n4r9

3 ms·

1.
by n4r9 5d ago
Imo it depends on the type of change required. If it involves churning out 1000 line of new boilerplate, an LLM will win every time. If it means tracing a subtle edge-case bug and carefully planning a deep one-line change in a 100k codebase
2.
by n4r9 5d ago
To be fair, I personally only said that I'd like a label that indicates whether an app has "had thought put into it". That could involve e.g. another human reviewing the code and architecture and giving it a stamp.
3.
by n4r9 5d ago
It's possible to code thoughtfully with LLMs, but the process strongly encourages superficial signing-off on its activities.
4.
by n4r9 5d ago
I'd love a label like that which tells me when an app has had thought put into it
5.
by n4r9 5d ago
I asked Claude and it showed me various examples of other words that may well be preceded with "an", such as heroic, habitual, and hereditary. It's to do with how strongly the "h" is pronounced.
6.
by n4r9 6d ago
Trivially false. Let P be the set of maths problems and I be the interesting subset of P. If I is finite, then there exists an element x belonging to P\I whose description is minimal among P\I. Then x is interesting. QED.
7.
by n4r9 6d ago
It's bad faith to assume someone didn't read your message.
8.
by n4r9 6d ago
Yes, this is an exception to the rule in British English, although possibly old-fashioned and becoming obsolete. Some British people put "an" before any word starting with "h" and longer than three syllables. EDIT I&#x2
9.
by n4r9 7d ago
I wonder if you're reading too far into one example. Hume's original example was of the sun rising every day - nothing to do with sentience or arguably even complexity. The chicken/farmer example is one that Russell came up w
10.
by n4r9 7d ago
I'm no expert, but my understanding is that modern neuroscience has moved away from the idea that specific areas of the brain control specific behaviours.
11.
by n4r9 7d ago
The idea of the brain having a primitive layer and more refined layers is much more than ten years old [0]. The new finding here is that distinct layers have different progenitor cells. [0] https://en.wikipedia.org/wiki&#x2F
12.
by n4r9 8d ago
I'm pretty sure that I don't worship some sort of God, unless perhaps you substantially dilute the meaning of the word "worship".
13.
by n4r9 8d ago
Judaism evolved out of earlier Canaanite polytheistic beliefs with a pantheon of deities, some of whom had malevolent or destructive aspects. As the Israelites transitioned to monotheism, Yahweh - previously one of many gods - took dominanc
14.
by n4r9 8d ago
That's Russell rather than Hume, and I don't see how it links to complexity.
15.
by n4r9 8d ago
Could you cite your reference to Hume here please? I'm not aware of him linking complexity to regularity.
16.
by n4r9 9d ago
Is there a difference between worshipping something and simply thinking positively about it, in your mind?
17.
by n4r9 9d ago
Yes, vaguely related. Dirac is not wondering why measurement itself works, but why mathematical reasoning applies so effectively to the measurement results. I think it's valid to be surprised that all observations of a physical phenome
18.
by n4r9 9d ago
This is probably a reference to Hume's problem of induction, i.e. that observing a regularity in the past cannot justify assuming that it will hold in the future. It's vaguely related to the article, but talking about "time w
19.
by n4r9 10d ago
I loved seeing "Le Vent du Nord" at Cambridge Folk Festival 2023, they'd probably make a decent Eurovision entry.
20.
by n4r9 11d ago
Fair, but I think potential international sabotage of transport systems in a wartime context is interesting. Even if speculative at this point.
21.
by n4r9 11d ago
International politics is intellectually stimulating.
22.
by n4r9 11d ago
In which case what do they gain here?
23.
by n4r9 11d ago
Don't you think those groups would have claimed responsibility by now?
24.
by n4r9 13d ago
In my experience it depends on the nature of that client. A drawback to having a dominant client is that they have a lot more leverage in roadmap, prioritisation, and product fit. That can seriously impact the internal coherence and maintai
25.
by n4r9 15d ago
Maybe I misunderstood, but it seemed to me that the smoking example was evidence for your claim: > colloquially, people say something is bad when they mean worse than the alternative [as opposed to absolutely bad] I don't think it
26.
by n4r9 15d ago
Fine it has some cognitive benefits, but in terms of physical health it's absurdly net negative. Long term smokers have a huge likelihood of dying from a smoking-related illness. That's what people mean when they say it's b
27.
by n4r9 15d ago
LLMs require massive datasets and massively parallel algorithms for processing them.
28.
by n4r9 15d ago
I'm not convinced by the smoking vs caffeine analogy. Smoking has no physical upside and in the long-term is extremely detrimental to health. The formula vs breast milk feels more comparable to something like still vs sparkling water.
29.
by n4r9 16d ago
Achievement can be good. But an upbringing which heavily idolises achievement can well be problematic, and probably isn't worth it. And your arguments to the contrary feel epistemically circular. You're using your belief that achi
30.
by n4r9 17d ago
I'm not saying that any value judgement is a bad thing. (Although such a statement would be amusingly ironic). I'm saying that you're viewing your upbringing strictly through the lens it handed you. As far as I can make out,

More ›