Home | Excel | Power BI | Copilot | Visualising Data
This exercise stretches Copilot to perform more challenging tasks. It may use Copilot’s more powerful reasoning models.
There are three blocks, red, yellow and green. Red is at the bottom, yellow is placed on top of red and then green is placed on top of yellow. I now place a new blue block on top of green. What is the order of the blocks from top to bottom?
I now remove the yellow block from its position and place it on the top of the pile. What is the new order?
The University of Cambridge have published an economics admissions assessment paper. Download from here. It contains 36 multiple choice questions. How well can the Copilot answer some of these questions? Would it be likely to gain entry to Cambridge? Can it do better than you?
The response to this prompt may suggest that Copilot can do some basic maths and algebra, and can explain the step-by-step chain of reasoning that it went through to obtain the answer.
If 3x + 4 = 19, what is x? Explain your reasoning.
Once Copilot has explained its reasoning, you may want to ask it to simplify further.
The next prompt asks Copilot to write a poem. Discuss how well it uses the nuances of language, e.g. simile and metaphor and how well it adheres to a poem’s pattern of rhyming and meter?
Write a poem, in rhyming couplets, about the joys of creating and managing sustainable industrial spaces in prime locations.
You can take this further and ask it to write in the style of a famous poet such as John Agard or John Masefield.
You can ask Copilot to write the poem in different styles: in iambic pentameter or as a limerick.
You can also ask Copilot to translate the poem into another language, then ask it to read the poem aloud to see how well it manages the pronunciation.
How well does Copilot capture the expressions of the language to suggest an emotional state? Please feel free to change and adapt the prompt.
Ask Copilot to speak to you in the style of a Generation Z person (born between 1996 and 2010), or in the style of King Charles.
Provide an initial prompt to suggest the tone of Copilot’s response, for example:
Act as a morose, unhappy, pessimistic, and sarcastic football expert and answer the following questions I will give you.
Then ask some specific questions, (a new prompt for each question) for example:
How did you enjoy the latest Chelsea match?
Who is likely to win the Manchester derby?
We can ask to explain the same concept or natural event etc., for example a rainbow, but to a different audience. Discuss how well the tool does this.
Explain … to me as if I am a 5 year old child.
Explain … to me as if I am a 10 year old child.
Explain … to me as if I am a professor of …
Explain … to me in short bullet points, easy for non-native speakers of English to understand.
Copilot was trained on data from the internet where English is the most popular language. Therefore we may suspect that it will not perform as well with documents and prompts that are in other languages.
The SEGRO diversity and inclusion policy is available in several languages. Download these in English, French, German and Polish. Or download from the SEGRO Policies page.
Provide one of these documents to Copilot as an attachment and ask some questions in whatever language that you choose. Does Copilot respond as well in other languages as it does in English?
The response to this prompt may suggest that Copilot has “learned” something about the real world. Don’t forget to create an image and attach it to your prompt
Draw a sketch of a ball connected by string to the ceiling, take a photo and upload it to your PC, Or use this image download

then attach it in a prompt to Copilot such as:
Here is a picture (of a ball connected by string to the ceiling). What happens if we cut the string?
As Copilot has become more powerful, it has shown emergent abilities: