feintruled

I have identical twin boys - when signing up for a bank account app and while completing all the associated hoops for ID, the second to do it got the rather dystopian error message 'an account has already been registered with this face'

And worse yet, you'll find the code peppered with comments relating to 'phase 3' and 'section 11', ephemeral stuff that had meaning in the moment but now enshrined forever. And what happens when the LLM stumbles on this and working off a whole different phase 3 or section 11?

So the critique from the other story holds true - it really is just running the Linux driver inside docker, and (perhaps) dressing it up as something more - though I note the claim on this page is walked back from the original 'writing the driver'. On the other hand, it really did get him printing from his Mac, so in this new LLM results based world we live in it did its job admirably.

Nice, I will give this a go. This is what /resume should be but because sessions are tied to directories it's very easy to lose track of where you started what. Until now I've been asking Claude 'what was the conversation we had about [X]' which works but is clunky in the extreme.

> The Aga chipset of the 1200/4000 stupidly only added 2 more bitplanes. The CD32 chip actually had byte-per-pixel (chunky) graphics modes but the omission from the 1200 was fatal.

The intention was good, but the Akiko chip was functionally almost useless. It was soon surpassed by CPU chunky to planar algorithms. I don't think it was ever even used in any serious way by any released games (though it might have been used to help with FMV).

I was impressed to see Irish was included, but it soon became clear it's all LLM generated. That's not enough to write it off on its own (ChatGPT can be surprisingly good when explaining Irish sentences) but I spotted so many mistakes in the demo questions alone I can't imagine this is going to help anyone learn anything.

I would start to enumerate the mistakes, but it's not even worth it. It's really terrible. Can't sugar coat it at all, sorry.

Interesting article, though is there really much new there? Also it discounts the alternative hypothesis of some bright spark acquiring wolf pups and doing it purposefully because that would take 'weeks'. Weeks, you say?

Surely some enterprising hunter-gatherer had sufficient time on their hands. I can't help but think strutting around with a feared predator at your beck and call would have been the ultimate status symbol, and once you saw it would have to be the must have accessory for the self-respecting hunter. Aficionados would no doubt breed their stock amongst themselves to save the hassle of having to abduct more wolf cubs, which would naturally tend to the more suited specimens (friendliness being one trait as you don't want them eating the kids). Once it was realised what an incredible force multiplier they are in hunting and their utility in defence, any time investment would pay for itself many times over.

I find this no less as unlikely as thinking humans would let wolves help themselves to their excess food. Fascinating subject all round, no matter the reason. I hope they can figure out more.

Reminds me of Werner Herzog's autobiography. In the introduction, he muses on a life being cut short by a snipers bullet, and when he sees a bird flying past his window as he is writing his book makes him imagine it is a bullet and he thinks it would be a nice device to cut his final chapter short at that exact moment, so he is giving fair warning that the book will end abruptly.

And so it does, but in a totally Herzog moment he then almost immediately intones afterwards "and that is the end of the book as I indicated in the foreword".

I dunno, man. You get to work for months to completion on a deeply fulfilling project, get well compensated, and then the plug gets pulled before it goes to production? That's living the software engineering dream!

Jokes aside, the article resonated with me (before the LLM vibes got overpowering) as I am learning Irish as my own personal challenge, which as a minority language is similarly derided by some as useless (there is essentialy no Irish speaker that does not also speak English) but which I have found tremendously intellectually invigorating and the most pumped I have been for a project in a long while. So it rang true for me before half way through a distinct "linkedin parable" nature started to come to fore. So alas I rather doubt the author is learning Sumerian at all. Cynical perhaps.

Some missing context (pun intended) is that Augment code has recently switched to a per-token instead of per-message pricing model. This hasn't gone down particularly well, but that's another story. But it may well be that users drop back to older models in the expectation it will use less tokens.

Personally, I stopped using GPT-5 as it would just be tool call after tool call without ever stopping to tell you what the hell it was doing. Sonnet 4.5 much better in this regard. Albeit it's too verbose for the new token based world ('let me just summarise that in a report')

I was wondering if it was a way to 'flaunt' his avoidance of LLMs. Could be in future genuine human conversation will be so rare that we will seize upon mistakes and typos like finding a shiny jewel in mud.

(But then I saw he used the formation - 'Honestly?' which made me think he WAS using LLMs!)

Interesting. Though it seems they are themselves building Agentic AI tooling. It's vibe coding all the way down - when's something real going to pop out the bottom?

Inspired by the recent post to describe relativity in words of 4 letters or less, I asked ChatGPT to do it for other things like Gravity. It couldn't help but throw in a couple 5 letter words (usually plurals). Same with Claude. So this could be a good one?