Big Tech Programmers: AI is turning us into high-speed assembly line workers, we can't take it anymore

Never would have thought that programmers at big companies are really becoming assembly line workers.

This is not just someone being alarmist, but rather the personal experience of many programmers at Amazon recently—

From early on, Amazon officially encouraged everyone to use AI in their work.

As a result, it has developed to the point where “encouragement” feels more like a form of coercion: senior management believes AI can improve efficiency, so many teams have had their personnel cut in half, while project deadlines have been significantly moved up.

This has led many programmers to not just have AI assist them, but to rely on AI to write code directly for the sake of efficiency.

However, as we all know, AI programming cannot achieve 100% accuracy.

Recently, Microsoft’s well-known open-source project .NET Runtime faced ridicule from programmers worldwide because the newly released Copilot code agent attempted to help automatically fix bugs, but ended up making things worse!

The situation at Amazon is similar; internal engineers using AI to write code not only face concerns about accuracy but also force programmers to become “code reviewers,” requiring them to spend a lot of effort checking and fixing bugs.

It is far from the imagined scenario of “using AI to liberate oneself and engage in more advanced tasks.”

At this point, Amazon engineers' work has become more complicated, but the hard work is theirs, while the credit goes to AI, leaving them unsure about their future career advancement…

Who wouldn’t be worried in such a situation?

Sometimes they have no choice but to let AI write complete programs

Over the past year, Amazon's management has encouraged internal engineers to use tools like Microsoft Copilot and Amazon's own AI assistant.

The reason is simple: it is of course with the goal of “increasing productivity.”

However, this kind of “encouragement” seems to have turned into a form of “having no choice” in actual work.

On one hand, for various reasons, some teams have had their personnel nearly halved, but the amount of code everyone needs to produce has not decreased.

This means that each person must write more code than before.

On the other hand, with AI assistance, management has very “thoughtfully” moved up the deadlines for each project.

In other words, code that originally required several weeks to develop now has to be delivered within a few days.

There’s no choice; developers sometimes can only let AI write complete programs.

The result is that employees report they must rely on AI to keep up with project timelines, or else their performance will be affected, and they may even risk being let go.

Faced with this situation, many people are suffering in silence.

An Amazon engineer who wished to remain anonymous bluntly stated, "We basically have to be forced to accept using AI to generate code, reducing the discussion time about projects and code, and immediately putting solutions into production."

Writing code is much more interesting than reading code.

But when you use AI tools, reading code _ (reviewing code) _ replaces writing code and becomes your main job.

Moreover, the current AI programming tools are clearly not perfect; they often encounter recurring programming issues...

Despite this, a current Amazon engineer stated in an interview that many of the memos and testing software are now written by AI.

In traditional workflows, these tasks are generally assigned to junior engineers, allowing them to gain experience through practical tasks.

Everyone is concerned that these processes, once automated by AI, will cause engineers to lose important skills, and opportunities for promotion will significantly decrease.

In response to this situation, Harvard University professor Lawrence Katz also expressed his disagreement.

In his view, excessive use of AI may prevent beginners from mastering basic skills, "If AI writes the first draft for them, they may never grasp the fundamentals of programming."

Within Amazon, an organization called "Amazon Climate Justice Employees" has begun to voice concerns for anxious engineers, focusing on topics such as "the impact of AI on work" and "the career prospects of engineers in the AI era."

Amazon Official: AI Has Saved Us Thousands of Years of Development Time

Amazon is not the only large company experiencing this situation; Google, Shopify, and others are undergoing similar issues internally.

Some technology historians have compared this phenomenon to what happened in the 19th century.

At that time, craftsmen were assigned to specific positions, focusing solely on handling small-scale segments of work, and it was repetitive.

The result was a faster work pace, while the quality of work declined.

This feels similar to the experience of using AI in coding today: fast, modular tasks replace deep, independent thinking.

But companies and management clearly do not think so.

Amazon CEO Andy Jassy praised a study conducted by Microsoft and others in 2023, which found that AI assistants could improve programmer productivity by over 25%.

At that time, Andy Jassy stated that "speed" is key for a company to maintain a competitive advantage, and he praised generative AI for truly saving a lot of costs.

In response to the complaints from internal engineers, Amazon spokesperson Brad Glasser told the New York Post that the goal of AI is to enhance engineers' professional capabilities, not to replace them; the internal promotion channels remain very clear and will still be considered based on individual performance Amazon officially stated that the company will regularly evaluate teams and increase staff as needed; with the development of AI technology, Amazon will continue to adapt to the changes of integrating AI into workflows.

The official also mentioned that Amazon employees have given positive feedback on the company's investment in AI _ (manual dog head) _.

Indeed, not everyone holds a pessimistic attitude towards the large-scale involvement of AI in work. Some feel that AI has freed them from tedious tasks—such as updating old code and writing tests—allowing them to focus on advanced development.

"AI has saved us thousands of years of development time," Andy Jassy wrote in a letter to shareholders.

Netizens engaged in heated discussions about this.

Some said that using AI to write code only seems to save time, but in reality, it doesn't make things easier at all.

After using AI for programming, you still have to debug it yourself... In short, you always end up needing to manually resort to old methods to complete programming.

Of course, there are also those who stand on the other side:

I'm not saying this, but software engineers generally know nothing about the power of general artificial intelligence.

This statement was quickly rebutted by those who feel that AI programming is not useful.

The counterargument is:

You're saying this because you're not professional and are just writing some very amateur code to complete very small projects~

Risk warning and disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investing based on this is at your own risk