[islewis]

> The cold-start training procedure begins by prompting DeepSeek-V3 to decompose complex problems into a series of subgoals

It feels pretty intuitive to me that the ability for an LLM to break a complex problem down into smaller, more easily solvable pieces will unlock the next level of complexity.

This pattern feels like a technique often taught to junior engineers- how to break up a multi-week project into bitesized tasks. This model is obviously math focused, but I see no reason why this wouldn't be incredibly powerful for code based problem solving.

[bearjaws]

It's actually pretty hilarious how far into detail they can go.

For example, I made a bot that you could give it a problem statement, and then it would return an array of steps to accomplish it.

Then you could take the steps, and click on them to break them down and add them to the list. If you just kept clicking you would get to excruciating detail.

For example taking out the trash can become over ~70 individual steps if you really drill into the details.

Some of the steps:

Stand close to the trash can – Position yourself so you have stable footing and easy access.

Place one hand on the rim of the can – Use your non-dominant hand to press lightly on the edge of the trash can to hold it in place.

Grip the top edge of the bag with your other hand – Find the part of the bag that extends past the rim.

Gently lift the bag upward – While your one hand stabilizes the can, slowly pull the bag up with the other.

Tilt the can slightly if needed – If the bag sticks or creates suction, rock or tilt the can slightly while continuing to lift.

Avoid jerking motions – Move steadily to prevent tears or spills

[eightysixfour]

This used to be part of one of the intro to engineering courses at my school - write an XX page document describing how to make a peanut butter and jelly sandwich.

[larrysalibra]

This was a homework assignment in my second grade class!

The next day we had to follow our instructions exactly in class to make the sandwich which was hilarious. A formative experience for me!

[amelius]

A dad trying this out on his kids:

https://www.youtube.com/watch?v=cDA3_5982h8

[voiper1]

I've been using that as a test of new LLMs - and do it in a specific style.

[lugu]

This is how I imagine llms are used in robotics, with one or two more levels of description.

[thrance]

This feels like a manual for infiltrated aliens: "How to pass as humans, Vol. I"

[roywiggins]

or for goblins:

https://goblin.tools/

[jrvarela56]

http://www.drawtoast.com/

[Alifatisk]

Is bot something I can try?

[otabdeveloper4]

Yes, an LLM can generate infinite amounts of bullshit if you ask it to.

[criley2]

Imo current models can already break things up into bite sized pieces. The limiter I've seen is twofold

1) Maintaining context of the overall project and goals while working in the weeds on a subtask of a task on an epic (so to speak) both in terms of what has been accomplished already and what still needs to be accomplished

and 2) Getting an agentic coding tool which can actually handle the scale of doing 50 small projects back to back. With these agentic tools I find they start projects off really strong but by task #5 they're just making a mess with every change.

I've played with keeping basically a dev-progress.md file and implementation-plan.md file that I keep in context for every request and end each task by updating files. But me manually keeping all this context isn't solving all my problems.

And all the while, tools like Cline are gobbling up 2M tokens to make small changes.

[jhrmnn]

> Maintaining context of the overall project and goals while working in the weeds on a subtask of a task on an epic (so to speak) both in terms of what has been accomplished already and what still needs to be accomplished

This is a struggle for every human I’ve ever worked with

[mmis1000]

This is probably the biggest difference between people who wrote code and people that should never write code. Some people just can't write several connected progtam file without logical conflict. It's almost like their brain context is only capable for hold one file.

[drob518]

True, but if AI only gets as useful as an average developer, it isn’t that useful.

[pertymcpert]

Yes. I wonder if the path forward will be to create systems of agents that work as a team, with an "architect" or "technical lead" AI directing the work of more specialized execution AIs. This could alleviate the issue of context pollution as the technical lead doesn't have to hold all of the context when working on a small problem, and vice versa.

Shit. Do we need agile AI now?

[Rudybega]

This is kind of what the modes in roo code do now. I'm having great success with them and having them as a default just rolled out a couple days ago.

There are a default set of modes (orchestrator, code, architect, debug, and ask) and you can create your own custom ones (or have roo do it for you, which is kind of a fun meta play).

Orchestrator basically consults the others and uses them when appropriate, feeding in a sensible amount of task definition and context into the sub task. You can use different LLMs for different modes as well (I like Gemini 2.5 Pro for most of the thinking style ones and gpt o4-mini for the coding).

I've done some reasonably complicated things and haven't really had an orchestrator task creep past ~400k tokens before I was finished and able to start a new task.

There are some people out there who do really cool stuff with memory banks (basically logging and progress tracking), but I haven't played a ton with that yet.

Basic overview: https://docs.roocode.com/features/boomerang-tasks

Custom Modes: https://docs.roocode.com/features/custom-modes

[dataviz1000]

Here is the tippy top of my copilot-instructions.md file

```

# Copilot Instructions

## Prompts

### General Coding

- *Boyd’s Law of Iteration: speed of iteration beats quality of iteration*: First and foremost, break every problem into smaller atomic parts. Then make a plan to start with one small part, build it, give the user an opportunity to run the code to quickly check the part works, and then move on to the next part. After all the parts are completed independently, check that they all work together in harmony. Each part should be minimal.

```

With any big problem the LLM responds first with ..... Body's Law of Iteration ..... and proceeds to break the problem into smaller parts.

I've discovered keeping file size under 300 or 400 lines helps. The AI is great at refactoring.

[ethbr1]

Everything that is 1950s is new again: dynamic programming https://en.m.wikipedia.org/wiki/Dynamic_programming#Computer...

[cadamsdotcom]

And it should be powerful for breaking down reasoning chains of thought too.