Location>code7788 >text

How Programmers Can Leverage AI to Improve Themselves: a Comprehensive Guide to Going from Working Productively to Upgrading Your Skills

Popularity:473 ℃/2024-11-11 13:54:33

It's 1024 again, time is running out, and it's that special day again. Frankly speaking, I didn't really want to write this article because I didn't find the secret of making money efficiently through AI that is universally applicable to the programmer community. However, reflecting on my past work, I realized that utilizing AI did allow me to become more flexible in my work, and as a result, I had more time to think about other issues. As such, I wanted to share some of my thoughts and experiences in using AI, although these are just my personal opinions and may not apply to everyone. If you have a different point of view, you are of course free to express your opinion. The main purpose of this article is to hope that my experience can provide some references to help you better cope with the challenges in this fast-changing era.

image

All in all, based on the various assistants I am currently using, I do feel that they have made a significant difference in shortening my work time, as well as significantly enhancing my learning ability and knowledge. By utilizing these tools wisely, I am not only able to complete my tasks more efficiently, but also free up time to explore new areas and gain a deeper understanding of more complex concepts.

I hope you can draw inspiration from my experience, find a learning method that suits you, improve your work efficiency and knowledge base, and work together to make progress in this age of information explosion!

Unprecedented change

From machine learning to natural language processing, AI applications are continuing to permeate every field, however, the impact for programmers doesn't seem to be as significant. It wasn't until the emergence of Large Language Modeling (LLM) that it really sparked a technological revolution, entering the general public's view with a fresh perspective, including my own, which was deeply affected by it.

image

However, as time went by, LLM faded away its great power and seemed to have reached a top height. At the same time, the concept of intelligent bodies began to attract widespread attention, and I was fortunate enough to participate in this wave. From the initial unrestricted phase, in which the public was free to create all kinds of intelligences and participate in all kinds of competitions and all kinds of money-consuming explorations in an attempt to find a real entry point.

image

Now, over time, the direction of AI development has gradually changed, and it has begun to focus more on commercialization possibilities, such as ad placement, lead conversion and other areas of exploration. This shift suggests that while AI technology is still advancing, the pace of its development appears relatively cautious, failing to recreate the powerful shock that GPT brought when it was first introduced to the public.

Among the many tools I have come across, I would like to list in detail some of the resources I can draw on, such as Large Language Models (LLMs), mapping tools, intelligences, etc. Through in-depth learning and effective application of these AI tools and frameworks, I can better adapt to current technology trends. These tools will not only increase my productivity, but also optimize my workflow and mindset for greater creativity.

LLM

In fact, the tool I use most often remains the Chat Big Model. When I think back to the time when it first emerged, I developed a deep fascination with it, which led me to develop a habit: no matter what problem I encountered, my first impulse was to go and ask for advice. As a result, this habit has become so ingrained that almost every time I encounter a challenge, I feel a little uneasy if I don't go and ask about it.

image

Of course, the reference here is mainly to GPT, while in the domestic environment, I prefer to use Kimi. this tool is really a pretty good choice, not only powerful, but also excellent in many application scenarios.

Introduction to Technology

In fact, initially I most often used the Chat Big Model for the purpose of asking all sorts of basic questions, especially about Python as a development language. During my learning of Python, I hardly referred to any videos or documentation, mainly because I already had a foundation in Java development, so those materials seemed a bit lengthy and verbose to me. During the process, I often asked GPT questions and used its instant feedback to solve the problems I encountered, which enabled me to master the basics of Python very quickly.

image

From the very first installation to the gradual in-depth study, I found myself increasingly interested in web crawlers. Crawling techniques are not only very useful in practical applications, but also help me to access all kinds of data.

However, in the learning process encountered some error reporting problems, with the aid of search engine tools to assist is very important. Thinking back to the stage when I learned Python from scratch, it took me about less than half a month of spare time to master some basics. In addition, I have also written some Python refresher guides for Java developers. It is worth mentioning that I have shared some valuable resources in the Python boutique learning library of Tencent Cloud community, hoping to communicate and share experiences with more learners.

image

Currently, I'm still using GPT to learn some new techniques, especially since I've recently developed a strong interest in machine learning. When it comes to anything I don't understand, GPT's help tends to be more reliable than a search engine, especially when it comes to providing concrete life examples. Notably, I think it's particularly effective for beginners because it can explain complex concepts in a more intuitive way.

image

However, despite having acquired the basics and mastered the skills through the GPT, I still believe it is vital to scrutinize the official documentation. After all, the GPT is currently no substitute for the depth required for systematic learning.

Programming Assistant

Apart from the above studies I often ask about GPT, programming code at work is my main daily task. If conditions permit, I will definitely try to use GitHub Copilot to improve my development efficiency. However, as an ordinary developer with a limited financial budget, I actively look for practical programming assistants in China. In this regard, I chose Tencent Cloud Programming Assistant, which excels in code auto-completion and smart tips.

