1,077 karma · joined November 20, 2013
That is not utilitarianism, it's just tribalism and xenophobia.
Speaking from experience as we once tried to help her to connect to a open WiFi which needed an OTP-based login via the phone and gave up after about 15 frustrating minutes for all three of us!
For example my MIL is a medical doctor, so is obviously educated, speaks English well and uses a smart phone but wouldn't be able to send email to a new contact. Same with WhatsApp, she can reply to messages from us but I don't think she knows how to add a new contact to her phone.
I totally disagree. The quality of papers at the "elite" conferences is way higher than most things I've read on HN. What is an example of an HN post that in your opinion is better than equivalent academic research in that area?
> Elite status happens exactly because there are conferences like the ones you mentioned. If you work with an advisor that publishes in Oakland, your chances of getting a paper in Oakland gets increased multiplicatively. And hint, that's not because your ideas (or papers) are better than anybody else's.
Sure, having an advisor on the Oakland PC helps a great deal. But it doesn't follow that your work is just the same as everyone else. Have you peer reviewed papers for these conferences? A majority of submissions, even at the "elite" conferences are just junk. That doesn't mean everything that gets published is not junk, but the stuff that does get published is significantly better than the average submission.
> Who cares? If the work is worth anything, people will cite it. If not, it will remain as is. Why does it matter? Why do you care if 10 people cited your work or 100 people if you are happy with the work?
Because the point of my research is not to sit in an ivory tower and produce academese that no one cares about. The goal is to have real impact on computer system design, and in my specific case, push practitioners towards methodologies that make systems more secure. That's not going to happen if no one reads our work.
Another way of looking at it is that a lot of our work is funded by taxpayer money. They aren't paying us to have fun proving lemmas that no one else cares about, the taxpayer would like us to produce research that results in tangible improvements in computer system design. In the system that we have today, the only way to have this tangible impact is to produce high quality papers that other people read, cite and build on top of.
Today, if I have a grad student interested in security who wants some ideas for things to work on, I could ask them to go look up papers in Oakland, CCS and NDSS over the last couple of years and see if anything catches their fancy. This works because there's a reasonable number of papers that I could reasonably expect a PhD student to look through. Asking them to look up all the stuff that ends up on arXiv or the IACR's ePrints is not reasonable. There's just too much stuff there and most of it is not worth looking at that.
So, you might say the endorsements will take care of this. CCS et al. could just endorse some limited subset of the IACR ePrints. But this leads to problem (ii) above. Right now, we have a bunch of high-profile researchers (who are mostly at places like MIT and Stanford) who just put up stuff on arXiv without any peer review and they start picking-up citations right away because of their "elite" status. Some of this is deserved, because these people have done good work. But some of this is also just a publication cartel where everybody cites their friends' work and make it impossible for others to break into a field.
The larger point is that in a scenario where there are so many papers that nobody could possibly look at all of them will lead to a few groups accumulating all the citations and all the awards. This will especially be unfair to researchers in the developing world -- people at places like Bilikent, Tsinghua, and the IITs -- where researchers don't have the PR muscle power to highlight unpublished stuff.
> Then learning the answer revealed that it's actually something that can be fixed --- as is always the case with these things if you think about them enough.
The author focuses on the consequent, thinking long and hard about something, but misses the antecedent, acquiring a deep understanding of a problem area.
I'd say the antecedent is as important as the consequent, if not more important. If you went to grad school you likely knew tons of brilliant people who worked long and hard on important and difficult problems and got nowhere. What they were missing was not some secret effortsauce, but that they picked a problem on which they didn't have an opening or insight into, so all their hard work was them spinning their wheels while standing still.
One thing I tell my students now is that they should always pick problems where they have something everybody else working in the same area doesn't. Like if you want to do formal verification but your background is years spent in industry as a kernel developer, you should avoid the temptation to chase the latest fad (e.g., adversarial machine learning or Spectre/Meltdown), and instead pick a poorly understood and validated module in the kernel and try to find out what you can do to make that module more secure.
There's a tutorial + web interface at: https://rise4fun.com/Dafny/tutorial/Guide. The official repository is here: https://github.com/Microsoft/dafny -- you might want to switch to a local installation once the online tutorial whets your appetite. A good initial challenge, once you've gotten past the baby stuff in the tutorial, is implementing insertion and deletion on a binary search tree with appropriate pre- and post-conditions.
I was part of a group that developed signal processing algorithms and my boss's boss pulled me aside and asked me what we ought to do about this stuff. I told him that there are a few applications where this works very well, but it's pretty niche overall and we shouldn't do more than what we'd already done. He was a bit surprised because he thought I'd be one of the kool-aid drinkers given how much I knew about and liked playing with the tech. But he deferred to my judgement.
It turned out to be one of the most significant conversations I've had. I was still a junior engineer fresh out of college, but my bosses realized that I could be trusted put aside naive enthusiasm about the tech and make cool-headed decisions about what was strategically important. And I learned to trust my intuition about what was useful and what was just hype.
Should I go on?
