bunderbunder

6 ms·

61.
by bunderbunder 19d ago
Yeah. And TLA+ can confirm that the design is sound, but it can’t confirm that the implementation conforms to the design. QuickCheck style tests can’t solve that problem, but perhaps they can mitigate it.
62.
by bunderbunder 19d ago
I've seen a lot of technical points about reduce, all of which are true. But I think the real reason might be even simpler: you can't tell what it does just from the name. What `map` does is consistent with well-known programming
63.
by bunderbunder 19d ago
You also have the problem of potentially having to re-verify everything by hand for every little change. Maybe fine for the kinds of projects Dijkstra was working on, but less practical in a business setting. Tools like QuickCheck and Hypot
64.
by bunderbunder 20d ago
Yes. One interesting example is that compulsive gambling is a side effect of a certain class of drugs for treating Parkinson’s disease. https://parkinsonsjournal.com/en/parkinsons-disease/gambling...
65.
by bunderbunder 20d ago
Yes, but also all too often “useful” is interpreted to mean “moves the metric”. If that metric is merely a proxy for some more tangible outcome then that may not be good enough. The one that tech tends to stumble on most often is velocity-t
66.
by bunderbunder 20d ago
Anecdotally, I’ve found that codebases that enforce code coverage metrics often have worse behavior coverage than ones that don’t. It’s a classic example of Goodhart’s Law in action. Code coverage metrics only measure what percentage of c
67.
by bunderbunder 20d ago
The thing is, AI has no idea when an abstraction is good or not. The reductio ad absurdum here is that, if abstraction can just be assumed to be bad for quality and maintainability, then perhaps we should go back to hand writing machine cod
68.
by bunderbunder 20d ago
A while back Hillel Wayne did a talk (whose name I forget) on what empirical evidence on software quality actually says. As I recall, he concluded that there’s really no support for then-popular ideas like short functions, reducing cyclomat
69.
by bunderbunder 21d ago
For 100% of the root cause analyses for critical production defects that I’ve done over the past 18 months or so, a key finding was that AI-generated tests asserted that the defective behavior was correct.
70.
by bunderbunder 21d ago
Also just a lot of overemgineering. The last project I kicked off, I straight up told my manager that in the pre-Claude days I’d have pushed back pretty hard on automating this particular task. And even with AI I’m still pretty sure that bu
71.
by bunderbunder 21d ago
I don’t know if that’s strictly true. But one thing I’ve seen since AI coding has taken over is that the additional features that do get built are all the lower-value ideas that would previously been left on the cutting room floor. I’m not
72.
by bunderbunder 21d ago
So there’s an argument to be made that when the same person designs, implements and tests the change, any purported defense in depth is largely illusory. Once upon a time I disagreed with this argument, insofar as I believed it wasn’t nece
73.
by bunderbunder 22d ago
Economics is generally not a zero sum game. At least according to what they taught me in Econ 101, on a macroeconomic scale skilled workers tend to increase job availability. They spend a larger portion of their income on services, which ha
74.
by bunderbunder 24d ago
That's also a good point. When I'm using caveman (and especially cavekit), I don't have to spend quite so much energy on dealing with it building features I didn't ask for and don't want.
75.
by bunderbunder 25d ago
I'm not so sure that's a fair comparison. So much "bad" enterprise code evolved into that state over years or even decades of small changes. Meanwhile, last year I got to watch an LLM-authored codebase speedrun itself in
76.
by bunderbunder 25d ago
I am actually rather fond of caveman. I haven't evaluated it for token cost, in part because frankly I think that part of the pitch is a load of malarkey. Output that's shown to the user is such a small percentage of overall token
77.
by bunderbunder 25d ago
It's possible to step out of your area of expertise in a way that's confident but also humble. Arrogance is not necessary. The person who signs everyone's paychecks might be making it mandatory, but that's not quite the
78.
by bunderbunder 25d ago
I've experienced it in industry, too. I recently got out of data science in part because I got tired of working with people who, emboldened by their PhDs in some completely other field, liked to patiently but condescendingly mansplain
79.
by bunderbunder 27d ago
I think that quantitative researchers have known this for a while, too. My perennial experience as a machine learning practitioner working in industry is that the ML and statistics folks raise concerns about the models learning social biase
80.
by bunderbunder 1mo ago
I disagree with the claim that "AI agents don't get lost." What I've observed instead is that they don't experience the sensation of feeling lost. Which is quite different. This summer I spent quite a while using a
81.
by bunderbunder 1mo ago
More information on the specific remedies in the USDOJ announcement: https://www.justice.gov/opa/pr/department-justice-wins-signi... Sounds like it's not nothing, but also not much.
82.
by bunderbunder 1mo ago
I spent some time working with that approach of using LLMs to generate synthetic labeled data for use in training more specialized models. It mostly didn't work. The problem was that getting the LLM to generate training data that suffi
83.
by bunderbunder 1mo ago
I almost wonder if it's even possible now that so many of us get our information through algorithmic feeds. I listened to one Zitron interview on YouTube and my home page was immediately crammed full of similarly foamy-mouthed AI criti
84.
by bunderbunder 1mo ago
IANAL, but it seems like the wording in the public-facing version of the policy is a little muddy. As far as I'm aware, "tax exempt donation" not a thing that actually exists in US tax law. It seems to be a conflation of &quo
85.
by bunderbunder 1mo ago
I don't think it over-claimed anything? Maybe over-simplified for certain purposes. But to expect a press release to be a complete accounting of the research it reports on is to misunderstand the purpose of press releases. A press rele
86.
by bunderbunder 1mo ago
I've discovered, for my part, that wearing yellow glasses works even better than f.lux when I'm working in an environment where I can't control the ambient lighting.
87.
by bunderbunder 1mo ago
And the most underappreciated kind of science. There have been estimates that up to 80% of published medical and psychology results arrive at incorrect conclusions due to methodological flaws, analysis flaws, and type I errors. And peer rev
88.
by bunderbunder 1mo ago
From the paper's abstract (emphasis mine): "Wavelength influences multiple aspects of visual performance, yet its role in spatial resolution remains incompletely understood due to confounding factors such as luminance difference
89.
by bunderbunder 1mo ago
I don’t love the “forced upon them” framing; if that’s really how people are thinking about it then maybe they should pause and reflect for a moment yhat it isn’t their money being spent. Amd the default isn’t always having the latest and g
90.
by bunderbunder 1mo ago
My sense from talking to people who’ve lived in both countries is that there’s also a perception that there’s at least a semblance of fairness in China, which can make it relatively more palatable. Chinese acquaintances are rather horrified

More ›