d4rkp4ttern

Love it. Like some others here, I've been making html presentations with code-agents and they come out really nice. Other alternatives for that suit different purposes are these:

- slidev markdown-based slides, I often have a code-agent make these. https://github.com/slidevjs/slidev

- typst slides are great looking too, especially with math etc. Again easy for agents to do

One minor nit: Based on the title, I thought this was literally a tool that had something to do with PowerPoint, but "PowerPoint" is used here as a synonym for "Presentation", which is ironic/unfortunate given how many dislike PowerPoint.

I use the Hex[1] app on MacOs for near instant transcription with Parakeet V3. This is how I speak to Code agents at least 80% of the time. The idea is - I tend to be lazy if I need to type lots of details, so speaking lets me get into details that I otherwise wouldn’t and this helps give more useful context to the agent. This often tends to be an unstructured brain dump so I sometimes ask the agent to repeat back what it understood, so I can make sure, and this also likely helps it stay on track.

[1] hex https://github.com/kitlangton/Hex

I used to use Handy but as of a few months ago it had stuttering and other issues so I switched to Hex. Even for relatively long minute long dictation, transcription is near instant and accuracy is more than sufficient, especially when talking to AI since it that can “read between the lines “.

Using voice with code agents is a huge unlock and I’m surprised to see some people I recommend it to, still resist it.

As for TTS I found it fun to make a voice plugin for Claude Code that uses PocketTTS, so it can give brief updates whenever it finishes a turn:

https://pchalasani.github.io/claude-code-tools/plugins-detai...

The other day I was trying to find out exactly how much context size 5.6 has in codex CLI (via subscription) and I didn’t find that documented in any of the obvious places. Sort of ridiculous that we have to find this out in X or Reddit, e.g:

https://www.reddit.com/r/codex/s/hVv29obfFD

In the Reddit post there was clarification that this was done with Sol Pro not Ultra - curious what is everyone’s mental model of the difference.

My understanding is that ChatGPT Pro is effectively a multi agent system, or somehow uses multiple LLMs in parallel and selects a best answer. And Ultra is more similar to Claude-Code UltraCode where the main agent can choose to create a dynamic JS workflow that deterministically orchestrates multiple agents to handle different parts of a task and have adversarial checkers etc.

Is that more or less the difference? Any substantiating sources would be great to see.

Nice, it works for discussing concepts baked into the LLM but fails when you want it to read contents of an article or do web search. ChatGPT Live doesn’t have this limitation. Their Voice mode did have this limitation, but the Live mode released a couple weeks ago works exactly as you’d want, with link-following and web search

Too much friction in notebookLM - several minutes to make a podcast, and they disappear after a while, and the interrupt feature is super-janky.

Related book by Blum, Hopcroft, Kannan:

Foundations of Data Science (2020):

https://home.ttic.edu/~avrim/book.pdf

Yes Claude voice refuses to actually read the article and relies on snippets it finds in web searches and often pretends to have read the article, and when pushed admits it really didn’t. ChatGPT Voice used to have this issue as well, but the new Live mode (released just a few weeks ago) fixed all that, and it actually pauses and reads the article and does real web searches during the conversation.

Actually for me it’s more cognitive load when the AI just dumps everything at once, which is why I prefer to have it give me a nugget at a time, so I can take it in, pause and discuss before moving to the next one.

Me:

    Ok so I'm going on a walk. I'll dump a link to a Hacker News   
    discussion about an article. 
    You have to read the article and the discussion and walk me thru   
    all the interesting details, nugget by nugget, and move on when 
    I'm ready for the next piece.

    ok, Great show me the link, I'm waiting.

Me:

    Ok I pasted it. Now go.

For the Socratic quiz I say:

    I want to understand this more deeply. So instead of you just telling me
    everything, lay out the problem and a question for me to think about, and 
    I'll try to answer. Even if I answer wrong, you should resist giving me the
    answer, and instead keep digging with more questions, so that I eventually 
    arrive at the answer myself.

https://pchalasani.github.io/claude-code-tools/plugins-detai...

When notebookLM was new, it was interesting to listen to the podcasts. Then the novelty wore off, and I wanted something where I can interact with the podcasters but it was janky as hell.

My current “audio-learning” hack is ChatGPT Live which has become shockingly good after being awful compared to Claude Voice (Let’s not even talk about Gemini voice which is still bad).

I go on a walk and dump a paper or article link in the chat, and ask chatGPT Live to walk me through the content in small nuggets, so I can discuss them interactively. For deeper topics I have it quiz me Socratic style so I’m not just passively listening, and actually thinking through problems or ideas.

