dirkc

6 ms·

1.
by dirkc 13d ago
Sorry, I wasn't implying your advice doesn't carry weight. Was more just thinking about the things that (used to) happen when you introduce more parties to process of creating software.
2.
by dirkc 13d ago
But why stop at rewriting in another language. Get another sub agent to invent a new language, create a database, query language and maybe another few DSLs. Then you've got an ecosystem! You can now re-position your initial solution an
3.
by dirkc 13d ago
And then another sub agent that argues for the whole system to be re-written in another language
4.
by dirkc 19d ago
Would you be able to taper by stretching out dosages, take a pill every 25 hours instead of 24?
5.
by dirkc 21d ago
Just don't put it in charge of a company making paper clips
6.
by dirkc 23d ago
Would building an LLM from scratch imply writing code by hand?
7.
by dirkc 1mo ago
> circular economy I get what you mean, but in general circular economy means something different, positive in my opinion. Here is what wikipedia has to say: > a model of resource production and consumption that involves sharing, leas
8.
by dirkc 1mo ago
Many times I wonder if it's possible to collapse abstractions? The idea of making the whole stack simpler. I'm not overly optimistic of LLMs achieving this, but it they do, I will be paying more attention.
9.
by dirkc 2mo ago
This immediately made me think of nand2tetris.org, but as a game. Is that what NANDGame is?
10.
by dirkc 2mo ago
I'm curious about ways to entice agents to perform weird and wacky things when working on a codebase.
11.
by dirkc 2mo ago
Isn't the problem with formal methods that it isn't clear whether or not most of the useful code we use are actually formally verifiable? The article mentions NP-complete, but is it actually a solvable problem in general? > For
12.
by dirkc 2mo ago
Earlier today I found this - https://arxiv.org/abs/2606.11755 I've been thinking lately about different ways to get agents to do interesting things when let loose on a code base. Think mischief, not malice. Someth
13.
Adversarial Code Obfuscation for Defending Against LLM-Based Analysis(arxiv.org)
3 points by dirkc 2mo ago|1 comments
14.
by dirkc 2mo ago
Thanks, clearly my caffeine intake is at issue
15.
by dirkc 2mo ago
I see a reference to Khan Academy, but not Go?
16.
by dirkc 2mo ago
For small applications having an observability stack like ELK or Prometheus + Grafana tends to outweigh the application itself ito cost and complexity. I haven't found something I'm 100% happy with to self host and tend to just go
17.
by dirkc 2mo ago
I get a cloudflare 502 error for standard.site
18.
by dirkc 2mo ago
Isn't that just saying CapEx is a bigger part of their costs as if that is a positive thing?
19.
by dirkc 2mo ago
I did something similar, for similar reasons. In the end I just did whatever to get it published. If you show up at my door, I'll pour you your choice of beer/coffee. But I agree, it feels very invasive! On the web side of things
20.
by dirkc 2mo ago
> As for me, I’ve found that the community and activity proxies are still good. Definitely still something to look into. A project I'm checking in on from time to time is https://github.com/emdash-cms/emdash&#x2
21.
by dirkc 2mo ago
Keep the timer, but after 30 seconds, deduct a point, after 60 seconds show an option to reveal the word / give up.
22.
by dirkc 2mo ago
I got curious and had a look at some of the code (>1m .rs). I was surprised to see code for a S3 client in there? I clearly don't get the value proposition of bun? And even if I accept that you want to bundle your run time, package
23.
by dirkc 2mo ago
I like the idea, but I didn't like losing after a few words. Now it might just be me not being good at losing, but who is? Maybe the game can always progress to the next word with your total score being reduced. So if you get all withi
24.
by dirkc 2mo ago
> as we aggressively adopt AI for engineering Why do we need to aggressively adopt things rather than thoughtfully adopt things? It sounds like they are probably punching AI and engineers in the process
25.
by dirkc 2mo ago
The challenge is that more and more people are producing project like this - 1,000s of commits and > 200k lines of code - and saying it was carefully created using agent based workflows and not vibe coded.
26.
by dirkc 2mo ago
Thanks for all the info you've provided! Maybe I'm just being a little grumpy. If I really need to look into a repository, I clone it and use vanilla git command line tools to have a look. It's just annoying that the modern w
27.
by dirkc 2mo ago
Cloudflare seems to be an example of a company massively upping their output in the last year or so. And a lot of it seems AI driven. I'm definitely keeping an eye on them to see if it works out for them. And if I will need to start ro
28.
by dirkc 2mo ago
How would one go about reviewing a piece of code like this? One of the things I'd typically do is peek at the commit history. Seeing what people worked on and how they did it tends to say a lot about a project. But with LLMs generating
29.
by dirkc 2mo ago
Wow, I'm surprised that doing an online course through coursera is still that expensive - £17,000 or $22,000 USD over 3 years! Interestingly it does seem comparable with what it would cost through The Open University - https:/&#x
30.
by dirkc 3mo ago
For me the worst scenario is when a kitchen sink of non-existent functionality the customer never asked for was sold. And in all likelihood it will never be used. But some project manager is hellbent on getting it through the pipeline and c

More ›