Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bglazer
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
36 ms
·
421.
▲
by
bglazer
5y ago
Do you not feel responsibility for the mental health or well being of your students? I’m trying to interpret your comment charitably, so I hope I’m not misinterpreting.
422.
▲
by
bglazer
5y ago
> "they are not exactly having a massive impact, even as they do a lot of cool stuff" AlphaFold has changed the course of structural biology research and some descendant of that technology will absolutely be used to design drug
423.
▲
by
bglazer
5y ago
I’m particularly excited about nanopore protein sequencing. I don’t think there are any startups yet but the technology is commercially available for DNA sequencing.
424.
▲
by
bglazer
5y ago
What do you think about these direct linear actuators [0]? They seem super simple, and like a closer match to the way biological muscles work. Is there some reason why rotary (?) actuators are the default choice for robotics? Edit, also wha
425.
▲
by
bglazer
5y ago
Colorado trail? It would be quite the drive. Also probably a bit long for 20 days.
426.
▲
by
bglazer
5y ago
Always happy to see a comment from YeGoblynQueenne on a program synthesis post. I’m optimistic about the prospects of some combination of logic solvers and deep learning. As you mention, natural language specs are only approachable with a l
427.
▲
by
bglazer
5y ago
There are two possibilities for lab leak: 1. The Wuhan institute for virology collected a naturally occurring virus then did gain of function research which subsequently resulted in infected lab workers. There's not a known virus with
428.
▲
by
bglazer
5y ago
I'm quite astounded and confused at the sudden shift in discourse towards assuming that the lab leak theory must be true. It's not so much the theory itself that's surprising. Lab leak remains plausible but less probable then
429.
▲
by
bglazer
5y ago
You're incorrect that neuroscientists and computational researchers don't collaborate. In fact, "computational neuroscience" departments have existed at major universities for a number of years. DeepMind was a spin-off o
430.
▲
by
bglazer
6y ago
Soda constructor It looks like the original is not online anymore, but here's an open source version https://peterfidelman.github.io/constructor/
431.
▲
by
bglazer
6y ago
Twitter stopped showing "liked" tweets from people you don't follow on the timeline. It's been an enormously positive change, in my opinion. It also happens to be a change that their users have been requesting for years.
432.
▲
by
bglazer
6y ago
Yes, hopelessly extending this analogy, see the change in the way artists represented galloping horses before and after photography. https://www.amusingplanet.com/2019/06/the-galloping-horse-pr...
433.
▲
by
bglazer
6y ago
Agreed! I laughed out loud.
434.
▲
by
bglazer
6y ago
I'm curious about your thoughts (if you have any) on the boolean modeling formalism. Basically, you represent bio-molecules as having two states: active and inactive. Their states change according to boolean logic update rules that are
435.
▲
by
bglazer
6y ago
A confidence interval is a sort of meaningless notion in this case. You can just do the derivative of the algorithm's answer, and see if you get back an expression that is equivalent to the input. If you did, then the answer is correct
436.
▲
by
bglazer
6y ago
You might be interested in the line of work around differentiable relaxations of formal solvers that was kicked off by Zico Kolter et al. They add a Max=SAT solver as a differentiable layer. This allows the neat trick of solving handwritten
437.
▲
by
bglazer
6y ago
I've read a fair number of papers on neural program synthesis lately. To me, these seemed to be obviously cherry picked examples, so you can't really evaluate the whole system based on them. However, this is fairly impressive for
438.
▲
by
bglazer
6y ago
This is fantastic. I've been looking for a strategy for graph subsampling for my research on protein-protein interaction networks. It's a surprisingly tricky problem, that I now don't have to worry about as much, thanks to th
439.
▲
by
bglazer
7y ago
https://www.markbittman.com/recipes-1/no-knead-bread
440.
▲
by
bglazer
7y ago
openFrameworks
441.
▲
by
bglazer
7y ago
One argument against this position is that economic environmentalism cuts both ways. Sure Sure, marine preserves produce an economically beneficial surplus of salmon. Also, the oil reserves in the Arctic national wildlife refuge are worth m
442.
▲
by
bglazer
8y ago
I had the thought that perhaps there's some sort of voting in flocks of migrating birds. I had this idea while listening to a flock of starlings make a tremendous amount of noise. It seems like a waste of precious calories to chirp and
443.
▲
by
bglazer
8y ago
https://en.m.wikipedia.org/wiki/Diamond_anvil_cell
444.
▲
by
bglazer
8y ago
Reformatted list for mobile: 1) Divide up that goal into pieces. 2) Take those initial pieces and divide that up. 3) Take one of the bits from 2, and create a lame, incomplete version of it 4) Keep polishing that bit until it's great
445.
▲
by
bglazer
8y ago
That's the point. YC should not limit themselves to only SV based companies
446.
▲
by
bglazer
8y ago
"Beckman and Shockley ... selected the location in Mountain View, near Palo Alto, California. The place did not prove very successful: Shockley's colleagues at Bell Labs and RCA refused to move to a rural location with no long-dis
447.
▲
by
bglazer
8y ago
To be totally honest, I always find these stats about some x% > 50% failure to program at all totally unbelievable. What do these people do all day, if not design and build software? Yet they're apparently completely unable to write
448.
▲
by
bglazer
8y ago
This is generally called the k-Nearest Neighbors problem. You should check out the various data structures for doing this, like the ball tree and kd tree.
449.
▲
by
bglazer
8y ago
I don't know if this is "easy" but Microsoft Research has this free and interactive tool: https://msbnx.azurewebsites.net/
450.
▲
by
bglazer
8y ago
I agree totally that a "non-pure" deep learning approach is going to be necessary in pushing the field forward. World shattering results from new deep architectures like AlexNet are becoming more and more rare and improvements are
More ›