programjames

17 ms·

151.
by programjames 1y ago
From [1], these are the rankings for CS: 1. Stanford 2. UC Berkeley 3. MIT 4. Caltech 5. Harvard I'm a little surprised MIT and Caltech are lower than Stanford and UC Berkeley. I know that MIT has a culture of sending its undergraduate
152.
by programjames 1y ago
If you're measuring the value in dollars, it would be very surprising if people maxxed on the INT stat rank higher than those on the USD stat. But, for example, if your goal is to secure a professorship at a top university, or do the m
153.
by programjames 1y ago
Clearly you need to reread The Bible: > To the one who is victorious and does my will to the end, I will give authority over the nations—that one 'will rule them with an iron scepter and will dash them to pieces like pottery'—j
154.
by programjames 1y ago
I believe you're being purposefully dense here. I get it, and probably do the same sometimes, but the connection is pretty obvious: someone gets in (allegedly) due to ideological fit instead of merit, and then becomes one of the most p
155.
by programjames 1y ago
Alternatively, the gap between the classes for access to educational resources has significantly declined in the past twenty years (due to stuff being posted online).
156.
by programjames 1y ago
TLDR; Train an "energy" model that checks if the output is correct (rather than just outputting something), and gradient descent to find good outputs. Using transformers.
157.
by programjames 1y ago
Negligence is enough for libel against private citizens in the USA.
158.
by programjames 1y ago
I wouldn't expect behaviors from mass shooters to carry over to serial killers.
159.
by programjames 1y ago
Note: the police do not have the suspect in custody. The comments about, "here's the assassin being arrested," are libel.
160.
by programjames 1y ago
Note: it was an assassination, not a mass shooting. There was only one shot.
161.
by programjames 1y ago
We can only ever have proxies. The issue (at least here) isn't Goodharting, it's that the proxy does not even measure what the researchers claim it measures. For example, pretty much the only study of its name, "High-Achievin
162.
by programjames 1y ago
In my experience, the "top tier of kids" is more cultural than school-specific. Even in schools like TJHST there's usually 10–30 students in the school that really care about achieving, while the other 90% don't put in m
163.
by programjames 1y ago
Gifted programs dropped from ~72% of elementary schools to ~65% by 2013, and probably have continued declining. Given it takes 10+ years to educate a child, the school culture to change, and so on, we should expect to see quite a lag betwee
164.
by programjames 1y ago
Homeschool or China.
165.
by programjames 1y ago
This was trending long before the pandemic.
166.
by programjames 1y ago
The core problem is actually very simple. Education studies do not measure what they claim to measure. When they say, "education outcomes improve when..." they usually mean the pass rate, i.e. they only measured a signal among t
167.
by programjames 1y ago
The core problem is actually very simple. Education studies do not measure what they claim to measure. When they say, "education outcomes improve when..." they usually mean the pass rate , i.e. they only measured a signal among
168.
by programjames 1y ago
The conclusion I drew is that even schools in the "top 2 states for K-12 education" are piss poor at education.
169.
by programjames 1y ago
Are your kids old enough to run amok at home instead of going to school? Would the police arrest you if you left them home alone instead of sending them to school?
170.
by programjames 1y ago
I graduated from high school less than ten years ago. I'm sure screens have become a big issue in many (or most) schools, but that was not the case at my high school. It still was mostly daycare, not education, so banning screens wil
171.
by programjames 1y ago
Particularly, the biggest incentives are test scores and passing rates, which incentivize attention only to the bottom 50% and 20% of students (respectively). This means: - You do not diversify classrooms by academic ability---the high-perf
172.
by programjames 1y ago
In the less affluent areas, I'd expect children not attending daycare to just not be getting anything at home. Presumably their parents are both working and cannot afford daycare. In the more affluent areas, I'd expect children on
173.
by programjames 1y ago
Suppose childcare is $15k/year and you work minimum wage making less than $15k/year. Then there's less wealth to go around, just more in your pocket. But actually, you probably don't take home all the wealth you create,
174.
by programjames 1y ago
I'm confused, what cultural values do you want them to learn?
175.
by programjames 1y ago
For me at least, it's a general dislike of the wider educational system. My parents taught me to read, play chess, multiply, and write in cursive before elementary school. I didn't really learn anything at preschool or kindergarte
176.
by programjames 1y ago
Lots of two-parent working families do the maths, and realize they would pay more in childcare than the income from a second job. This incentivizes one of them to stay at home. Here, the incentive is gone. This is worse for the economy and
177.
by programjames 1y ago
Not the guy you replied to, but here are some improvements that feel obvious: 1. Memory indexing. It's a pain to avoid banking conflicts, and implement cooperative loading on transposed matrices. To improve this, (1) pop up a warning w
178.
by programjames 1y ago
It took about the same amount of time to evolve human-level intelligence as human-level mobility. Pretty much no other animal walks on two legs...
179.
by programjames 1y ago
I think you are referring to > The strategic thinking that goes into longer-horizon tasks may be something LLMs aren’t as good at, which aligns with why entry-level workers are more affected than experienced workers. I think the article
180.
by programjames 1y ago
I wonder if there's a way to do something similar to Rectangular Surface Parameterization[1] with voxels. It would allow you to get pretty even-volumed voxels, and also simplify vertex identification (same three coordinates, nonlinear

More ›