fnl

20 ms·

211.
by fnl 9y ago
Excellent article, thanks. HN discussion: https://news.ycombinator.com/item?id=14781426
212.
by fnl 9y ago
And quantum computing, as well as fusion generators... :-)
213.
by fnl 9y ago
"Never" is a strong prediction. But yes, ANNs have nothing in common with BNNs (biological ... :-)) at all, other than taking them as a very rough abstraction for teaching the basic intuition of the chained up tensor transformat
214.
by fnl 9y ago
Put a lot simpler: Even DL is still only very complex, statistical pattern matching. While pattern matching can be applied to model the process of cognition, DL cannot really model abstractive intelligence on its own (unless we phrase it a
215.
by fnl 9y ago
This. I still cannot forget the disappointment of my parents and some family friends, all retired scientist or MDs, when I explained them how deep learning and natural language processing works a few years ago. They were truly upset that al
216.
by fnl 9y ago
Because that has nothing to do with consciousness ... Every living cell can perceive such inputs, even the simplest of prokaryotes can "sniff" out their food sources.
217.
by fnl 9y ago
True - until some clever guy proves us all wrong and finds ways to train some multidimensional/complex/deep/... kernel/forest/swarm/... that can learn those nonlinearities that currently only deep nets can be t
218.
by fnl 9y ago
That's just another version of the trap GP spoke about. About a decade ago everybody was expecting emergent complex behavior from all kinds of evolutionary, intelligent ("swarm") systems. Didn't happen, seen that. https
219.
by fnl 9y ago
I completely agree with you on that take. However, if you accept that take, I see little chances of convincing management in such a company to switch to a new, fledgling language - even if it were much safer to use. If anything, your hopes
220.
by fnl 9y ago
As already posted in another thread for this GP, I completely disagree with the overall sentiment of this whole discussion that is throwing (modern) C++ (i.e., C++11 ff.) into the same bucket with (any version of) C. What I can agree to, ho
221.
by fnl 9y ago
There is a big difference between modern C++ (C++11 ff.) and even the most current iteration of C. The two languages might have been quite similar back in the day when "C with classes" was created, but they have diverged signifi
222.
by fnl 9y ago
Pretty much agree, and particularly on the budget/time aspect.
223.
by fnl 9y ago
Oh, that's why I had mentioned tree-ish structures, sort of referring to skip-lists & friends on a broad level. Sorry for the misleading expression. Rather, I'm interested if any ranged/set retrieval approach can be found
224.
by fnl 9y ago
Not sure I understand your reply - which of it explains how NitroGen (i.e. the indexing algorithm described in the linked paper) has evolved?
225.
by fnl 9y ago
Foremost, thank you for your work on this subject and your contributions. As this is work from 2010, and, to cite from your paper: > In this work, we will not consider finding multiple records at the same time. The algorithms given in ps
226.
by fnl 9y ago
Thanks for the work on and link to gperf! Looks interesting, to say the least.
227.
by fnl 9y ago
Absolutely correct! I guess a comparison with various levels of "rows" and concurrent access might be needed before judging any way...
228.
by fnl 9y ago
Quick access? Tree lookup is O(log n) on a rb-tree. Hashtables are O(1). And this Diskhash thing claims to scale to billions of entries. Not sure your options are comparable, therefore.
229.
by fnl 9y ago
Neither of those seem to easily scale to billions of keys, though.
230.
by fnl 9y ago
Or until they get bought up and shut their doors. Now where have I seen that one... :-)
231.
by fnl 9y ago
Most uses are in rather obscure bioinformatics packages, where the exact matching of sequences is useful, including BWT compression of the arrays. Easy to Google, but I don't know if any pure-Python versions not written in C exist, an
232.
by fnl 9y ago
As nobody yet has mentioned it: lptrace -> strace for Python programs @ https://github.com/khamidou/lptrace For Python3: https://github.com/rixx/lptrace/tree/python3
233.
by fnl 9y ago
I am also interested in that - People seemingly are complaining a lot about broken RNN(Cell) functionality/insufficient regression testing. https://news.ycombinator.com/item?id=14576912
234.
by fnl 9y ago
I also completely forgot about Caffe(2), which I recall to always have been the most easily deployable library, and possibly DL4J. http://www.cio.com/article/3193689/artificial-intelligence/w...
235.
by fnl 9y ago
> If you are building products / services that use ML and developing/training your own models (as opposed to taking pre-trained models and using them), there is really no credible competitor to TensorFlow. MXNet has amalgamati
236.
by fnl 9y ago
Polyolefins are used to make stuff like plastic bags and are heavily researched due to their substantial and very negative impact on the environment. So that this stone paper stuff is sold as a more environmentally friendly alternative to p
237.
by fnl 9y ago
Your post itself is inciting, because you imply that those up-voting the post did so due to "their world views". I am sure you can think of other, less offensive reasons that are more likely explanations of the votes, like scienti
238.
by fnl 9y ago
For your core argument to hold, terminating a process would have to lead to software with fewer bugs than software that that can report and recover from bugs. I don't see how that would be the case. Second, you did not address my main
239.
by fnl 9y ago
There is no such thing as an "orthogonal problem" here, both things connect. That's why you (a) want to have your processes staying alive in about any software as long as possible (hence invalidating your argument for #1), an
240.
by fnl 9y ago
Just to clarify: I also don't think you should bring down the whole house when no financial interests are associated with the process dying. But when money gets involved, people typically start to behave and act more professionally, so

More ›