assistance and information. It can act as a virtual assistant, offering help with tasks, answering questions, and even providing real-time information updates.
Content creation and summarisation: ChatGPT can help with content creation by generating human-like text, which can be used for various purposes like blog posts, articles, or social media updates. It can also summarize lengthy articles or documents, allowing users to consume information more efficiently.
Online learning and tutoring: ChatGPT can serve as an online tutor, helping students with subject matter questions, explanations, and even guiding them through problem-solving exercises. This can make education more accessible and personalized.
As ChatGPT continues to evolve, it will likely lead to more advanced and user-centric ways of searching for and consuming content online, making the internet an even more valuable and accessible resource.
In conclusion, ChatGPT, a ground-breaking AI language platform, has sparked numerous debates around content creation, ethics, and online safety. With its ability to produce convincingly human-like text, it has proved challenging to distinguish AI-generated content from human-written pieces. As the platform raises concerns about harmful content and potential misuse in educational settings, it is crucial to explore its benefits as well as shortcomings. ChatGPT is likely to change the way we search for and consume content online, ushering in a new era of conversational search, personalized curation, and accessible education.