QuadmasterXLII

5 ms·

1.
by QuadmasterXLII 6d ago
Scandal as fraternity discovers 2/3rds of pledges were cheating by paying TaskRabbit temp workers to drink and dance in their stead, while the pledges stayed home and doomscrolled.
2.
by QuadmasterXLII 7d ago
its clocktower syndrome. they are fucked in the head and can beg us to stop them but cant stop themselves
3.
Ask HN: Does Gusto drop dependents randomly for anyone else?
1 points by QuadmasterXLII 7d ago|0 comments
4.
by QuadmasterXLII 8d ago
if meatspace stays hard, we are fine.
5.
by QuadmasterXLII 8d ago
Right now, if you want to pay money to a stranger on the internet, and have them draw you a high effort picture using a pencil, this is hard. Recently this was easy. Instead, what is extremely likely is that you will pay more than the cost
6.
by QuadmasterXLII 8d ago
It is actually pretty nice that now we can spin up an isolated copy of our whole system with one command line call- this was mandatory for letting agents run wild trying things, but mostly now it makes manual development and onboarding easi
7.
by QuadmasterXLII 8d ago
Every four months or so we try getting hyped on agents for ~ a week. Up to this point the answer has always been a pretty solid "nope this sucks back to manual code." The agentically coded parts of the codebase that came from thes
8.
by QuadmasterXLII 8d ago
This is a good FYI. I guess my microsoft trauma is showing.
9.
by QuadmasterXLII 9d ago
If one of the researchers forgot to check "don't train on this" a single time, the pretrain is going to know what they were working on. The question of whether they remembered to check "don't train on this" eve
10.
by QuadmasterXLII 18d ago
How much rote memorization did it take to harm you? the addition of 1 digit numbers is kind of mandatory to function in everyday life, the multiplication table is at most (8*9)/2 36 facts.
11.
by QuadmasterXLII 18d ago
Whatever the reason, it seems specific to the intersection of space and telescope. Space probes, landers, and crewed vehicles all get made in multiples, as well as ground based telescopes like Keck or the gravitational wave observatories.
12.
by QuadmasterXLII 19d ago
but that can just be executable heap?
13.
by QuadmasterXLII 24d ago
LLMs are a magic machine to turn commons into tragedies.
14.
by QuadmasterXLII 27d ago
"I am against these tools generally. I use them myself" would be a more interesting take
15.
by QuadmasterXLII 27d ago
Ok! I had parsed your comment wrong. I see now that you claimed that python and go projects always had at least one dependency, and that go and rust projects had similar dependency counts, but you were not claiming that python and rust had
16.
by QuadmasterXLII 28d ago
If a python example doesn't count, why were your two example languages python and go? But yeah, small number of large dependencies is the way to go.
17.
by QuadmasterXLII 28d ago
Pytorch has 9 transitive dependencies for cpu-only execution, or 29 to bring in cuda. candle, the most popular rust ml library I found in a cursory search, has 119 for cpu only, and 150 to bring in CUDA. I guess it's taste whether that
18.
by QuadmasterXLII 1mo ago
I suspect that because each RLVR episode injects ~1 bit into the models capabilities, and training on a reasoning trace injects ~megabyte into a models capabilities, distillation is powerful enough right now that they’re all basically the s
19.
by QuadmasterXLII 1mo ago
this idea that government regulation is the problem and the companies need economic incentives to self regulate is a religion around here, and after incredible amounts of evidence that is untrue, like all religions, it’s practitioners have
20.
by QuadmasterXLII 1mo ago
“everyone has reader mode” is I guess freeing?
21.
by QuadmasterXLII 1mo ago
Given an open weights model trained to sometimes bite kids, we can’t train it to not bite kids, even though billions of dollars of research have been thrown at this open problem. Given an open weights model trained to never bite kids, you c
22.
by QuadmasterXLII 1mo ago
I do appreciate the literal college lectures on there. Hard to get the algorithm to surface them but they’re golden for learning. Of course I would be unshocked if the authenticity signals I’m tracking (AB.203 Lecture 3 video title, general
23.
by QuadmasterXLII 1mo ago
Use a fucking typewriter?
24.
by QuadmasterXLII 1mo ago
a gun that goes off when dropped is a very bad gun
25.
by QuadmasterXLII 1mo ago
The memphis datacenter? Do you have a context limit of under 30 tokens or something?
26.
by QuadmasterXLII 2mo ago
this will also be true at deployment time
27.
by QuadmasterXLII 2mo ago
I wonder if it would work to make it illegal to sell ad impressions of underage people. Do whatever you want as long as you can prove its being done to adults, or prove you have no economic incentive to algorithmically engagement bait
28.
by QuadmasterXLII 2mo ago
I’m saying that providing a better standard of living to adults who cannot work is something we are spending ~3% of our GDP and 5% of all fulltime workers on. That providing these things is something we are doing, that is good, and that is
29.
by QuadmasterXLII 2mo ago
We could provide food for the poor and disabled who lack or are not using family support with a tiny fraction of our industrial output. We wouldn’t even notice this expenditure, let alone whine about it. But what we are providing is actuall
30.
by QuadmasterXLII 2mo ago
Why are comments turned off on the blog posts itself? I was planning to make 200 thoughtful comments with the cheapest AI I could find to prove a point, but the lack of an option to comment shows you already know the point: something has to

More ›