Merad

Most likely pretty bad. It's not all that difficult to sweet talk Fable into doing a security review of a code base. I've pointed it at a few of the major open source projects in the main tech stack I work with and it found dozens of bugs in each. Many were fairly minor, and some of the more significant issues relied on fairly narrow attack vectors (for example, a compromised DB server being used to attack a web server). But this was just a single search pass, literally a few lines of prompt and an hour's worth of agent time.

You're looking for Young's penetration equations, developed at Sandia National Lab in the 90s. Your favorite LLM can walk through the math step by step if you're interested, but a US WW2 era AN-M64 500 lb bomb gives results of 15-20 ft penetration in average soil. Reality might be a little less because free-falling bombs have forward velocity from the aircraft that dropped them, so they hit at angle instead of penetrating straight down.

This repo I started yesterday has most of the patterns I like to use: https://github.com/mbcrawfo/SqlSource

Really the single biggest piece I use today is using a tool like superpowers to ensure your task is well defined. Added bonuses are its TDD implementation skills and multiple review steps baked into the workflow. Before I discovered superpowers I was manually telling claude to do these workflows, the plugin makes it easy.

In the repo itself:

* Keep agents.md focused; claude likes to really bloat it by default. The "Adding guidance for AI agents" rules go into every project and have worked pretty well.

* Aggressively use nested agents.md files. Focused rules located where they're relevant seem to do much better than having a large whole-project rule dump in the root agents.md.

* Recently I've started using the tech debt and deferred work documentation rules. Claude frequently decides on its own that some piece of functionality or issue is "deferred until later" or "out of scope". These rules help make it clear so that you can make it fix the item or at least have it documented for a followup session.

* Have it test all the things. Most of my personal projects with claude have about a 3:1 ratio of test:application code and high coverage. Claude make it easy to build test harnesses for things that most of us wouldn't have bothered with 5 years ago.

* Strict type checking, linting, static analysis, warnings as errors, etc. Anything you can turn on or add to block questionable code, do it.

* Use a claude pre-tool hook to run all of your checks before you allow a git commit command. Left on its own it can and will decide that it has a valid reason to bypass git commit hooks (husky) to commit broken code or tests.

I'm not being flippant. If I need to sit and watch a claude session reviewing every edit it makes, that's slower than me coding the changes by hand. Either it's good enough to work independently or it's not worth using. I have no idea how Google is using AI, but from what I've seen at my own company I do know that using AI effectively _and_ setting up repos/projects for good AI usage (and with guardrails to keep AI in check) are both acquired skills. Skills that the majority of devs don't have right now.

If you need to review every edit that claude wants to make what's the point in using it? We're now treating it basically like a junior dev. Do our best to make sure that it's on the right path (leaning hard on the superpowers plugin) then let it work and look at the result. Though even that review is heavily AI assisted now. Our leadership has mandated that everything that can be done by AI, should be done by AI. We've only been in that mode for a couple of months but I'm starting to think that it's plausible with proper human steering and oversight.

There's a company based in Kentucky named Big Ass Fans that makes a lot of these. I talked to them at a job fair about 15 years ago, at least at the time they were writing a lot of software for motor controllers, etc. in house. Sounded pretty interesting.

Use the superpowers plugin. It's a game changer. You start by building out a spec defining the work you want to be done, then claude writes a detailed implementation plan. It can easily work for hours without interruption, and the process integrates testing (TDD when possible) and review steps along the way. I even did one fairly complex POC for work that had claude implementing for almost 3 days straight, with me stopping it and handing off to a new session when context hit about 800k. The process is slower and more methodical than the way claude works by default but the output is much better.

The A-12 was built in an era when the aerospace industry was flush with knowledge and active R&D into mach 3+ flight - everyone was convinced that it was the future. Today's industry is heavily optimized for subsonic airliners and fighter jets that have top speeds of "only" mach 1.5-2. No one in the west has built a mach 3 aircraft since the 60s and not in Russia since the early 80s. The knowledge isn't _lost_ (well, we hope) but there is a lot that will need to be relearned, and we all know how even the best documentation is usually missing a lot of details.

Is it able to work with non default apps like Gmail?

I don't think this will work reliably. Pretty sure it's not the LLM deciding to read the files, it's the harness in Claude Code itself injecting the instructions files based on the paths the model accesses.

Edit: sometimes if Claude lists files in a directory or does a search that shows it an AGENTS.md exists it will decide to read it. But it's not a reliable behavior.

I used to work at a company that was the industry leader in business management software for storage facility owners. I don't remember exact numbers but a lot of facilities are mom & pop type operations where people basically invested in the business as a retirement plan. It's a surprisingly stable and profitable industry. Economy good? People buy too much crap and want to store it. Economy bad? People have to move and downsize their housing and need to store stuff. Large incentive for tenants to avoid delinquency because they risk their stuff being auctioned off.

"Space weapons" here most likely means some combination of anti-satellite weapons and electronic warfare such as GPS jammers. AFAIK it's been widely believed that the USSR/Russia has had these capabilities in orbit since the 80s. I'd be willing to bet that the US has had them for decades as well, it just hasn't been publicly acknowledged.

