Life Before GPT3 Appeared
when i write blogs and papers
I thought: ah~~~~
What to do, Masaka, I really will never be able to knock out the results!
me after writing
I can finally sleep! ! !
God, is this the life I should have as an undergraduate (￢_￢)
As a result, when I went out to find a job in the Internet, I either met a liar, or I didn’t need my second-hand trash. In the end, I had to go to do business operations just for a broken internship that copied foreign experience.
Sure enough, writing code and blogging can only be a hobby, and life still has to bow.
GPT3 turned out
Xiaohongshu, Zhihu, Douyin: "It's over, our workers will be replaced by AI"
In fact, even if there is no GPT and a unified language model, the value of people is not only low or even non-existent, they may not be replaced by AI, but they will definitely be replaced by more obedient and hardworking people.
Actual: All industries have changed and jobs have changed
When I write a blog, I no longer need to search around to find a new idea. chatgpt itself can extract information from the Internet.
I suddenly thought about changing the language to write an algorithm. I don’t need to bother to find related ones, and then try to change it. I just need to ask chatgpt to give me a simple example, and I will follow as a student. According to the teacher's instruction, I slowly changed the template I needed.
Essay writing? -> You ask questions, chatgpt patiently help you to change
Copywriting planning? -> Various convenient instructions and models appear, chatgpt helps you refine
In short, not only will AI not replace people, it can even be said to further reduce your learning costs and time costs.
The time when one person is a team has come
We no longer need to be attached to studios or companies, just to learn about interpersonal relationships, for the anxiety of daily life, and for the lack of daily skills and knowledge.
limitations of use
1. The problems it can solve can only be part of the mathematical problems, and part of the computer problems that can see the content of the solution.
2. It can only provide ideas, and the person who provides the questions is still yourself, and you must refine the questions clearly
For example, if I write an abstract of a thesis, I hope that chatgpt will polish it. If the paper does not follow the standard or the template that has been shown to be effective, chatgpt itself can only choose nonsense.
3. chatgpt does not belong to anyone, nor does it belong to anyone.
Technology is human, we can only use it, and we cannot freely choose to let it answer any questions and solve our problems.
4. Insufficient data -> Still unable to think like a human and try and error
The written code, if there is not enough data, it can only be written casually, and it still looks good.
If you play tricks on it, it may choose to make up a story just because its knowledge base is not up to date.