Fragmented Podcast

Turning vibe coders into software engineers, one episode at a time. We share our learnings on how AI can augmenting developer, not
replacing their craft.

[Kaushik Gopal @kaushikgopal]

We all know @pidotdev is a masterclass in nailing the core primitives. How do they do it?

On the latest @FragmentedCast, I had the distinct honor of finding out the answer from the legends themselves - @badlogicgames & @mitsuhiko of @earendilworks.

They help unpack the engineering philosophy behind Pi and its new library, Pi Durable.

I learnt a tonne, and they helped formulate a few rules I'm going to carry strongly forward in this AI coding era:

1️⃣ Keep primitives simple enough to fit in your head.
2️⃣ Agents can deal with tedium no sweat; go with explicit state declarations over automagic checkpointing.
3️⃣ Sometimes, you just can't machine learn your way out of software engineering.

They shared a tonne of other insights in the most jovial way possible ❤️

Grab those headphones and listen to the full episode 👇

https://t.co/65ocNkNgeH

YouTube:
https://t.co/UIqztXfJ83

Spotify: https://t.co/z40YkShIFh

[Kaushik Gopal @kaushikgopal]

Self-learning harnesses are here.

After @deepseek_ai released their new harness, even the authors of Pi & OpenCode said there were some super compelling ideas in there. But what exactly?

@IurySza and I take a shot at explaining it in @FragmentedCast

[iury souza @IurySza]

This week @kaushikgopal and I had the pleasure to chat @mitchellh on the pod🎙️!

Refreshing to hear someone of his caliber bring such a grounded perspective to agentic coding.

We also talked about Ghostty, and how terminal performance gains make tools like Claude Code possible. (He even explains what's behind scrollback perf issues).

A lot of gems in this one. Check it out!

[Fragmented Podcast @FragmentedCast]

The calendar alerts he sets for warm starts. Hail Mary prompts across GitHub issues. Why AI leaderboards are BS and rarely translate to value. A vibecoder he respects who's actually shipped.

Listen to that and more 👉
https://t.co/bi7OJfdCi8

[Fragmented Podcast @FragmentedCast]

🎙️ Our first guest in the AI series is the legend @mitchellh 🎙️

We covered a lot of ground and learned a tonne from him: 👻 Ghostty's internals and why tmux & certain shells strip its best features. His path from AI skeptic to cautious optimist.

[Fragmented Podcast @FragmentedCast]

We're going to be releasing ep 310 (our 10th episode in the new series) in the next day or two.

🎙️ it's an ultra special one cause we're starting up the guest series again... and boy do we have a guest for you.

stay tuned!

[Fragmented Podcast @FragmentedCast]

Listen to episode 309 - Background Agents

https://t.co/HYUU3TO8R8

[Fragmented Podcast @FragmentedCast]

Karpathy's goal for agents: maximize how long they run without intervention.

While individual speed is going up, "system" speed isn't (the false summit ⛰️).

In ep 309 we talk about resolving this tension with Background Agents!

[Fragmented Podcast @FragmentedCast]

More than you'd think!

Listen to @kaushikgopal & @IurySza talk about this in episode 308, and find out.

https://t.co/zfuUrNwep6

[Fragmented Podcast @FragmentedCast]

Our LLM explainer (ep 303) got pretty popular, so we thought we'll follow up with the natural progression - how do image models work - in 20 minutes!

also... what does Michalangelo have to do, with understanding stable diffusion models? 🤔

[Fragmented Podcast @FragmentedCast]

@kaushikgopal & @IurySza dug into:

- shaping the agent environment & 5 themes for harness engineering
- then on building a harness

For your ears 🎧

https://t.co/G5y0ExEp1x

[Fragmented Podcast @FragmentedCast]

OpenAI spent 20% of every week cleaning up AI slop from their own coding agents. So they built a "harness" shaping the agent's environment so the slop never ships in the first place.

New episode 307 on Harness Engineering !

[Fragmented Podcast @FragmentedCast]

Listen to our newest episode: Keeping your agent instructions in sync and effective

https://t.co/T2ygBv2Ned

[Fragmented Podcast @FragmentedCast]

Are AGENTS[.]md files even useful? There’s finally a paper asking the obvious question.

@kaushikgopal and @IurySza get into the paper, the sync problem, and the small amount of discipline required to stop your agent setup from turning into a mess.

[Fragmented Podcast @FragmentedCast]

No episode this week dear listeners but fun newsletter with that story and 10 promised tips from @IurySza and @kaushikgopal
https://t.co/e8yjtQBNI7

Subscribe to our newsletter for more of this content!

[Fragmented Podcast @FragmentedCast]

We almost shipped an episode with a silent master track 😬

Adobe Podcast returned 58 minutes of completely silent video for one of our episodes. No audio. Just two guys miming a podcast. ��

Here's how we recovered it 🧵

[Fragmented Podcast @FragmentedCast]

A script that analyzed mouth movements frame-by-frame from the silent video, correlated that motion signal against audio energy bursts in our local recordings, and recovered the exact timing offsets.

... only felt like this simplest solution 😅

[Fragmented Podcast @FragmentedCast]

If you're curious to learn more about subagents, new @FragmentedCast ep out with @kaushikgopal & @IurySza!

what subagents actually are, a convenient RTS scout analogy, when to spawn them vs when not to. all in this week's episode!

https://t.co/UCevKBz1No

[Fragmented Podcast @FragmentedCast]

Master subagents and your AI coding sessions stay sharp for hours with clean results and no context pollution.

Under the hood though? they're just a tool call. One model prompting another in a scoped context. 💡

[Fragmented Podcast @FragmentedCast]

thanks for sharing our episode last weekend @androidweekly 🙌

Ep 303 explains how LLMs work in a simple and relatable way. It really is one we'd love all devs to listen to.

[AndroidWeekly @androidweekly]

Latest Android Weekly Issue 713 is out!
https://t.co/8aHjC5QPjg #AndroidDev #Kotlin