It may be an artistic/engineering choice to demonstrate what minimizing bloat to an extreme degree looks like.
165 karma · joined August 15, 2025
He was against the unofficial Mac port of Notepad++ but that had many problems other than being AI-assisted.
Hatchet has technology - the titular hatchet, along with the airplane that crashed and the various items the protagonist collects at the end. But the presence of technology in the work does not determine if it is science fiction or not.
This is straying away from the article, but if Hatchet were dictated to a tribe who didn’t have metallurgy, perhaps it could be perceived as science fiction.
> For years now, the testers at DoesItPlay have been documenting this trend, testing thousands of physical game releases to see which ones live up to the promise of full “plug and play” functionality without an Internet connection. Thus far, a full 27 percent of the physical releases they’ve tested require some sort of download to fix game-breaking bugs or obtain core game content that is not stored on the physical release itself. That ratio balloons to 34 percent for tested PS5 games and 50 percent for those on the Xbox Series X.
Is this training occurring with commercial AI models or open-weight models?
It seems like if this were all being done in an above-board way with commercial models, OpenAI and Anthropic would be able to determine which texts were being used in the training process.
> our dataset of known human text is drawn from 2021 and earlier
That’s great for now, but that corpus is already missing five years of slang and won’t get any younger.
I was under the impression that the lack of specialization is an aspect of the models themselves, not merely the UI or harnesses.
For example, the recent OpenAI success in a mathematics proof was accomplished with a general purpose model [0].
> The proof came from a new general-purpose reasoning model, rather than from a system trained specifically for mathematics, scaffolded to search through proof strategies, or targeted at the unit distance problem in particular.
Once the model is trained on general purpose data, specialization is just another kind of guardrail, as vulnerable to “jailbreaking” or prompt injection as any other content-based restriction.
[0] https://openai.com/index/model-disproves-discrete-geometry-c...
> The decoy message serves as a final trick for a determined agent. When looking for a hidden message, it might first find the decoy message and think that that is the real embedded message in the video.
> What is clear is that once the collapse started, it was contagious, likely for two reasons: first that collapsing areas produced invading forces and refugee flows that destabilized their neighbors and second because as you will recall above, these states are interlinked and their rulers rely on trade to furnish the key military resource (bronze) as well as to acquire key prestige goods necessary to maintain the loyalty of the aristocracy.
> The friendly interaction between the Ello character and my kids gives them a fun motivation
> AI drives a love for reading
I cannot trust someone to assess a love for reading, if I also loathe their writing.
That’s good discipline, but with better tooling and languages such conditions could be enforced by the compiler such as with Dafny. But Dafny is overkill for those who only want to avoid underflow and overflow without verifying everything else about their program…
That’s what I mean when I lament the state of affairs for underflow. Subtracting numbers is such a basic operation, but for the tech stacks used by most teams it still requires a level of care to avoid errors or corruption. Our industry is building million line behemoth codebases but these basic operations still have to be designed around.
> Typically, in real algorithms, what you want is to not only know the delta, but also which number is larger.
I agree with you, and the article was more concerned about avoiding underflow than actually having a practical use for its various example idioms.
What I’d actually prefer is a guarantee that subtracting two (X) bit numbers results in an (X+1) bit number. A sort of compile-time-checked bignum implementation where the programmer never thinks about the integer sizes or which number is larger until the final step of a calculation. Obviously this gets wasteful when multiplying numbers several times. EDIT: and I’m referring to an abstraction of integer bit sizes in a programming language, I’m not expecting the instruction set or hardware to have arbitrary sized registers.
It’s pretty sad that after all these years, dealing with fixed size integers is still so complicated. Yes, many of the problems are specific to low level languages with undefined behaviour and numeric for loops. But the issue of subtracting two numbers and possibly having an underflow is both common and a bit absurd.
The code in the article for “safely” calculating the difference of two unsigned numbers, which is simpler than the equivalent for signed integers, is this little ritual:
> delta = max(x, y) - min(x, y);
Seriously??? Two function calls just for the difference of two numbers??
Why can’t such “safe” operations have some of the sweet syntactic sugar, and the underflow-rampant “ordinary” operations have the bitter medicine of ritual?
Honest question: why? Why would anyone important enough to hold that power actually use it to force resales to exist? Who actually has the incentives to make this happen?
That isn’t as convenient as an all-in-one tool, and might not be what the user you’re responding to is doing. But it’s doable.
But I still don’t understand what you actually expect from maintainers of free and open source software. In prior comments you have used words like “empathy” but I’m wondering in more specific terms what you mean. What should be maintained, and what should be abondoned? What should stay stable and what should be updated?
I ask because when you say stuff like
> I am just deeply saddened how any attempt to suggest that there were and should still be some implied cultural norms regarding expectations
I’d feel very differently about that comment based on the “expectations” that are “implied”.
And perhaps your answers might help me out a bit, to be honest. There are plenty of times when I use an open source library and get annoyed by the decisions of upstream developers, but I often wonder if my annoyance should be directed at upstream at all, or at myself.
> People now yell at you that their only obligation is whatever is spelled out in the license they attached to the code.
Let’s turn your implied question around: if a person wants to share code without any expectation of care and maintenance, what should they do? Is the entire concept bogus, and the developer should just keep the code to themselves forever? Or put a “DO NOT USE FOR YOUR BILLION DOLLAR COMPANY” in a README? What communication other than the license could reasonably be provided?
I think the person you’re responding to made a pretty good point that open-source is sometimes the passion of an unpaid 1-person “team” and sometimes the product of a VC-funded attempt to buy goodwill. The idea that asking for more from maintainers is entitled is clearly suitable for the former case but not the latter. Now that Bun has been bought out perhaps they are more deserving of scrutiny.
Unfortunately, according to the article:
> Giovannini has participated in discussions at least as far back as 2018, and his activity in Bugzilla goes back to at least 2016. He does not appear to have been a particularly active contributor to the project, but his involvement clearly predates the agentic AI era. Whether his account is now being operated by a human attacker, an agentic AI, or a mix of both, it has a legitimate history prior to its recent activity.
So people would have to not only verify the age of Giovanni’s accounts, but judge whether his behaviour was normal.
In the entire linked article, where is the explanation for what a savings account is? Most of the early paragraphs are just waffling about how “Types of Accounts” are important. I’m pretty sure I read the phrase “money is emotional” before even getting to any description of any type of account. The word “savings” almost never appears and none of the instances seemed to define a savings account.
Honestly, is this content written by AI? In my opinion it’s acceptable to use AI to replace the boilerplate HTML, JavaScript and CSS of your site. But using AI for the actual writing risks turning your “educational tool” into a tool for misinformation.
EDIT: according to Pangram, 100% of the first two paragraphs are AI generated. Which is not a surprise at all, I don’t see how a human tasked with describing types of bank accounts would struggle this hard.
In comparison CodeBerg [1] and SourceHut [2] both offer Git hosting but don't merely describe themselves as "GitHub but X".
When I first saw the 26k changes statistic I was shocked. It made me think a large chunk of code running on people’s machines was AI-generated.
But the knowledge that a lot of the changes might be testsuite changes made me change my perspective. If for instance 25k of the changes were test changes and only 1k of the changes actually affected the .so and other artifacts used downstream, that would be a lot less dramatic.
I haven’t reviewed the code, only the messages, so I don’t know if these changes were removing or adding test cases. And there are a minority of Claude-assisted changes which are not listed as tests.
Honestly, why? If AI actually becomes capable of replacing large sections of the workforce, why wouldn’t a business composed entirely of AI “employees” outcompete their rivals?