mwkaufma

4 ms·

1.
by mwkaufma 5d ago
"Advisory Group" Importantly, the group will not be responsible for advising us on how to pace our internal progress on mathematics.
2.
by mwkaufma 5d ago
"The attack surface wasn't big _enough_"
3.
by mwkaufma 8d ago
If it's one of those two it's not fully undefined behavior it's just implementation defined. UB can do anything.
4.
by mwkaufma 8d ago
The good uses cases for trivial infinite loops is a vanishingly small subset of infinite loops.
5.
by mwkaufma 11d ago
When your "app store" is just a chatbot role-playing being an app store.
6.
by mwkaufma 15d ago
Speculating about future technological risks is at least more fun than acknowledging immediate systemic financial risks.
7.
by mwkaufma 18d ago
If you add "fits in a register" to your list of correctness requirements, then it's no-go even if the source has less cognitive overhead.
8.
by mwkaufma 18d ago
Three days late homie
9.
by mwkaufma 22d ago
So, next-token predictors
10.
by mwkaufma 24d ago
Question isn't how can a function closure be represented anywhere, but how it can be represented in a way that can be aliased by a vanilla C function-pointer, which is a more specific and challenging question.
11.
by mwkaufma 24d ago
The lady doth protest about the lady doth protesting too much, too much.
12.
by mwkaufma 28d ago
Complexity measures the worst-case, not the amortized case. If you want to report proportional cycles for more fine-grained per-feedback, fine, report proportional-cycles, but that's not Big-O, so don't use that notation.
13.
by mwkaufma 28d ago
need to represent the "captured-variables"/closure
14.
by mwkaufma 28d ago
You've introduced an idiosyncratic definition of "worst-case" that nobody else uses to redefine proportional cycle-counting as "complexity", so yeah, I guess in your novel terminology that makes sense, but it isn&#x
15.
by mwkaufma 28d ago
Every O(n) is O(1) if n = 1. Complexity measures worst-case by-definition, not all cases.
16.
by mwkaufma 28d ago
Isn't O(n - k) or O(len(l1) + len(l2)) just O(n)? Instead of blurring the line between complexity-analysis and cycle-counting, just print both the complexity and the est proportional cycle-count as separate measures.
17.
by mwkaufma 29d ago
Awkward prose of GenAI.
18.
by mwkaufma 1mo ago
I know it's not just for the branded inverted asterisk, but is it just for the branded inverted asterisk?
19.
by mwkaufma 1mo ago
Cosmetic improvements. Term originates from observations of asian-associated car decorations, which based on context is either a bigoted pejorative or an ironic compliment.
20.
by mwkaufma 2mo ago
Tellingly, the legal tendency The Economist finds alarming involves workers asserting their rights.
21.
by mwkaufma 2mo ago
The "delusions of grandeur" phase of the hype cycle.
22.
by mwkaufma 2mo ago
The ram usage in this case is an order of magnitude lower if you install ublock in edge, as other comments have noted. Why would a system level GC affect wasteful adtech? Any marginal efficiency gains will be wiped out with more adslop. Nat
23.
by mwkaufma 2mo ago
So the best way to run a controlled trial is to maximize the external noise and confounding variables?
24.
by mwkaufma 2mo ago
It's not strictly true that all physical disks are just license keys - those are marked on the packaging.
25.
by mwkaufma 2mo ago
Won't an LLM, prompted for a CVE report, always hallucinate one even if it isn't there? Why would they "settle down"?
26.
by mwkaufma 2mo ago
This very article mentions windfall profits tax multiple times.
27.
by mwkaufma 2mo ago
Why?
28.
by mwkaufma 2mo ago
If only there was some link to a readme which would answer this question.
29.
by mwkaufma 2mo ago
Buzz-word soup.
30.
by mwkaufma 2mo ago
No, just a much more mundane rhetorical bait-and-switch. Ranking the "greatest logician" is purely-subjective, not a truth-value. Barging in like the kool-aid-man to bellyache that Harper isn't applauding Frege in an article

More ›