Can ChatGPT write your documentation?
I’ve been fascinated with AI and its implications since the 90s. I’ve been skeptical, but the level of sophistication in the large language models today is pretty amazing. As a quick test of how well it can create documentation, I copied the text from a Zap we use in our production environment. The page had very little content.
A quick copy and paste into Notepad gave me:
Email by Zapier logo Trigger 1. New Inbound Email in Email by Zapier Twilio logo Action 2. Send SMS in Twilio iService (3.0.0) logo Action 3. Find Contact in iService (3.0.0) iService (3.0.0) logo Action 4. Create Ticket in iService (3.0.0).
This looked a little cryptic to me so I wasn’t expecting much. I put together a basic prompt and fed the hungry ChatGPT monster my little bit of description about the zap. The result was pretty impressive.
Write an explanation for this zapier.com zap: Email by Zapier logo Trigger 1. New Inbound Email in Email by Zapier Twilio logo Action 2. Send SMS in Twilio iService (3.0.0) logo Action 3. Find Contact in iService (3.0.0) iService (3.0.0) logo Action 4. Create Ticket in iService (3.0.0)
- New Inbound Email in Email by Zapier (Trigger): This step is triggered when a new email is received at a specified Email by Zapier address. The zap monitors this email address for any incoming messages and initiates the workflow once a new email is detected.
- Send SMS in Twilio (Action): After detecting a new inbound email, the zap proceeds to send an SMS notification via Twilio. The SMS can be customized to include specific information from the email, such as the sender’s name, email address, or subject line. This step helps to notify the appropriate personnel promptly when a new email is received.
- Find Contact in iService (3.0.0) (Action): In this step, the zap searches for a contact in the iService (3.0.0) system, using the sender’s email address from the inbound email. If a matching contact is found, the zap will use their information in the next step; if not, the zap can be configured to create a new contact based on the email sender’s details.
- Create Ticket in iService (3.0.0) (Action): The final step in this workflow is to create a support ticket in iService (3.0.0). The ticket is populated with relevant information from the email, such as the subject line and message body. If a contact was found or created in the previous step, their information will also be associated with the new ticket.
That was better than I expected. With a better prompt and some minor editing, that’s quite an explanation from the bit of information provided. I decided to press my luck and go for a more interesting version. ChatGPT doesn’t let me down. I gave it the prompt.
Rewrite that as a British professor in his late 70s
Ah, splendid! Allow me to elucidate the intricate workings of this remarkable contraption known as a “Zap” from the ingenious digital platform, Zapier. The purpose of this remarkable system is to expedite the process of handling incoming electronic mail, dispatching a telephonic message of the short variety, and generating support tickets within the esteemed iService (3.0.0) software. The procedure can be broken down into four distinct steps, which I shall now expound upon forthwith.
- New Inbound Email in Email by Zapier (Trigger): This particular stage is initiated upon the receipt of a fresh electronic mail communication at a predetermined Email by Zapier address. The clever device continuously surveys said address for incoming correspondence, thus setting the entire mechanism into motion when a novel message is discovered.
- Send SMS in Twilio (Action): Posthaste, after the interception of the aforementioned electronic mail, our intrepid Zap proceeds to dispatch a Short Message Service notification via the esteemed Twilio telecommunication service. The contents of said message may be tailored to incorporate specific intelligence gleaned from the original email, such as the sender’s moniker, electronic mail address, or even the subject line, ensuring that relevant personnel are duly alerted with alacrity.
- Find Contact in iService (3.0.0) (Action): Subsequently, this ingenious invention endeavours to locate a corresponding contact within the iService (3.0.0) system, utilising the electronic mail address of the original sender. Should a match be found, their particulars will be employed in the ensuing step; conversely, if no such individual is located, the Zap may be configured to generate a new contact based upon the sender’s information.
- Create Ticket in iService (3.0.0) (Action): In the denouement of this remarkable process, the Zap proceeds to formulate a support ticket within the highly-regarded iService (3.0.0) application. The ticket shall be endowed with pertinent data from the initial electronic mail, such as the subject line and the body of the missive. Additionally, the contact details discovered or generated in the previous stage will be duly associated with the nascent ticket.
By means of automating these rather laborious steps, this Zap significantly streamlines the handling of incoming electronic mail, ensuring that vital messages are swiftly acknowledged via a Short Message Service notification and an iService (3.0.0) support ticket. This, in turn, greatly enhances both efficiency and responsiveness when addressing customer enquiries or difficulties.
This is usually where I would write some type of summary to wrap up this post. But this time I decided to let ChatGPT-4 write it for me.
In conclusion, the astonishing capabilities of ChatGPT, as demonstrated in this article, attest to its potential for revolutionising the realm of documentation and knowledge management. The AI’s capacity to generate clear, coherent, and engaging explanations from rather cryptic information is a testament to the advancements in large language models. Moreover, ChatGPT’s flexibility in adapting to various writing styles, as exemplified by the British professor persona, showcases its versatility in catering to diverse audiences and contexts. Consequently, the implementation of ChatGPT in support processes, documentation, and knowledge management domains can lead to marked improvements in efficiency, responsiveness, and overall user experience. As AI technology continues to progress, the possibilities for its applications in these areas appear to be virtually limitless, with the potential to transform how we approach and manage information in the digital age.