I've been looking around for skills that help with improving AI prose, and found these:

(1) Avoid AI Writing - https://github.com/conorbronsdon/avoid-ai-writing

(there is a similar "humanizer" skill that (1) subsumes)

(2) Agent style: https://github.com/yzhao062/agent-style

(3) Journal-adapt (style transfer from your writings): https://github.com/WantongC/journal-adapt-writing-skill

I find it useful to set up dynamic workflow loops with (1) and/or (2) as checkers.

On (3) - I think "style-transfer" that was all the range in image-gen is going to be very useful for prose-generation. Curious if there are LLMs specifically trained for this.

I too used to be a skeptic of the whole "agent-swarms" hype. But I do find some patterns useful with ultracode:

- parallelism: "if it makes sense, use ultracode to do this thoroughly" Claude decides if this is the type of task that can benefit from a dynamic workflow. For example where there is a large number of similar independent sub-tasks, it can choose to set up a dynamic workflow, and it even sets up verifiers/tests etc to validate its work. In some cases it pushes back and decides not to set up a dynamic workflow and instead works on the task directly.

- verification loops: "Set up a dynamic workflow with adversarial checkers, and iterate until they are all green, up to some sensible limit". Similarly, I can include code-simplification steps and documentation-update steps.

These patterns help me leave it unattended for longer periods of time and return to a reasonable assurance of correct code.

I’m curious why many tech bloggers choose Substack, when it’s almost trivial to set up your own blog with infinitely more style control, and math/syntax-highlighting etc.

My understanding is that in the generated JS code, the subagents are invoked as headless Claude, equivalent to Claude -p, or the agents SDK.

single line addition to prompt to use subagents proactively.

This misses an important detail. In Claude Code [1], ultracode suggests the agent create a JavaScript code to deterministically orchestrate sub agents. This is different from just having the main agent launch sub agents and (non-deterministically) manage them.

The resulting workflow is called “dynamic” because CC creates this orchestration script dynamically, “on the fly”.

[1] https://claude.com/blog/introducing-dynamic-workflows-in-cla...

Another useful thing about dynamic workflows is you can ask Claude to make them durable as skills (or slash command) that can be invoked later.

I believe inside Google they have a similar concept called “deterministic workflows”.

I find ultracode extremely useful. Of course you have to watch how your 5 hour and weekly session usage percentages are getting used. So I had Claude make a status-line with 3 progress bars: for context window, 5h session, 7d session:

https://pchalasani.github.io/claude-code-tools/tools/statusl...

That’s a good point

100% Also good to note that Cherny and Steipete have said in interviews that they keep it simple and do not use any of these shenanigans.

I didn’t know you could get the “Chat-GPT-5.5 Pro” (the one that’s been solving Erdos problems) inside codex-cli, or maybe I misunderstood?

Ok I’ll ask since nobody else has — are they not giving their devs a Claude code max or Codex Pro subscription? If so, why is token cost approaching billions? And if not, why not?

When I tried handy a few months ago the transcription had stuttering issues, and I later discovered Hex and it was even faster and did not have those issues, so I stuck with it. This was with the same underlying Parakeet model, I think the differences are in the surrounding engine and post processing etc.

Nice collection! Do you know if any of those can do meeting transcription with turn detection with consistent speaker ID ? Otter killers in other words.

My current fav STT is Hex, MacOS only:

https://github.com/kitlangton/Hex

Near-instant accurate-enough transcription using local Parakeet V3.

Handy is multi-platform, also great:

https://github.com/cjpais/Handy

Every time a new STT model or app is announced, I see if it displaces Hex for me but so far none have.

I’ll check out Yap though.

It’s available today in beta for Claude Enterprise and Team customers.

You’re welcome.

The engine decomposes the traces to understand common failure modes across harness executions and produces a report with its findings.

What are some examples of these common failure modes?

Is this a generally well known thing?

TechCrunch articles should be ignored into oblivion.

Thanks, not sure what went wrong, but fixed it now

I sometimes have claude or the coding-agent quiz me, Socratic style.

It persists in asking questions at deeper levels until you arrive at the answer yourself. This forces you to think hard about a problem, and this effort helps with understanding, learning and retention. Of course I made a Socratic-quiz skill for this, to use with any coding agent or similar:

https://pchalasani.github.io/claude-code-tools/plugins-detai...

For example I’ve used this to better understand counter-intuitive things about diabetes/insulin, dopamine and motivation, catching up with a codebase, Claude’s implementations, etc (to combat so-called cognitive debt).

Strong LLMs are surprisingly good at this type of quizzing, they display a semblance of “theory of mind”.

[Edited - link fixed]

You're thinking of the legendary IIT (Indian Institute of Technology) JEE (Joint Entrance Exam).

https://en.wikipedia.org/wiki/Joint_Entrance_Examination