Github Copilot has finally arrived on the front lines. As a mid-level engineer, I'm very happy, but...

I work at a typical IT company, and we have finally received the order to “please use Github Copilot for development.”

Even in Japanese companies, which tend to be slow to adopt AI, we have finally reached this situation.

In this article, I would like to talk about what I have actually felt while coding on the front lines and my outlook for the future.

📝I'm very happy, but...

When I first started using Github Copilot, my initial reaction was

🧑‍💼“It can write such complex code in an instant!? And it actually works!”

that was it.

The products we handle are a bit tricky, and it's hard to write code if you're not used to them.

I also struggled quite a bit to get used to it, but the AI doesn't care about that at all.

🧑‍💼“Anyway, this has made the development process easier!”

That made me happy.

However, when I look around, there are many new graduates and employees with only 2-3 years of development experience.

At that moment,

🧑‍💼“I feel sorry for these kids because they have fewer opportunities to learn by doing it themselves...”

is what I thought.

Even though there are many ways to study, I still believe that the knowledge and experience gained by working with your own hands and through trial and error are things that can be used for a lifetime.

Things like how to recover from failures and how to apply things in advanced ways are gained through such steady accumulation, I think.

However, I feel that AI has robbed them of the opportunity to engage in that trial and error.

Right now, it's fine because mid-level employees who can read and write code are doing the code reviews, but what will happen when the current 1-3 year employees become mid-level?

📝 I've thought about what will happen from here on out

From a developer's perspective,

I think these are the major issues.

💻 Development man-hours and personnel will be significantly cut

First of all,development man-hours will be set lower than they are nowas a matter of course.

It wouldn't be surprising to hear demands like these at any time.

In fact, I'm sure many of you have heard news from overseas about mass layoffs of developers.

Companies should reduce development man-hours to make money more efficiently.

However, I think every company is currently in the process of figuring out how to reconcile this with the challenge of 'training employees with little coding experience'.

🤖 Demand for talent capable of fact-checking will skyrocket

I think this is true regardless of the industry.

However, for system engineers, to perform fact-checking,knowledge of coding is essential.

This means that, in the end, we need talent thatcan write and read code.

In Japan, at least in large companies, delivering products of this quality to users is not permitted.

This is an extreme example, but I fear that such incidents will become increasingly common.

💰 The gap between those who can handle upstream processes and those who cannot will become more pronounced

First of all, I want to state thatthis is not a discussion about 'which is better, upstream or downstream processes!'that I want to convey.

I know the joy of both, and my boss, who is a PM (Project Manager), often says

👨‍🦲“Ah, I want to code...”

so I think it just depends on the person, haha.

There are people in every company who are not PMs or PLs like my boss, but rather so-called "craftsmen" who make a living solely through coding skills.

However, now that we live in a world where anyone can code in a flash, I feel that the value of those people is relatively declining.

In this world, the value of work that anyone can do decreases, so the value of people who can handle upstream processes (= people who can solve customer problems) increases.

This means that a wage gap will emerge.

With the arrival of AI, it has become harder to get paid for coding.

📝 What companies/workplaces should do now

So, what should we be doing now?

“Let's understand the characteristics of AI and learn to use it!”

I won't write that, as it's already too obvious, but

are things I think we should put particular effort into.

⏱️ Securing time for coding education for new employees

If we end up flooded with talent who say, “I can write code, but I can't read it, haha,” it could lead to a decline in quality and, eventually, even lawsuits.

“It's AI's fault...”

You can't say that, after all.

For those in their first to third year or those about to enter the workforce, don't rely solely on your company; engage in self-improvement so you can acquire coding skills at a level that allows for fact-checking.

🚫 Sharing the risks of the AI development era in advance

Information leaks and liability issues, there are several risks unique to AI development.

If you don't take measures to avoid such risks in advance, I think that in the worst case, it could develop into an irreversible situation.

AI is a very convenient tool, but if you jump into it without thinking, you could get badly burned, so I believe it should be handled with particular caution when used in a corporate setting.

🧠 Proactively take on upstream engineering tasks

I feel this is true from both a corporate and an individual perspective.

If there is an opportunity, you should definitely go for it.

On one project, I was hired as a developer, but I voluntarily supported my supervisor's work and gradually took over their tasks as well.

My supervisor was grateful because they had other work to do, and I was able to gain experience with upstream tasks, so it was a win-win relationship.

If you have a desire to "earn more," skills in upstream processes are essential, so you should be greedy and take on the challenge.

📝 In conclusion

I wrote this article because major companies have finally started using generative AI.

Writing this made me keenly aware that we are in an era where skills are in higher demand than ever before.

It is an era where the gap between those who take action and those who don't is likely to widen even further, but precisely because it is such an era, let's enjoy it and challenge ourselves with various things.

If you found this even a little bit helpful, please be sure to follow and like.

See you next time 🙌