AIPRM’s user interface offers many prompt templates and specific instructions to help users get the best out of ChatGPT.

For example:

A user can run the keyword “AI software development” using the Keyword strategy prompt template and get results on how to market software development:

You have an industry overview, key Concepts, tools and Frameworks, and business applications. Going down further, you’ll see it is already structured.

Regardless of all the already returned ideas, the user can still ask chatGPT to continue.

You can also copy any of the ideas returned and ask ChatGPT to write an article on it. Example Let’s take the first one, “the beginner’s guide to EI software development” :

ChatGPT returns an output immediately, and as you can tell, it’s quite fast in doing so.

You can also go further to instruct ChatGPT, using AIPRM’s features, to explain the output:

You can get much faster output on ChatGPT plus than with the free version. A user can decide to further work on the output by instructing ChatGPT to expand it:

A user who wants a TLDR summery of the output can go ahead and search AIPRM for the TLDR prompt template. then copy the output into ChatGPT and run:

This returns gives a quick summary again from the created content, ready for LinkedIn with nice emojis. Using the ChatGPT features in the “continue” drop-down menu, as well as the change of style and tone menu, one can instruct ChatGPT to return desired outputs.

Another Example is AIPRM’s time machines.

There are a couple of time machines like “the time travelers Guide to World History” prompt template created by David Birss. Given a destination and a time keywords, this prompt returns what to see, what to do, and what to care about:

What else can we do with AIPRM prompt templates?

AIPRM offers a lot of generative art for mid-journey prompt templates.

If you give them a concept, they would take that concept and extract the relevant details for the mid-journey art generator and then return an output,  eg, “A robot dancing in the snow” :

 

Prompts are the new code. You cannot simply slap a sentence in and expect wonders. You have to fine-tune it. You must learn how to work with the machine, including how this specific model responds to your prompts.

This means that a fantastic prompt for ChatGPT may not work for other large language models. We can see it today when a great prompt on ChatGPT cannot be used on versions before the GPT-3, which is available on the OPenAI playground. Prompt engineering is here to stay. Therefore, a prompt engineer must learn and qualify for a specific language model.

AIPRM offers different categories of prompt templates:

The Software engineering category is very popular amongst users, especially the Python pro and HTML & CSS Hero prompts. Based on the python pro prompt template, one can ask, “how can I read a file and print each line” and get excellent, detailed step-by-step output in very popular languages like python or C, or Java – they work really well with ChatGPT.

Newer language and templating systems like the Hugo templates don’t work with ChatGPT simply because chatGPT has a cut-off date- summer 2021. Consequently, It doesn’t know anything newer than that.

Users who come across prompt templates that don’t work can always give them a thumbs down to help AIPRM investigate them. The more thumbs down a template get, the more it will be filtered away.

You can also report a template to AIPRM and choose the appropriate reason for the report.

All the reports are taken into AIPRM’s ranking algorithms to give users a better experience.

AIPRM is also a search engine for the prompt. Therefore, like other search engines, e.g., Google, AIPRM experiences problems such as duplicate content, but these problems can be tackled underneath in AIPRM’s backend system.

Furthermore, through AIPRMs templates, users can interact with ChatGPT to learn different languages:

AIPRM also offers templates that get your HTML corrected or where you can ask questions about HTML:

There are also prompt templates that rewrite articles to make them human-written. It is important to note that most AI content detectors can not easily detect ChatGPT “Human” written content.

A good number of these detectors don’t work really well since they are trained on hundreds or thousands of examples and function by matching the output content to the patterns it has been trained to recognize; it then returns an estimation of how much that pattern matches to the output pattern. Using AIPRM’s prompts to rewrite your content makes it harder for the content to be flagged down by AI content detectors.

A user can just change the tone of content or ask ChatGPT to expand and rewrite it. That way, many AI detectors usually fail detection and cannot detect content coming out of ChatGPT reliably.

Christoph Kemper, the founder of AIPRM, stated that they are only getting started.

AIPRM is a Marketplace and a community, and as such, everyone is encouraged to join it and use it towards learning how to prompt AI, how to prompt chatgpt, and other AI platforms.