hatmanstack

Why are we using cli wrappers if you're using Claude Code? I get if you need something like Codex but they released sub agents today so maybe not even that, but it's an unnecessary wrapper for Claude Code.

I don't get why people need a cli wrapper for this. Can't you just use Claude skills and create everything you need?

No brother, the Claude plans aren't the right path, they're for hobbyists.

Completely agree. Until recently I only let LLMs write my commit messages, but I've found that versioning the plan files is the better artifact, it preserves agentic decisions and my own reasoning without the noise.

My current workflow: write a detailed plan first, then run a standard implement -> review loop where the agent updates the plan as errors surface. The final plan doc becomes something genuinely useful for future iterations, not just a transcript of how we got there.

Too much free stuff already and anything new will eventually become free. I'd rather wait or direct money to the projects I support.

this.

That's why I'd love to get stats on load/hardware/location of where my inference is running. Looking at you Trainiuim.

Where are my glow sticks?

That's the first piece by Didion I've read, after her death, I'd always meant to read her more. The mask-less account was refreshing, the only counter-weight to flower-power I knew about was Altamont, was getting heavy Hunter S. Thompson vibes.

Does anybody have a better understanding of activation capping? Simple cosine similarity?

I've always been a Dilbert fan, didn't get to any of his books until later. I think Scott was someone unafraid to share his thoughts, unfiltered.

They were valuable to me because it gave me perspective on a way of thinking I would never have considered. I disagreed with the majority, but some had the subtle beginnings of truth that helped to expand my world view.

I'm grateful he was part of the world, and will miss his comedy.

Epic training for reading output tokens.

This looks impressive, haven't had a chance to give it a go, would love a consumable "counters" tutorial for this type of intrusion..."Be a researcher, not a criminal." might be wishful thinking.

Had no idea this existed and IMHO what the internet was made for. Well played sir. https://frinkiac.com/video/S12E09/5fI0bxc31gXMaJNjLhbIeyRQP6...

If you want insight into why they haven't deleted "old garbage" you might try, The Age of Surveillance Capitalism by Zuboff. Pretty enlightening.

Hell, even simpler. Termux + Caddy + cloudflared with a domain you own. Serving in 15 min.

Nah mate, it's all about the Wayland Trust model. No keylogging, consent-based screen recording, and no window spying. Isolation.

Full on cringe shivers

The Simpsons did it.

more a schlimazel, Charles Schultzie, Lucy's everywhere

I feel like I'm schelpin' through these comments, it's all mishigas

Not here to schlep for AWS but S3 Vectors is hands down the SOTA here. That combined with a Bedrock Knowledge Base to handle Discovery/Rebalance tasks makes for the simplest implementation on the Market.

Once Bedrock KB backed by S3 Vectors is released from Beta it'll eat everybody's lunch.

Darwin started with survival of the "fit". It changed to "fittest" in later editions.

Life is good.

That's exactly what it is - formalizing and creating a standard induces efficiency. Along with things like AGENTS.md, it's all about standardization.

What bugs me: if we're optimizing for LLM efficiency, we should use structured schemas like JSON. I understand the thinking about Markdown being a happy medium between human/computer understanding but Markdown is non-deterministic for parsing. Highly structured data would be more reliable for programmatic consumption while still being readable.

Am I dating myself by thinking Kurzweil is still relevant?

2029: Human-level AI

2045: The Singularity - machine intelligence 1 billion times more powerful than all human intelligence

Based on exponential growth in computing. He predicts we'll merge with AI to transcend biological limits. His track record is mixed, but 2029 looks more credible post-GPT-5. The 2045 claim remains highly speculative.

This piece is absolutely a strong candidate for David Brooks' Sidney Awards. Thank you for sharing it.

In my experience with Kiro's spec-driven approach it generated massive task lists (12+ tasks with 4+ sub-tasks each). The workflow was decent but it deleted code unpredictably and wouldn't revert changes. Being a full IDE likely diverts resources to UI edge cases rather than core reliability.

So much simpler to just iterate without the puzzle box of tasks. "a sledgehammer to crack a nut"

Jump the line and just install it. who needs to read stuff. https://github.com/ChromeDevTools/chrome-devtools-mcp?tab=re...

Have no empirical feedback but subjectively it reads as though the authors are trying to proof their own intelligence through convolution and confusion. Pure AI slop IMHO.