Almost sounds like script kiddies when they start learning programming for the first time in their lives they assume that the test of successfully learning how-to-program is that your code should compile on first try without errors.
1,022 karma · joined August 10, 2014
Almost sounds like script kiddies when they start learning programming for the first time in their lives they assume that the test of successfully learning how-to-program is that your code should compile on first try without errors.
In order for outsiders to understand tech, you need to invite engineers/researchers, not CEOs.
And you need marathon sessions, not sprints.
Go figure!
I do not believe she's coming clean on her learning journey.
If this wasn't horsesh*t to begin with, she went on to work in non-physics areas after graduation, and never did any research work in physics (no grad school either).
How convenient.
(likely explanation: either her undergrad program was super lax, passing pretty much everyone who shows up in class and exams, hence useless for a serious career in physics, or she's misrepresenting her background)
[0] https://web.archive.org/web/20170314073043/https://fledgling...
There is plenty of magnesium in foods.
As a STEM grad student (not in math), I had more than a couple such moments of crises, when I posted my questions on various stackexchange websites. I got either useless replies, or no replies.
- C: You will create the mean function
It's the difference between using a calculator and knowing how to multiply two 5 digit numbers with pen and paper.
You might say, "Oh but we can do that in R". Well, if you start doing that one function at a time, very soon you find yourself with R grinding to a halt.
Like when Google decided it's going to host comp.lang.c, can there be only one comp.lang.c on the internet, or can someone else start hosting comp.lang.c as well?
I can guarantee you that deep NN/RL researchers are thinking of these hypotheses every other week. But they don't publish them. Even I, who got more active in this kind of research very recently, came up with the exact same hypothesis. I didn't publish.
We don't publish because we have a strong sense of empiricism. There proverbial proof has to be in the pudding. We should be able to setup a DNN and run experiments. If it confirms our hypothesis, we publish. If it disproves our hypothesis, we move on to something else.
These armchair neuroscientists/psychologists come along and type out a dozen pages of brain-dump, no pun, and think they're doing science. They need to do better than that, and do some real work with hands-on deep learning, and/or experimental neuroscience (or collaborate with someone who can do that).
- I don't think they can get away with it when it comes to Top-500 rankings. Those rankings are based on LINPACK scores and this supercomputer would end up not scoring high enough and will be placed in the right spot in the list. So it's not a big concern to me.
- "ML isn't scientific computing" goes both ways. Sure tensor-tera-ops are not teraflops, they're specific to operations involving artificial neural networks (ANNs). But when you take the claims from the past about how much computing power it'll take for rivaling that of a human brain, and folks came up with 100 petaflops, or 1 exaflop. Well those should not be called "f"lops either. Because when it comes to brain inspired computing, ANN tensor-tera-ops are a more reliable number than FP32 or FP64. And if it's 100 peta-ops of ANN compute, well we're already past that, and can easily create a supercomputer with 1 tensor-exa-ops. And that means we have reached the hardware capacity required to emulate human-level intelligence in a machine (i.e., the only thing missing is the right set of algorithms).
Conclusion: Tensor-tera-ops are not FP ops and should not be used for placement in the Top-500 list. But tensor-tera-ops have enough significance that it warrants creating a new list Top-500-Tensor and ranking Tensor-supercomputers on that list.
Have to put up with high-school kids' "I asked GPT-3 to ..." posts for the next gazillion months.
Here, you have your 15-minute of fame, here you have yours ... Everyone gets 15-mintues of fame.
Note-to-self: Stay away from vulture-minded execs.
Note-to-investors: Keep an eye on and go the extra mile in rooting out the vulture-minded execs in your companies, before shit hits the fan.
Is HN overrun by coding noobs who are pathologically handicapped in the size of their vocabulary? How the eff is 'IntelliJ' an enterprise s/w?
It'll convince more and more that healthful longevity is a viable future.
> ... N independent event types ...
So if the robot blows a $0.03 (i.e., 3 cents) electrolytic cap [0], that's an independent event type that "killed" the machine? Just trying to get a sense of how you define "machine kill" (and as I said I mentioned a big caveat of right-to-repair laws).
- 4-year robot lifetime is overly pessimistic. If upgrade is not the reason (which has nothing to do with lifetime, it's your choice to upgrade), a robot could last way more than 4-years (especially with right-to-repair laws in place).
- That said, such a robot will most likely be available for less than $21k and very soon in the near future.
So what do you think the definition of 'privilege' is?
There are dozens of amazing open source projects posted on HN every week, and on average, the ones out of these that are outstanding and could be turned into a business model are countless every month.
So I don't understand what makes this project stand out so much that it's like leaving money on the table. (this is the first time I'm seeing a comment where the money is apparently on the table because the project is so good).
Good software product/service is a tiny aspect of doing business and turning a profit. The money doesn't appear magically on the table!
Case in point, Freeman Dyson, brilliant mathematician/math-physicist, never got a PhD degree, but did research work all his life. Note, he was deeply embedded in the research community, and pretty much everyone he collaborated with, was a PhD or eventually got a PhD (meaning he went through the motions just like other researchers).
Final note, the process and training required to be able to conduct research is massively undernurtured, i.e., in my opinion, most PhD graduates are barely an iota better-trained than MS graduates these days. This is getting increasingly true as world population grows, and PhD diplomas get handed out willy nilly (a topic I could go on at for hours). In short, Sturgeon's law is in full swing.