A well-known researcher once joked to me that programmers are associated with a company for a few few years, programs live only for a few more years, but data lives forever.
Learning English is viewed as a "job" thing -- it's what you do to have a good career. And English isn't used in the same ways as native languages are. A lot of the serious literature, art and politics happens exclusively in languages that are not English and using English in these contexts outs you as the equivalent of the "condescending coastal elite" in the US.
OK, that's fair. Is the claim that such a framework did not exist before? Or is it that this new framework was somehow "fairer" and closer to our modern ideals of justice than the ones that existed before? Because neither of these claims seems defensible.
> To deny the concept of human rights, and the validity of science because some foundational thinkers fell short is ridiculous.
But where did I (or anyone else for that matter) deny the concept of human rights or the validity of science?
What I'd also like is for you to stop putting words into my mouth, I never advocated for disregarding all knowledge generated before 1970, nor did I suggest that we must disassociate ourselves from anyone.
I just want to point out that the so called values of the enlightenment were never really put into practice. Rule of law and democracy didn't exist for African Americans or colonial citizens of color, the power of nonviolent protest and it's ability to effect change was mostly thanks to leaders who stood in opposition to those who supposedly represented the values of the enlightenment, and the relative peace of the last 60 years is a result of the devastation caused by great wars that were entered into by societies that were supposedly enlightened.
I'm just calling bullshit on the notion that the ideals of European nationalism fuelled by the industrial revolution can claim any credit for the social progress we've made as a nation. Technological progress? Sure. Social progress and justice? Not at all.
I looked up wikipedia to see who the intellectuals of the enlightenment were and it seems many of these people are racists, slave-owners and colonizers who literally thought that Gandhi/MLK were sub-human. Their military philosophy seems to boil down to might is right and their values AFAICT are pretty close to Richard Spencer's. Sure, some of these people were great scientists and politicians, but I'm having a hard time seeing them as moral exemplars.
It's the likes of who Intel figure out how you can put a billion little things each with a few tens of atoms on a chip that somehow work together to produce a useful computing device. And then Apple finds a way to create a really pretty GUI and walk away with 90% of the profit in the Industry.
As a hardware guy, I always feel like this is somehow unfair.
Every high performance processor in the world performs speculative cache updates. That did not cause meltdown, nor did it cause spectre.
I believe the number of people who would have looked at Intel's design decisions related to meltdown and thought to themselves this was a credible security holeis very close to zero. I can say this with some confidence because I used to do processor performance modeling for AMD, security verification for Intel and now do research in hardware security verification. This is not a bug until you look at the exploit.
I am certain AMD/ARM/RISC-V or whatever other Intel competitor you can think of have similar security bugs in them. It's just a question of whether people will find them. In fact, knowing what I know about AMD and Intel, I'm willing to bet these other companies have more security holes, not less than Intel
This eventually led to Gandhi marching to Dandi to deliberately violate the salt laws and was one of the key moments in the struggle for Indian Independence.
It really breaks my heart that the British killed tens (maybe hundreds) of millions of Indians for no reason at all and they've gotten away with it both in terms of any actual consequences as well as in the realm of public opinion. And to make matters worse, we still have so-called British intellectuals who like to argue that colonization was somehow a good thing for India.
Sure. I'm not arguing that bitblasting should never be done, it's probably still the best way of solving bitvector problems. My point is that the pedagogical objective of an toy SMT solver is lost if doesn't teach DPLL(T) and only focuses on bitblasting.
Most SMT solvers today use the DPLL(T) technique which transfers information back and forth between a so-called theory solver (e.g., a simplex solver that handles real constraints) and a SAT solver which handles the "Boolean part" of the problem . I feel that an SMT solver that intends to be educational must implement DPLL(T).
Link to a recent syllabus is here: http://mrmgroup.cs.princeton.edu/cos583/syllabusS15.pdf
The papers in it are all well-worth reading and are mostly accessible (with some effort) to a first-year grad student (i.e., anyone with an undergraduate degree in CS or something related.)
PS. Don't make the mistake of reading easy papers just because they're easy to read.
Trump is by no means a temporary phase. The demographic trends that led to Trump will continue for a long while.
The "big sort" means that 70% of US population will be represented by 30 senators by 2050. Guess which party the other 70 senators representing primarily rural populations will belong to? The republican advantage in the electoral college will continue to grow as the rust belt becomes older and whiter. The supposed democratic flips of Arizona, Georgia and Texas due to rising Latino population are still far away in future and in the meantime, republican voter id legislation and gerrymandering of state-level districts to control state-level legislatures will ensure they remain red for the foreseeable future.
What all this really means is that the current situation -- where winning the primary is far more important than the general for republican politicians -- will continue for at least a decade or two more. Which means we WILL have more Trumps. The scary part of this is that the next guy won't be as incompetent at implementing his own agenda as this one.
I'm not aware of anything of note coming from Baidu research. All they've done is a splashy celebrity hire of Andrew Ng, which is the exact opposite of how one would set up a serious research organization.