AI = Artificial Intelligence
https://www.browndailyherald.com/article/2024/02/computer-science-economics-experts-discuss-ai-influence-on-job-market-amid-tech-layoffs
https://www.axios.com/2024/01/18/tech-layoffs-ai-2024-google-amazon
https://edition.cnn.com/2023/07/04/tech/ai-tech-layoffs/index.html
https://time.com/6248290/microsoft-layoffs-ai/
https://amherststudent.com/article/cs-majors-troubled-by-tech-layoffs/
https://www.businesstimes.com.sg/working-life/layoffs-and-ai-are-changing-techs-once-invincible-job-market
https://en.wikipedia.org/wiki/Technological_unemployment
https://www.youtube.com/watch?v=9Z1rMgu2LV0
https://www.youtube.com/watch?v=ddpxy8qAZV8
https://www.youtube.com/watch?v=RLZcdftYNr4
www.youtube.com/watch?v=qrvK_KuIeJk
Most software developers or programmers in the current sense may become obsolete due to AI, and many of them will become or be replaced by "prompt engineers" who ask generative large-language models to design algorithms and write code (and maybe test the software) for them. AI will most likely take over most junior-level developer roles, so the field of software engineering may become increasingly competitive - though while a future "coder" still needs critical and logical thinking, communicating clearly with humans might become equally important because AI conversations become increasingly human-like. Artificial intelligence will probably also affect the jobs of technical writers and documentation specialists, as it can take up most of the actual writing work.
At the current level of maturity, artificial intelligence (AI) can definitely assist programmers in boosting their productivity by providing helpful hints, generating code snippets, and automating certain repetitive tasks. This can help programmers work more efficiently, ultimately leading to more output in less time. Thus individual programmers become more productive with the help of AI and as a result there may be a decrease in the overall demand for programmers.
At the same time it is unlikely in the near future that AI will be able to autonomously develop fully finished software products or complete complex part of projects based solely on text business requirements. Like it can do any professional programmer and even many with junior level
no there is still space for programmer in the market after AI revolution
First and foremost: AI is not Intellect. Today it is not. Can we call computational algorithms created by human (i.e. intelligence) intelligence? Likewise, a program that finds an extremum by a derivative is also intelligence. So my cat is more intelligent than this program.
About the horror in the future, let me bring an analogy: did coachmen disappear from the world when cars replaced horses? Yes. But who are more (even in percentage, even taking into account the years): drivers now or coachmen in old time? How many drivers and at what level were needed after the advent of cars? Have they all become “prompt engineers”? These superficial discussions about AI are scary - what material will future neural networks be trained on without "human" programmers? First they must become Intelligence and then let's be afraid. But today they are not.
Yes, We everybody will have to relearn, and a lot of it - ask former coachmen about this. But ask the drivers - aren’t their cars alive for them?
The short answer is “no” in foreseeable future.
Regarding some optimism about Generative Pre-trained Transformers (GPTs) and Large Language Models (LLMs), I am sorry to disappoint you. Such technologies are not suitable for automatically generating reliable code. And they are not suitable for this in principal, regardless of their maturity. They are very useful for other tasks: natural language translation, simple refactoring of existing code etc., but not for tasks requiring deep conceptualisation. It’s because GPTs and LLMs have no logical reasoning model.
Current GPTs and LLMs are based on just neurocomputing. To generate programs automatically, AI should be based on neurosymbolic paradigms. But this is a long story.
From my everyday experience, I see how a GPT fails to complete simple coding tasks, generates obscure code and inserts bugs.
However there are some nuances of what we mean by “coders”. If we mean narrow coding, i.e. just coding without other important integrated steps in software development, then such coders have already become obsolete even without involvement of AI. If we mean broad coding, including integrated steps of requirement elicitation, conceptualisation, design, implementation, debugging, testing, evaluation, deployment etc., then the demand for such skills getting higher and higher, especially when AI grows.