Interesting! Have you tried it on any large code bases? I have to wonder if json is going do become unwieldy at the scale of millions of LOC - a SQLite db might be a better choice.

My three year old 15 Pro is already down to 85% capacity.

I definitely didn't want one 30 minutes ago, but now that I've seen it the idea of a phone that can be smaller in the pocket but unfold to give a large screen is pretty darn appealing. Appealing enough to pay 50% more for it? Eh... probably not.

> eSIM-only models deliver up to 36 hours of video playback on iPhone 18 Pro and up to 45 hours of video playback on iPhone 18 Pro Max.

That might convince me to upgrade. Not because I care about watching video for two days straight, but hopefully it means the battery is large enough that it'll last 5+ years before battery degradation becomes annoying.

And both of them rely on government welfare to keep their employees from starving so that they can pass more wealth on to shareholders. A study earlier this year (below) estimated that more than a million Walmart and Amazon employees nationwide depend on Medicaid.

https://www.ips-dc.org/report-americas-20-largest-low-wage-e...

I'm not sure how old you are, but shortly after Obama took office in 2009 there was quite a bit of reporting claiming that the Bush administration had been so disastrous and young people had been so thoroughly poisoned by Republican policies, that it would likely be a generation before the Republicans had the chance to win the presidency again. Remember, Bush left office in the depths of the Great Recession. Instead, it took less than 8 years for us to get MAGA.

The real value of a junior engineer is that they will become a senior. That's really always been the case IME. Juniors in the first year or two of their career are routinely a net negative, maybe break even. But the thing is, unless you're making the bet that LLMs are going to _entirely_ replace human SWE, we as an industry have to maintain a pipeline to take in juniors and level them up. If we don't, we're going to be in real trouble in about 10 years.

That's kind of risky considering the radical differences between types in SQLite and Postgres. There are plenty of situations where EF will need to be configured differently in order to map your .Net types correctly to each DB. Or subtle differences in behavior due to storage differences (particularly SQLite's predilection for storing things as strings). It's so trivially easy to run Postgres in Docker that I don't really see the advantage of using SQLite for local dev.

That's not at all unusual in the US. My dad saw a neurologist for chronic severe headaches. It was very common for us to go to an appointment, have the doctor request a follow up appointment in 3 months, then be told during checkout that the earliest available time was in 5-6 months.

I don't get it either. Earlier this week I had Claude build a POC for a tool we're evaluating. I spent a few hours building out a requirements doc in a chat with Fable. Then Claude Code built it with Fable as the manager and the code being written by about 80% Opus/20% Sonnet subagents (this was using superpowers). It ran continuously for 15 hours on the implementation and didn't hit any limits.

Oh yes, the 1990s was the height of massive fast food drinks. Wendy's had 32 oz "biggie" drinks and a 40+ oz "great biggie". McDonalds had "super size" tho I don't recall its volume. Most of those got renamed by the mid 00's and eventually downsized.

Flock data can be accessed by cops nationwide who have different (or no) restrictions on data retention or privacy compared to local police.

> It's some permutation of all the bits of available memory. It's in there somewhere.

True, but the number of possible permutations is so large it's incomprehensible. Let's limit ourselves to say a 50 MiB program, meaning the size of the compiled binary. If I'm remembering the math correctly, that amount of memory can hold 2^(2^20 * 50 * 8) different states. That's 2^419430400 or roughly 10^126000000. For comparison, the heat death of the universe is predicted to occur in about 10^100 years, or 10^107 seconds. You'd need to be able to look at more than a million programs per second to examine them all before the end of the universe.

The average car in the US is about 13 years old. Just like with phones, some people upgrade every few years even though it's a poor financial decision. Many frugal or less wealthy people only buy used cars and drive them until they die.

Serious electric aircraft are really gated behind major advancements in battery technology or some alternative power storage tech. The energy density of batteries is an order of magnitude less than fossil fuels, and you have to carry the heavy batteries with you the entire trip (vs burning off fuel as you fly).

It was built in 1898, for starters.

Are you trying to split it with an axe? You need a sledgehammer and a few splitting wedges. The sledge lets you apply a lot more force than an axe and striking the wedge focuses that force onto a small area. The first wedge will open a crack, then you use additional wedges to expand that crack until she splits.

Source: grew up in a wood burning family, helped split many stubborn hardwood trees (all by hand).

> I enforce CC on my projects because I don’t have the energy to police horrendous commit messages.

And does it actually accomplish that goal? I've been on several projects where someone pushed CC on the team with this reasoning. Every time my experience has been that you get the same crappy messages with a tag that may or may not be accurate.

BTW, AI absolutely knows how to bypass pre-commit hooks and will do so when they come up with some reasoning why their situation is an exception to the rule. I've watched them do it. The only way I've found to strictly enforce things on an agent (tests, linting, whatever) is to use a claude pre-command hook that will block git commit if the checks don't pass.