Can I generate photos of people through DALL∙E?
Yes, but DALL∙E will take creative liberty to ensure the person does not look like any public figure. For example, if you request “Cristiano Ronaldo”, a well-known footballer, it will return an invalid result. However, if you request “a face of a footballer”, it will return a fictional face. This safety measure is introduced to […]
Are DALL∙E images copyright free?
According to OpenAI’s policy, you own the images you create with DALL∙E and thus, own the copyright. However, you also need to be aware of the local and University law. While each country has different laws governing the copyright of AI generated images, in Hong Kong, under section 11(3) of the Copyright Ordinance (Cap. 528), […]
Do I need to cite DALL∙E for images I create?
Yes. It is important for the audience to be aware that the picture generated is not by any human but rather, AI. OpenAI services recommend users to word the citation as: “This image was created with the assistance of DALL∙E”. However, it is also allowed if users do not want to specify they used DALL∙E […]
I have a very specific image I want DALL∙E to create. Will it be able to do it?
Although we recommend you to be as specific in your request as possible, having more specifications lowers the chance of getting the image you want. In other words, this tool is more useful when looking to generate simple images, such as the scenery of Victoria Harbor rather than a specific graphic for the cover page […]
Can I specify the artistic technique to be used when generating an image through DALL∙E?
Yes, absolutely! You simply need to input the technique into your response (again, DALL∙E mimics human interaction so there is no template to phrase your request upon). If you want realistic photographs, you can also request the same.
How reliable is ChatGPT for information?
ChatGPT is a Large Language Model (LLM) and is trained upon the data which is fed to it. Consequently, ChatGPT is subject to misinformation and is not completely accurate nor up to date. It is a good practice to verify information provided by ChatGPT.
Can I upload files to ChatGPT for it to analyze?
Staff users can upload document files (.docx, .pptx, .xlsx, and .pdf) to any ChatGPT model for analysis. Please note that images in document files may not be processed. Document file uploads are not currently available to students. All students and staff users can upload image files (.png, .jpeg, .gif, and .bmp) to gpt-4.1-nano, gpt-4.1-mini, gpt-4.1, […]
I am looking to generate an output in Shakespearean English. Can ChatGPT do this?
Yes, there are different options to select the manner and language with which ChatGPT generates response. To change this, go to Settings at the top right and from the drop-down list for “System message”, choose your preferred mode of delivery. This means ChatGPT can even write poems in Shakespearean English!
Does ChatGPT only work in English?
No, ChatGPT can communicate in multiple languages. However, language proficiency and accuracy can vary language to language.
What is the difference between GPT-3.5 and GPT-4?
Like with any technology, there is constant development done to ensure improvement in the product/service. Similarly, GPT-4 is an improved version of GPT-3.5. There are no differences in the use case but, the responses generated by GPT-4 would be more accurate. Furthermore, the context in GPT-4 is 8,000 tokens, compared to the 4,000 tokens in […]