ekidd

14 ms·

121.
by ekidd 8mo ago
Yeah, if a cartel actually used anti-aircraft weapons on a US passenger plane in US airspace? It wouldn't even matter if MAGA or the Democrats were in charge. The US would collectively lose its shit and spend the next 10 years and seve
122.
by ekidd 8mo ago
The wrinkle is that the AI doesn't have a truly global view, and so it slowly degrades even good structure, especially if run without human feedback and review. But you're right that good structure really helps.
123.
by ekidd 8mo ago
I have a wide range of Claude Code based setups, including one with an integrated issue tracker and parallel swarms. And for anything really serious? Opus 4.5 struggles to maintain a large-scale, clean architecture. And the resulting softwa
124.
by ekidd 8mo ago
A significant number of AI companies and investors are hoping to build a machine god. This is batshit insane, but I suppose it might be possible. Which wouldn't make it any more sane. But when they say, "Win the AI race," t
125.
by ekidd 8mo ago
In January 2026, prototype code is cheap. Shitty production code is cheap. If that's all you need—which is sometimes the case—then go for it. But actually good code, with a consistent global model for what is going on, still won&#x
126.
by ekidd 8mo ago
I want to compliment Anthropic for doing this research and publishing it. One of my advantages(?) when it comes to using AI is that I've been the "debugger of last resort" for other people's code for over 20 years now. I
127.
by ekidd 9mo ago
As I keep pointing out, if the model ever stops needing you to complete ambitious goals, then what does the model actually need you for? People somehow imagine an agent that can crush the competition with minimal human oversight. And then t
128.
by ekidd 9mo ago
Having actually run some of the software produced by nearly "dark software factories," a lot of that software is completely shit. Yegge's Beads is a genuinely good design, for example, but it's flakier and more broke
129.
by ekidd 9mo ago
All on my own, I hand-craft pretty good code, and I do it pretty fast. But one person is finite, and the amount of software to write is large. If you add a second, skilled programmer, just having two people communicating imperfectly drops q
130.
by ekidd 9mo ago
Yes, as an American, I could point out that the side of US politics represented by Biden, Obama and Clinton is very real. It's internationalist, cooperative, and reliably so. Clinton was, in some ways, more willing to intervene in Ea
131.
by ekidd 9mo ago
If there's a human in then loop, actually reading the plans and generated code, then it's possible to have 90% of me code generated by an LLM and maintain reasonable quality.
132.
by ekidd 9mo ago
Yes. Also, it's a fairly common trope that if you want to pilot a mech suit, you need to be someone like Tony Stark. He's a tinkerer and an expert. What he does is not a commodity. And when he loses his suit and access to his mone
133.
by ekidd 9mo ago
There are some interesting possibilities for LLMs in math, especially in terms of generating machine-checked proofs using languages like Lean. But this is a supplement to the actual result, where the LLM would actually be adding a more rigo
134.
by ekidd 9mo ago
Yeah, I agree, the time zones are killer , and this can't be ignored. I work at a company spread over most of the world, with SMEs coming and going as the globe spins. Back-and-forth iteration and consultation is a genuinely hard prob
135.
by ekidd 9mo ago
I think the actual problem here is that Opus 4.5 is actually pretty smart, and it is perfectly capable of explaining how PR disasters work and why that might be bad for Anthropic and Claude. So Anthropic is describing a true fact about th
136.
by ekidd 9mo ago
Yeah, I have zero problem getting Opus 4.5 to write high-quality Rust code. And I'm picky.
137.
by ekidd 9mo ago
I think vibe coding isn't quite good enough for real products because I usually have 4 AI agents going non-stop. And I do read the code (I read so, so much code), and I give the AI plenty of feedback. If you just want to build a li
138.
by ekidd 9mo ago
I have a version of this without the GUI, but with shared mounts and user ID mapping. It uses systemd-nspawn, and it's great. In retrospect, agent permission models are unbelievably silly. Just give the poor agents their own user accou
139.
by ekidd 9mo ago
If you give Claude examples of good and bad property tests, and explain why, it gets much better than it was out of the box.
140.
by ekidd 9mo ago
It is really easy to say something incredibly wild like "Imagine an AI that can replace every employee of a Fortune 500 company." But actually imagining what that would actually mean requires a bigger leap: The AI needs to be
141.
by ekidd 9mo ago
I have been reading through this thread, and my first reaction to many of the comments was "Skill issue." Yes, it can build things that have never existed before. Yes, it can review its own code. Yes, it can do X, Y and Z. Does it
142.
by ekidd 9mo ago
> You will need the CEO to watch over the AI and ensure that the interests of the company are being pursued and not the interests of the owners of the AI. In this scenario, why does the AI care what any of these humans think? The CEO
143.
by ekidd 9mo ago
Yes. I have been building software and acting as tech lead for close to 30 years. I am not even quite sure I know how to manage a team of more than two programmers right now. Opus 4.5, in the hands of someone who knows what they are doing,
144.
by ekidd 9mo ago
> But if claude can reliably reorganize code, fix patterns, and write working migrations for state when prompted to do so, it seems like the entire way to reason about tech debt has changed. Yup, I recently spent 4 days using Claude to
145.
by ekidd 9mo ago
One easy way to test different models is purchase $20 worth of tokens from one of the Open Router-like sites. This will let you asks tons of questions and try out lots of models. Realistically, the biggest models you can run at a reasonable
146.
by ekidd 10mo ago
I have been overusing em dashes and bulleted lists since the actual 80s, I'm sad to say. I spent much of the 90s manually typing "smart" quotes. I have actually been deliberately modifying my long-time writing style and use o
147.
by ekidd 10mo ago
In my professional life, somewhere over 99% of time, the code suffering the error has either been: 1. Production code running somewhere on a cluster. 2. Released code running somewhere on a end-user's machine. 3. Released production co
148.
by ekidd 10mo ago
Read Brooks' argument in detail, if you haven't. He has spent decades getting robots to play nicely in human environments, and he gets invited to an enormous number of modern robotics demonstrations. His hardware argument is prima
149.
by ekidd 10mo ago
> Programmers suddenly need backup plans. Yup, Claude Opus 4.5 + Claude Code feels like its teetering right on the edge of Jevon's Paradox. It can't work alone, and it needs human design and code review, if only to ensure it
150.
by ekidd 10mo ago
> I couldn't for the life of me tell you what dd stands for. Traditionally, according to folklore? "Delete disk" or "destroy data". (Because it was commonly used to write raw disk blocks.)

More ›