In addition to this, Tongyi Spirit Code is also worth mentioning. All of these tools are highly recommended options for developers looking to be more productive.

image

Since my current role is primarily a CRUD programmer, understanding and implementing this complex logic sometimes seems a bit overwhelming. In order to be more efficient, I would like to use a programming assistant to help me generate the appropriate code so that I can save a lot of thinking time. After completing the implementation of these logics, I just need to perform a simple boundary test on the result to make sure it works properly, and then I can smoothly move forward with the progress of my project.

add a few finishing touches to an article

As a blogger with a passion for technical writing, I spend my spare time every day immersing myself in the world of writing, documenting each of my favorite technical topics and keeping them as notes in my writing platform. However, one of the things that bothers me the most is the uncertainty of whether my statements are perfect enough. Although readers tend to find my articles logical and well-organized, in reality, they are carefully embellished.

image

My base drafts are essentially complete, with the touch-ups only enhancing and optimizing the original base draft. Typically, my base drafts are about 85% of the words after touch-ups, while the remainder relies heavily on the help of the GPT in order to expand and enrich the content, thus making the article more fluid and readable. In this way, I not only retain the core of the original idea, but also improve the readability of the text.

I really don't use a specialized writing tool here. These days, many writing tools are equipped with AI amplification. However, I prefer to take my own time to finish writing and then touch it up appropriately. If you are also a writing blogger, you may also utilize this feature.GPT solved one of my writing challenges - poorly organized sentences.

image

As a result, I've grown to love sharing through consistent writing, and I've met many bloggers and joined various communities. However, my favorite is the user-friendly Tencent Cloud community. If you're interested in this, feel free to join the Creative Star group chat and I'll be waiting for you there. The experience is unmatched by any other technical writing community.

AI Cue Words

I really don't want to get into the game of prompts because I foresee that prompts will eventually be templatized and simplified, which for the time being will undoubtedly raise the bar for people to use large-scale Language Models (LLMs), a problem that will be solved sooner or later by large enterprises. However, for the time being, we don't seem to be able to take these potential problems into account. As a result, I've been seeking help in generating prompts from national prompt word experts like Kimi, whose prompts are very versatile and adaptable to a wide range of scenarios.

image

But even then, we need to tweak the prompts to ensure they work best on different smartbody platforms, as the same template can work very differently on different platforms. We must take into account the characteristics and applicability of each platform to ensure that the cue words are widely applicable. This is exactly the current situation in the domestic smart body field, and I have no choice but to adapt to it. Since I have been studying and using smart bodies in depth, I have been using this feature quite frequently.

fantasy (fiction)

The rest is a variety of other problem solving. This includes not only generating creative ideas using GPT to inspire and possibly adopt certain ideas, but also solving technical problems, such as when I forget exactly how to write a certain server command, I will seek help from GPT. The answers provided by GPT are often more precise and easier to understand than traditional search engines.

image

Additionally, I turn to GPT for advice when I have a clear question about a topic but am unsure how to structure an outline or expand on an idea. These suggestions help me determine the direction and structure of my writing, after which I can write content based on these guidelines.

intelligent body

Smart bodies are really popular this year, looking back at the beginning of the year, it was a wise move to get in touch with coze early. As time went by, I discovered more excellent intelligent body platforms, such as Yuanqi, Qianfan AppBuilder and Baidu Wenshin Intelligent Body, which enriched my horizons. This year, I have participated in as many as ten smart body competitions, and the prizes I have received are quite impressive, ranging from iPhone 15, $1000 JD card to Lofi keyboard, really everything. My love for smart bodies is growing day by day, and it really makes me feel that this is a bright way out for programmers in the future, or at least this is my personal deep experience.

image

After all, having the ability to code provides me with a natural advantage, and even though the application of intelligent bodies is often premised on zero code, if you know how to program, the implementations you can come up with will often be more creative and effective. In order not to be left behind by this trend, I also seized the opportunity to open the B station UP master, specializing in sharing knowledge related to intelligent body and tutorials for beginners, and strive to help more people understand and master this cutting-edge technology.

image

In order to explain the knowledge of intelligences in more depth, I also created a WeChat group chat and a QQ channel, but these platforms are mainly maintained and operated by me personally, so the management style is relatively casual, and the communication within the group is not very frequent. After all, I still have a lot of work to do, so I can't devote myself to the operation of these social platforms, which is also called "having the heart but not the power".

Of course, I have also actively joined several internal communication groups related to intelligent bodies, and in my spare time, I will browse other people's innovative thinking and insights, so as to broaden my own vision and ideas, after all, individual ideas and creativity are limited. After all, individual ideas and creativity are limited. Through continuous learning and practice of intelligent body authoring, I can gradually process and systematize my inner abstract concepts, and this way of thinking not only improves my work efficiency, but also saves me a lot of time cost, which can be said to be two birds with one stone.

image

If you're interested in the field of artificial intelligence, start by reaching out and trying to build simple intelligences - it's a very accessible and fun process.

