Developers are Split on Claude Code

Are AI agents really making you more productive? Plus, the search for silver bullets.

Ben Putano
August 13, 2025

The Daily Grind is brought to you by:

Damn Gravity: Book for Builders and Entrepreneurs

Get 15% off all books

Good morning!

Welcome to The Daily Grind for Wednesday, August 13.

Today’s headline actually comes from the comment section of a headline. Yes, that’s where the real news lies.

We’re talking about Claude Code and the search for silver bullets.

Let’s get into it.

📰 One Headline: Developers are Split on Claude Code

For today’s headline, I planned to cover Claude’s new, larger context window that allows users to run large-scale code analysis, document synthesis, and context-aware agents.

But I found the comments on the story on HackerNews much more interesting.

Claude Code is changing the way developers and computer engineers work, but results have not been evenly distributed. Some users claim large productivity gains, while others call AI coding a “net negative.”

The split of opinion shows how early we still are in the AI era. Below are some of the more interesting debates:

Is a larger context window even useful?

Part of the frustration around Claude Code is its inability to perfectly remember and maintain the context its given. Without that assurance, a larger context window only opens the way for more errors and problems.

“Short of evals that look into how effective Sonnet stays on track, it's not clear if the value actually exists here,” said user aliljet, who had the top comment on the story in HackerNews.

This comment set off a waterfall of discussion on how to maintain effective context, and whether or not one should expect an LLC to maintain full context at all.

“There's a tradeoff just like in humans where getting a specific task done requires removing distractions. A context window that contains everything makes focus harder,” responded user ants_everywhere.

Do coding agents really make you more productive?

One of the most widely debated questions on the story was whether or not coding agents like Claude Code were actually improving productivity. Some users shared that even after saving time on a coding project initially, they spent even more time fixing phantom bugs that ended up in the codebase.

There were also reports of Claude spellchecking lines of code and changing their meaning for no reason.

User benterix is one of the users that has had an overall bad experience with coding agents:

“Having spent a couple of weeks on Claude Code recently, I arrived to the conclusion that the net value for me from agentic AI is actually negative.

I will give it another run in 6-8 months though.”

Benterix was not alone in the sentiment, but many users reported productivity gains as well. One user claimed “3x” improvements—which other users immediately asked to be verified—while others saw productivity gains in some areas, but not others.

But even here, users did not agree.

weego: “My biggest take so far: If you're a disciplined coder that can handle 20% of an entire project's… time being used on research, planning and breaking those plans into phases and tasks, then augmenting your workflow with AI appears to be to have large gains in productivity…

If you're more inclined to dive in and plan as you go, and store the scope of the plan in your head because ‘it's easier that way’ then AI 'help' will just fundamentally end up in a mess of frustration.”

cmdli: “My experience has been entirely the opposite as an IC.”

So which is it? Are coding agents better for well-planned projects on known subjects, or for exploring new domains? The jury is split.

Using Claude Code is a skill all its own

Most users seemed to fall somewhere in the middle on all these debates. Claude Code worked great for some projects, but not others. It saved time here, but not there.

The biggest takeaway for me is that Claude Code is a tool that needs to be mastered like any other.

The most succinct point came from user wiremine, a CTO:

“[W]e're giving people new tools, and very little training on the techniques that make those tools effective.

It's sort of like we're dropping trucks and airplanes on a generation that only knows walking and bicycles.

Those who practice with the truck are going to get the hang of it, and figure out two things:

1. How to drive the truck effectively, and

2. When NOT to use the truck... when talking or the bike is actually the better way to go.

We need to shift the conversation to techniques, and away from the tools. Until we do that, we're going to be forever comparing apples to oranges and talking around each other.”

Claude Code and other AI agents (and all generative AI, for that matter) are still in the “Silver Bullet” stage of the hype cycle. The tools are so impressive that people rush to use them for everything.

As wiremine said, software teams need to shift from comparing tools to comparing techniques—learning how and when to use the tools to serve their ends.

And it will take time.

“I'm six months in using LLMs to generate 90 [percent] of my code and finally understanding the techniques and limitations,” said user jf22.

This is a hard lesson for AI optimists, and may be another signal that we’ve entered the trough of disillusionment era for AI.

But as I said yesterday, once the hype dies down, real progress can begin.

🔗 A Few Good Links

📚 One Page: War of Art by Steven Pressfield

If you find yourself using AI coding agents as a silver bullet, it might be time to pick up War of Art by Steven Pressfield.

Pressfield, a writer and author, discusses the challenges of creative pursuit: resistance, ego, and making the leap from amateur to professional.

In the excerpt below, he reminds us that amateurs think they can “pull off the big score without pain and without persistence.”

Like writing, learning to use Claude Code will take time, pain, persistence, and patience.

❓ One Question: Where are you looking for a silver bullet instead of doing the work?

I wish I could remember where I heard this question, because it’s a great one.

Think about an area in your life where you are feeling stuck. It could be professional, personal, or about a hobby of yours. You are no longer making the progress you once were.

Ask yourself: Are you looking for a silver bullet that will magically fix the problem?

We often do this without even realizing it. We fall into either defeatism or indeterminate optimism—either way, we’ve given up on truly doing the work to get ahead.

Once you see where you’re looking for silver bullets, you can put your head down and get back to work.

🗳️ Wrap Up and Feedback:

That’s it for day! As always, please share your feedback for today’s newsletter: