AI is an excavator for knowledge workers, significantly enhancing teaching and research productivity.
Below is an article by Teacher Wang Jue on AI teaching skills, feel free to check it out:
————————————————
When controlling large models to complete general tasks,
you only need to give instructions in natural language.
For example, in Kimi, we can directly type requests such as:
-
Help me generate a summary of a certain article/link, etc.
However, for some more complex, specified, and constrained tasks,
To put it academically, prompts are a way to “fine-tune large models,
without changing the inherent capabilities of the large model,
by inputting specific prompts,
you can guide the model to output according to specific requirements.
For instance, before asking a question, we first tell the large model:
Who you are playing, What you should do, How to do it,
and sometimes you also need to specify what you cannot do!
the omnipotent general AI
is transformed into a specialized AI for a specific scenario!
This is the simplest method for adapting AI to specific scenarios.
For example, everyone can try this “Celebrity Database” AI Bot (just input a celebrity’s name, like “Yang Zhenning”)
We can see that the AI output is no longer in a “narrative” style,
but in a “structured” presentation:
can accomplish different tasks with different prompts!
Let’s take a look at what the above prompt looks like:
(The following is the prompt written in coze)
Of course, coze is different from general large models,
it can enhance the capabilities of large models beyond the scope of “general language” through “plugins”, “knowledge bases”, etc.,
providing richer and more professional services!
—— This is why Teacher Wang Jue recommends everyone use coze to develop AI functions
you can not only specify output content and format
but also specify the “user interaction method”
What is the “user interaction method”?
Everyone can first scan the code to experience the “Chinese Idiom Solitaire Game” made with coze
how the large model plays a role,
interacting with the user;
Moreover, it can even specify
the dependency relationships between introductions and conclusions,
and the large model will check this relationship according to the prompt,
This function seems much more complex than looking up information,
it can still be controlled with prompts!
Below are screenshots of prompts in coze:
Perhaps some people think, Teacher Wang Jue is really amazing
to write prompts so well and comprehensively!
Actually, it’s not that I’m brilliant, but coze is powerful!
We just need to write down our requirements in the “character setting and response logic” (which is actually the prompt)
and write a few sentences as we think of them
Then click the upper right corner “Optimize” button
coze will automatically help you optimize the prompt.
For example, the long prompt for “Chinese Idiom Solitaire” above
I actually just input four characters “Chinese Idiom Solitaire” 
(Of course, after coze optimized it, I added a bit of new requirements myself, which can be flexibly adjusted and tried, somewhat like natural language programming, very interesting!)
Why write the prompt in such a strange way?
——Because experiments show
when the prompt exceeds 200~300 words
and the instructions are complex,
the large model’s understanding of the instructions will sharply decline.
Therefore, it needs to be written in a “structured” prompt
so that it is easier for the large model to understand
and perform tasks according to our requirements.
Additionally, we used coze as an example
to introduce the role of prompts and how to write them.
are applicable to various large models,
we can copy the above prompt
and directly paste it into Kimi’s text box.
It has become a “Chinese Idiom Solitaire Dialogue Machine”:
if you want to save this prompt,
for direct use next time,
you need to click the cube button
and then click “Add Common Phrases” (which is actually the prompt)
Copy the prompt into the text box:
Don’t forget to give it a name,
so that in the future, entering “@name” in the text box can call this prompt.
As shown in the following image:
Additionally, coze is a development tool for creating your own “AI agents”,
while we can directly use large models like Kimi for our work.
We have seen how coze optimizes prompts,
saving us a lot of trouble!
In fact, Kimi also provides prompt optimization functions.
Kimi+ has a powerful “Prompt Expert” feature:
(For more information on Kimi and Kimi+, please refer to: 《Kimi+ is here, no need to write prompts anymore!》)
Just input the prompts you want directly into the text box
(write as many points as you think of)
The Kimi “Prompt Expert” will automatically help us write standardized prompts according to writing norms:
Finally, “breaking news”:
The dozens of typical “scenarios” provided in Kimi+
including “Prompt Expert” and “PPT Assistant”
essentially consist of built-in prompts,
controlling large models to perform “specific tasks”!
If everyone has used the functions of Kimi+
or experienced the two mini-games of coze today,
you must realize how important prompts are
in making large models serve as specialized tools,
how powerful large models are as “knowledge tools”!
★Source: Research on Learning Science and Technology Author: Wang Yu
★ Typesetting: Zhang Yangyang; Proofreading: Guo Jing; Review: Yang Qi
This article and its images are copyrighted by the original author and source, the content represents the author’s views and does not reflect the views of this public account or its authenticity. Reproduction is for non-commercial purposes only, aimed at sharing. If there are copyright issues, please contact us immediately, and we will modify or delete related articles to protect your rights.
Previous Recommendations:
Report by Deputy Minister Wu Yan 2018-2022 (including download link)
Improving Teaching Ability | Quickly Create 100 Pages of PPT in 30s
Essentials | Sharing of “Big Unit Teaching” Design Template
Summary of 17 Excellent “Curriculum Ideology” Teaching Design Cases
How Outcome-Based Education (OBE) Promotes Curriculum Reform in Higher Education (with three cases)
Analysis of Several Basic Types of Innovative Teaching Methods
Gold Medal Sharing | Display of Award-Winning Works from Previous “Internet+” Competition in Higher Education Main Track
Teaching Plan Template Sharing | First Prize in Professional Course Group of Teaching Ability Competition
15 Teaching Strategies Centered on Students: Reshaping Classroom Teaching
Sharing of Essentials | Guidelines for Integrating “Curriculum Ideology” into Teaching Syllabus
How to Make Teaching Competition PPT Stand Out to Judges!
Teaching Competition | 30 Excellent Works of No-Student Teaching Display at National Competition
2023 Humanities and Social Sciences Fund Project Application Reference with 133 Topics
Teaching Achievement Award: Summary Report of Special Prize and First Prize
Secrets to Passing Research Project Application (40 Questions + 10 Principles)