Spring AI

For many programmers, myself included, I'm very curious about how the SmartBody platform is built. Despite already knowing a bit about LangChain, I personally don't tend to build projects in Python, preferring Java projects instead. Recently, I discovered a new framework called Spring AI and tried building a project myself. Through this process, I gained a more intuitive understanding of the Intelligent Body platform and realized that it is actually not as complex as I thought. However, I still encountered quite a few challenges in the details during the actual operation.

If you're equally interested in this, you might want to refer to my full breakdown of the process of building an assistant using Spring AI:/guoxiaoyu/p/18453559

image

The Spring AI framework is truly impressive, and even as a programmer who primarily deals with business logic, it would be beneficial to familiarize yourself with this new addition to the Spring ecosystem in advance. Given that Spring AI is still in the snapshot version stage and the official documentation provided is relatively limited, being able to master this technology ahead of time is certainly an advantage in the future competition for your career. After all, isn't it a way to get ahead of others when everyone around you is still figuring it out and you're already proficient in its use?

Platform tools

Originally, I intended to incorporate this section into the above discussion of intelligent body modules, but on reflection it seems that this is actually a form of self-improvement - i.e., discovering and utilizing quality tools. By tools, I don't mean gadgets that simplify a programmer's day-to-day development tasks, but rather a service that can be directly integrated into an intelligent body and realize specific functionality through API calls. For example, I recently stumbled upon an EdgeOne API that is in internal testing that allows users to generate certificates and merchandise templates by passing parameters.

image

This API is very useful because it allows developers to write code directly inside the smartbody to implement the desired functionality without having to make traditional HTTP request calls. This is especially important because on most smartbody platforms, it becomes critical to implement functionality directly within the smartbody since code nodes typically do not allow for the introduction of third-party libraries other than internal dependencies.

image

In addition, I think it's essential to stay on top of the cloud offerings of large tech companies. Personally, I am used to browsing new products and services on Tencent Cloud in my free time, not only because I serve as a promotional ambassador for Tencent Cloud Servers, but also because I know that if I don't know about these products myself, how can I effectively promote them to others? At the same time, as the trend of enterprises going to the cloud becomes more and more obvious, familiarizing myself with these popular cloud products in advance will undoubtedly have a positive impact on my career development.

image

Of course, such learning and exploration will inevitably take up some of one's spare time, but the investment is worthwhile in the long run, as it not only contributes to the enhancement of one's skills, but also strengthens one's competitiveness in the workplace.

picture (e.g. of life in the city)

Other tools that have significantly reduced my work time include mapping software. In this regard, I used to use MidJourney for up to a year, especially during its hottest period, and I was a loyal user for a while. However, due to the high cost of MidJourney during its peak period and my limited budget, I eventually had to look for alternatives. Luckily, I found an excellent product in the domestic market - Mixed Yuanbao, a mapping tool launched by Tencent.

image

The main thing I like about Mixed Meta is its unique interaction. It allowed users to continuously modify and refine the generated images through a dialog. This feature was not common in similar products at the time, and it definitely increased the flexibility and efficiency of the creation process. Since I started to use Mixed Meta, I stopped paying attention to other drawing tools, because this software has fully satisfied my needs.

image

Since I've been working on writing articles, I sometimes use AI-generated images to make them look better, especially for the first image of the article title.AI generates images very quickly, and you can get the image you need almost instantly, so you don't have to spend a lot of time searching for the right image. In the past, whenever I was ready to publish an article on WeChat, I had to spend a lot of effort to pick the matching image. Not only did I have to solve the copyright and watermark issues of the image, but even when I found a satisfactory image, it was often not in HD, which was very frustrating.

Now, using AI to generate images solves these problems completely. By simply typing in descriptive text, you can quickly get a high-quality image with no copyright hassles or lack of resolution to worry about.

summarize

Finally finished, although I did not find the so-called "programmer money tips", but through the application of AI technology, it does bring a lot of convenience and change for my work and life. Whether it is to improve work efficiency, expand knowledge, or stimulate creative inspiration, AI has shown its power can not be underestimated.

I hope my experience and insights can bring some inspiration to friends who are exploring AI technology, so that we can meet the challenges of the future together and seize every opportunity for growth. In this ever-changing era, the only way not to be eliminated by the times is to keep learning and progressing.


I'm Rain, a Java server-side coder, studying the mysteries of AI technology. I love technical communication and sharing, and I am passionate about open source community. I am also a Tencent Cloud Creative Star, Ali Cloud Expert Blogger, Huawei Cloud Enjoyment Expert, and Nuggets Excellent Author.

💡 I won't be shy about sharing my personal explorations and experiences on the path of technology, in the hope that I can bring some inspiration and help to your learning and growth.

🌟 Welcome to the effortless drizzle! 🌟