> In particular, Wren gives up dynamic object shapes, which enables copy-down inheritance and substantially simplifies (and hence accelerates) method lookup.
A general rule of thumb is that if you can assign an expression a static type, then you can compile it fairly efficiently. Complex dynamic languages obviously actively fight this in numerous ways, and so end up being difficult to optimize. Seems obvious in retrospect.
Seriously, when you're conversing with a person would you prefer they start rambling on their own interpretation or would you prefer they ask you to clarify? The latter seems pretty natural and obvious.
Edit: That said, it's entirely possible that large and sophisticated LLMs can invent some pretty bizarre but technically possible interpretations, so maybe this is to curb that tendency.
—So what would theoretically happen if we flipped that big red switch?
—Claude Code: FLIPS THE SWITCH, does not answer the question.
Claude does that in React, constantly starting a wrong refactor. I’ve been using Claude for 4 weeks only, but for the last 10 days I’m getting anger issues at the new nerfing.
Yeah this happens to me all the time! I have a separate session for discussing and only apply edits in worktrees / subagents to clearly separate discuss from work and it still does it
I sometimes prompt with leading questions where I actually want Claude to understand what I’m implying and go ahead and do it. That’s just part of my communication style. I suppose I’m the part of the distribution that ruins things for you.
To me too, if something is ambigious or unclear when I'm getting something to do from someone, I need to ask them to clarify, anything else be borderline insane in my world.
But I know so many people whose approach is basically "Well, you didn't clearly state/say X so clearly that was up to me to interpret however I wanted, usually the easiest/shortest way for me", which is exactly how LLMs seem to take prompts with ambigiouity too, unless you strongly prompt them to not "reasonable attempt now without asking questions".
I have a fun little agent in my tmux agent orchestration system - Socratic agent that has no access to codebase, can't read any files, can only send/receive messages to/from the controlling agent and can only ask questions.
When I task my primary agent with anything, it has to launch the Socratic agent, give it an overview of what are we working on, what our goals are and what it plans to do.
This works better than any thinking tokens for me so far. It usually gets the model to write almost perfectly balanced plan that is neither over, nor under engineered.
IME "don't ask questions and just do a bunch of crap based on your first guess that we then have to correct later after you wasted a week" is one of the most common junior-engineer failure modes and a great way for someone to dead-end their progression.
> However race is a cultural and social construct. [...] The superficial variations in average height, skin color, etc. do not vary enough to constitute species differentiation
Species is also a social construct. Calling race a social construct isn't the persuasive argument people seem to assume.
> the vast majority of genetic variation exists within populations (in other words, within "races") and not between them.
This is is a fallacious argument, because there is no such thing as the "average Norwegian" and the "average Pygmy", and so you cannot even construct a meaningful sentence like "the average Norwegian and the average Pygmy are more alike than any individual Norwegian is to any other Norwegian". People need to stop using this silly argument.
It's the established scientific consensus. Obviously it isn't convincing to racists, but no argument would be given that racists don't approach the subject in good faith to begin with.
I think you're being too pedantic, though, because the statement "the average Norwegian and the average Pygmy are more alike than any individual Norwegian is to any other Norwegian" is perfectly sensible within the context of this thread and relative to the supposition that genetic variability between human populations is a valid basis to justify a biological definition of race and further classifying human races as subspecies. That species is also a social construct is true, and you seem to think that it disproves the premise, but it really doesn't because species is a social construct in the sense that all scientific classification is a social construct. But when people say that race is a social construct, they mean it was created to justify white supremacy, slavery and colonialism, and that (unlike species) it serves no useful scientific purpose. You're intentionally omitting necessary context to create a false equivalence between race and species.
Here are some actual scientifically credentialed papers and statements supporting the thesis that race has no biological basis. I doubt anyone will bother reading them but here they are just for the record. Further reading is easy to find.
So was the fact that ulcers weren't caused by bacteria. "Established scientific consensus" is another argument people need to stop using.
> You're intentionally omitting necessary context to create a false equivalence between race and species
No, I'm not equating race and species, I'm refuting the argument that race being a social construct makes it meaningless or scientifically useless by pointing out that species is also a social construct while being meaningful and scientifically useful. Therefore the argument that it's a social construct is a red herring.
> But when people say that race is a social construct, they mean it was created to justify white supremacy, slavery and colonialism, and that (unlike species) it serves no useful scientific purpose.
So you agree that calling it a social construct is completely besides the point, and people are not saying what they mean and merely polluting discussions with pointless red herrings.
Now whether race serves a useful purpose is highly debatable. There are plenty of statistical associations with race that are used to this day, eg. race as a risk factor in sickle cell anemia. If your argument is that we usually have better classifications than race in many circumstances, then sure, but note that this still doesn't prove the intended point that race classifications are useless, which is a claim that they never have any use.
Edit:
> the statement "the average Norwegian and the average Pygmy are more alike than any individual Norwegian is to any other Norwegian" is perfectly sensible
Just want to be clear that this is still a fundamental category error. These are completely unlike measures and equating them properly yields different conclusions, eg. using pairwise genetic distance measures. See the paper, "Human Genetic Diversity: Lewontin's Fallacy" for where this misunderstanding originated.
> But when people say that race is a social construct, they mean it was created to justify white supremacy, slavery and colonialism
Yes, the modern context of what we call race today is inherently linked to notions like limpieza de sangre and casta in post-conquest Latin America - the true prototypical example of structural racism, where for several centuries and over several generations a "white" appearance was conflated with a socially elite status and a "racialized" appearance with poverty and marginalization. The Moors in Medieval, Renaissance and early modern Europe were of African origin, and sometimes even had what we would now call a Sub-Saharan appearance, but they were not considered "Black" in racial terms because that was not a notion that existed in that specific milieu.
It can even be more subtle, it's entirely possible that some rare members of A can C breed, and some members of C and B would not be able to breed. The "fertility" relation can only be decided between two individuals, not groups. Group-level fertility is a statistical average of individual fertility.
That said, I don't think that means that "species" is entirely subjective or meaningless.
No… seriously. Every model release is accused. Including Opus, GPT-5.4, whatever. And yes, including smaller models that are not the top in every benchmark.
I would almost be tempted to call it benchmaxed if that term weren’t such a joke at this point. It is a deeply unserious term these days.
Gemma 4 is worse than its benchmarks show in terms of agentic workflows. The Qwen3.x models are much better; not benchmaxed. I have tested this extensively for my own workflows. Google really needs to release Gemma 4.1 ASAP. I really hope they’re not planning to just wait another calendar year like they did for Gemma 3 -> 4 with no intermediate updates.
A general rule of thumb is that if you can assign an expression a static type, then you can compile it fairly efficiently. Complex dynamic languages obviously actively fight this in numerous ways, and so end up being difficult to optimize. Seems obvious in retrospect.
reply