mccoyb

8 ms·

91.
by mccoyb 8mo ago
You can use your Codex plan with it. OpenAI endorsed it several weeks ago, as far as I remember. That could change, however, but now seems safe.
92.
by mccoyb 8mo ago
Pi has made all the right design choices. Shout out to Mario (and Armin the OG stan) — great taste shows itself.
93.
by mccoyb 8mo ago
Here's my question: if agents continue to get better with RL, what is future proof about this environment or UI? I think we all know that managing 5-10 agents ... is not pretty. Are we really landing good PRs with 100% cognitive focus
94.
by mccoyb 8mo ago
Skills feel analogous to behavioral programs. If you give an agent access to a programmable substrate (e.g. bash + CLI tools), you write these Markdown programs which are triggered and read when the agent thinks certain behaviors will be be
95.
by mccoyb 8mo ago
I wanted to write the same comment. These people are fucking hucksters. Don’t listen to their words, look at their software … says all you need to know.
96.
by mccoyb 8mo ago
Does your experience include writing your own agent? Send a link
97.
by mccoyb 8mo ago
Is it? Look at pi, for instance. It turns out that "most of the bell and whistles" could amount to instructing models how to use tools like tmux
98.
by mccoyb 8mo ago
The opposite is true. There is barely any magic in the harness, the magic is in the model. Try it: write your own harness with (bash, read, write, edit) ... it's trivial to get a 99% version of (pick your favorite harness) -- minus the
99.
by mccoyb 8mo ago
"Training the specific harness" is marginal -- it's obvious if you've used anything else. pi with Claude is as good as (even better! given the obvious care to context management in pi) as Claude Code with Claude. This wh
100.
by mccoyb 8mo ago
OpenAI has endorsed OAuth from 3rd party harnesses, and their limits are way higher. Use better tools (OpenCode, pi) with an arguably better model (xhigh reasoning) for longer …
101.
by mccoyb 8mo ago
Broadly agree with the author's points, except for this one: > TypeScript/Node.js: Better concurrency story thanks to the event loop, but still fundamentally single-threaded. Worker threads exist but they're heavyweight OS
102.
by mccoyb 8mo ago
No, I agree! But I don’t think that observation gives us license to avoid the problem. Further, I’m not sure this elevates my understanding: I’ve read many posts on this space which could be viewed as analogous to this one (this one is more
103.
by mccoyb 8mo ago
Effectively everyone is building the same tools with zero quantitative benchmarks or evidence behind the why / ideas … this entire space is a nightmare to navigate because of this. Who cares without proper science, seriously? I look th
104.
by mccoyb 8mo ago
Thank you, very helpful!
105.
by mccoyb 8mo ago
My 5 minute read is that the divergences are primarily in the communication model and in transactions: - the SAM coordinates through the dataspace, whereas Goblins is focused on ("point-to-point") message passing - SAM (as present
106.
by mccoyb 8mo ago
How does this model compare to the syndicated actor model of Tony Garnock-Jones? (which, as far as I can tell, also supports capabilities and caveats for security) Neat work!
107.
by mccoyb 8mo ago
a better term might be “feedback engineering” or “verification engineering” (what feedback loop do I need to construct to ensure that the output artifact from the agent matches my specification) This includes standard testing strategies, bu
108.
by mccoyb 9mo ago
It's not in the harness today, it's a special RL technique they discuss in https://www.kimi.com/blog/kimi-k2-5.html (see "2. Agent Swarm") I looked through the harness and all I could find is a `Tas
109.
by mccoyb 9mo ago
Claude Code gets functionally worse every update. They need to get their shit together, hilarious to see Amodei at Davos talking big game about AGI and the latest update for a TUI application fucking changes observable behavior (like histor
110.
by mccoyb 9mo ago
Here's a separate, optimistic comment about Yegge and Huntley: they are obviously on the right track. In a recent video about Loom (Huntley's orchestration tool), Huntley comments: "I've got a single goal and that is aut
111.
by mccoyb 9mo ago
I think Yegge and Huntley are smart guys. I don't think they're doing a good job incubating their ideas into being precise and clearly useful -- there is something to be said about being careful and methodical before showing your
112.
by mccoyb 9mo ago
Supposing agents and their organization improve, it seems like we’re approaching a point where the cost of a piece of software will be driven down to the cost of running the hardware, and the cost of the tokens required to replicate it. The
113.
by mccoyb 9mo ago
Who knew that these massive high-dimensional probability distributions would drive us insane
114.
by mccoyb 9mo ago
It codes faster and with more abandon. For good results, mix Claude Code with Codex (preferably high or xhigh reasoning) for reviews.
115.
by mccoyb 9mo ago
I’m luckily in a situation where I can afford to explore this stuff without the concerns that come from using it within an organization (and those concerns are 100% valid and haven’t been solved yet, especially not by this blog post)
116.
by mccoyb 9mo ago
I think this read is generous: > something like gas town is clearly not attempting to be a production grade tool. Compare to the first two sentences: > Gas Town is a new take on the IDE for 2026. Gas Town helps you with the tedium of
117.
by mccoyb 9mo ago
My wishlist for 2026: Anthropic / OpenAI expose “how compaction is executed” to plugin authors for their CLI tools. This technique should be something you could swap in for whatever Claude Code bakes in — but I don’t think the correct
118.
by mccoyb 9mo ago
I see -- unoriginal, but it was accepted for publication at a leading programming languages venue? Do some experiments, stop speaking out of your ass.
119.
by mccoyb 9mo ago
There are no concepts in this blog post. It is the author's opinions in the form of a pseudo-Erlang program with probabilities. If one reads it like it is a program, you realize that the underlying core has been obfuscated by implement
120.
by mccoyb 9mo ago
I've tried most of the agentic "let it rip" tools. Quickly I realized that GPT 5~ was significantly better at reasoning and more exhaustive than Claude Code (Opus, RL finetuned for Claude Code). "What if Opus wrote the c

More ›