Share

AI is making “good coders” cheap, but great architects rare. 

Hey Pythonista,


We saw an interesting post from Principal Software Engineer, and Pybites Community member, Josh Engroff: the developers who thrive won’t be the ones who can write the most code, but the ones who can:

  • understand the domain,

  • design robust systems,

  • and use LLMs as power tools rather than substitutes for thinking.

What's gaining value is the ability to:

  • Frame and decompose problems - turn fuzzy business goals into clear technical problems.

  • Model domains well - data models, boundaries, invariants, failure modes.

  • Make trade-offs - latency vs complexity, coupling vs speed, build vs buy, etc.

  • Design systems & APIs - contracts, evolution, and how pieces fit together over time.

  • Provide technical leadership - deciding what not to build, sequencing work, mentoring, risk management.

This kind of observation from an expert in their field just reiterates what we've said all along - LLMs are just a tool, the real thinking is still up to you. You need to know your craft, and know it well.


This is why even with the advent of AI, our coaching is still unbeatable and still yields incredible results.


Read Josh's article in its entirety on LinkedIn using the link below, connect with him to follow his journey, and have a fantastic week.


Bob & Julian

Josh's LinkedIn Article

Marimo creator on our podcast


In this week’s episode I (Bob) interview Marimo co-creator Akshay Agrawal about how Marimo improves the traditional notebook and fits into a modern Python workflow.

We dig into:

  • The problems it solves vs Jupyter (hidden state, multi-purpose notebooks, running a notebook as a module 😍)

  • How its DAG-based design keeps execution predictable

  • Being Git-friendly and great for collaboration

  • Integrations with pandas/polars, DuckDB, uv

  • The design thinking behind the tool, plus open source, books, and more

I learned a ton from this chat — hit reply and tell me what you think after listening. 💬

Watch it on YouTube

Efficient counting in Python 📈


I solved a Bite the other day and used `collections.Counter()` 🙂

Good reminder: the right data structure saves a lot of code. 🐍

In this case: apart from a small helper function, you get the top stock by symbol using `most_common()`

Clean, fast, readable. 🚀

What built-in, stdlib, or tool helped you write way less (and better)? 💡

Reply and let us know or create a post in our community...
 

Master the collections module on our platform

ChatGPT 5.1 Launches


Chat GPT-5.1 launched this week and it seems the focus is on tone and warmth in an effort to make the LLM more enjoyable to talk to. It's interesting to see this "human" focus become a priority!


Even so, and per Josh's message above, LLMs are no substitute for humans. If you currently use LLMs to do all of your coding, this is the perfect time to check out our PDM program.


We'll help you build the understanding and experience you just don't get with an LLM.


Check out the PDM Program

Warren Buffett...

...sent his last email to shareholders over the past week. This is one of the more meaningful points he shared that really resonated with me. This is what Bob and I aim to do every day through Pybites and the calls we always tell people to book with us:


"Greatness does not come about through accumulating great amounts of money, great amounts of publicity or great power in government. When you help someone in any of thousands of ways, you help the world. Kindness is costless but also priceless."


Keep pushing forward on your coding journey with PyBites as your partner. We're here to help you succeed every step of the way.

\n\n\n\n

Email Marketing by ActiveCampaign