lunatechian (lunatech-ian)

the moat is the system, not the model

nospam@example.com (Raj Shekhar) — Sat, 11 Apr 2026 17:41:11 +0000

An interesting comment from hacker news in the discussion on Small models also found the vulnerabilities that Mythos found

This is from the first of the caveats that they list:

> Scoped context: Our tests gave models the vulnerable function directly, often with contextual hints (e.g., "consider wraparound behavior"). A real autonomous discovery pipeline starts from a full codebase with no hints. The models' performance here is an upper bound on what they'd achieve in a fully autonomous scan. That said, a well-designed scaffold naturally produces this kind of scoped context through its targeting and iterative prompting stages, which is exactly what both AISLE's and Anthropic's systems do.

That's why their point is what the subheadline says, that the moat is the system, not the model.

Everybody so far here seems to be misunderstanding the point they are making.

Link to the comment

Job reports

nospam@example.com (Raj Shekhar) — Fri, 06 Mar 2026 17:21:14 +0000

From https://ritholtz.com/2026/03/10-friday-am-reads-487/

ADP Jobs Report Shows White-Collar Losses in February: The professional and managerial class is getting hit hardest in the latest hiring data. The college-educated job market continues to deteriorate even as GDP holds up. (Quartz)

Checking on the Quartz article

The topline number — 63,000 private-sector payroll jobs added in February, for the best showing since July 2025 — appears, at first, to be good news.

But glancing even just a few lines down the report changes the story and brings something alarming into view, with the ADP table showing that professional and business services shed 30,000 jobs last month. This is the category encompassing lawyers, consultants, accountants, marketers, and administrative positions — a broad swathe of the white-collar knowledge economy, at least within this type of reporting.

The ADP report is gone, but it's available on archive.org

Some vitamix hacking

nospam@example.com (Raj Shekhar) — Mon, 17 Aug 2026 08:01:27 +0000

Had to work around the "SmartDetect" feature that Vitamix has to make it work again. Documented the steps here

Good Engineering Requires Failed Experiments

nospam@example.com (Raj Shekhar) — Sat, 09 May 2026 01:27:30 +0000

Interesting quote from a John Carmack tweet (xcancel link)

When failure is not an option, success can stay very expensive. You need to experiment to improve, and that fundamentally means being comfortable with failure. If you know it is going to work, it isn’t an experiment.

Start With What You Know

nospam@example.com (Raj Shekhar) — Sat, 15 Aug 2026 00:27:04 +0000

Quoting Daniel Stenberg

I decided that instead of trying to do this the best and optimal way – I shouldn’t let perfect be the enemy of good – I would start out by doing the things I know how to do and take it as far as I can one step at a time. Something should be better than nothing.

Interesting thought process from a great hacker and engineer.