laretluval

12 ms·

61.
by laretluval 6y ago
The discrepancy could be because Hacker News cares more about technology platforms than Alabama public TV. For an example of HN getting upset about government censorship on a technology platform, see https://news.ycombinator.com&
62.
by laretluval 6y ago
It’s only been a few months. If it really does robustly get 4.6 perplexity on PTB (less than I ever thought was possible) then it will receive its due recognition, at the very least via the Hutter prize.
63.
by laretluval 6y ago
Not great, not terrible
64.
by laretluval 6y ago
If the paper is on the arXiv then it will not be anonymous to the reviewers.
65.
by laretluval 6y ago
Where are people going in lieu of Youtube?
66.
by laretluval 6y ago
Solving problems doesn’t necessarily mean getting numerical answers. If you don’t know how to apply the equations you list algebraically then you don’t really understand them.
67.
by laretluval 6y ago
Modern methods for deriving word embeddings easily beat LSA.
68.
by laretluval 6y ago
People love making complicated models. It’s fun and feels like you are having insights. Fitting a boring old ARIMA feels like work.
69.
by laretluval 6y ago
> get things checked by a credible authority Who decides what's a credible authority? How about critical thinking instead?
70.
by laretluval 6y ago
The species is becoming more and more domesticated.
71.
by laretluval 7y ago
Once everyone is working from home it is going to be prime time for terminal-based yak shaving.
72.
by laretluval 7y ago
They only find the effect for cases where there is low flux in the magnetic field. This indicates another reason to be skeptical of the study. If an effect is found in one condition and not another this is a red flag for p-hacking.
73.
by laretluval 7y ago
Current academic NLP would have been considered quite out-of-the-current-box 10 years ago. Most academic progress is driven by young graduate students who think similarly to you.
74.
by laretluval 7y ago
What are the algorithms that can dismiss “indirect correlation”? How can you tell a direct from an indirect correlation in data?
75.
by laretluval 7y ago
Fun to wonder if this is what Plankalkül would have evolved into.
76.
by laretluval 7y ago
I’d revert to my prior, just like here.
77.
by laretluval 7y ago
Yes, I would bet that there is no causal effect.
78.
by laretluval 7y ago
100,000 lines was enough for Terry Davis.
79.
by laretluval 7y ago
Indeed, as revealed in his court documents, Ross Ulbricht paid a lot more more money than this (by an order of magnitude) to have his darknet hits carried out (although he was likely being scammed). The price mismatch makes me skeptical of
80.
by laretluval 7y ago
I suspect that psychometric tests would also be able to predict success in terms of having friends/relationships and living a low stress life.
81.
by laretluval 7y ago
Is there a citation for that?
82.
by laretluval 7y ago
Sign me up!
83.
by laretluval 7y ago
An alternative would be to present the version of borders which is most likely to be relevant for a traveler in terms of border crossings, visa requirements, etc.
84.
by laretluval 7y ago
Scales are universal, but the distinction between harmony and dissonance is not. https://www.ncbi.nlm.nih.gov/pubmed/27409816
85.
by laretluval 7y ago
Yes, it can be done. https://openreview.net/pdf?id=S1eZYeHFDS Natural language is harder.
86.
by laretluval 7y ago
That is the official Unix philosophy, but a glance at just about any man page will suggest that it's never been followed.
87.
by laretluval 7y ago
> Well except quantum physics might get in the way, This is a pretty huge well except. When people talk about information that can be stored in the universe this is exactly the limitation they have in mind.
88.
by laretluval 7y ago
Honestly hunting for the URL is half the fun
89.
by laretluval 7y ago
What should I google to read more about this kind of thing?
90.
by laretluval 7y ago
And evidence that sapiens x neanderthal offspring were not always fertile, see elsewhere in this thread.

More ›