HN user

f_sav

30 karma

My (main) interests: * machine learning, neural networks (I'm doing a masters in ML) * web dev (Django, jQuery notably) * personal knowledge management (best defined through examples, see my blog)

http://fsavard.com

Posts0
Comments15
View on HN
No posts found.

Also, having read a few (first year grad student here), I'm starting to think it might sometimes be a simple matter of lack of space. Explaining algorithm and method correctly requires examples, precisions etc. which might take away space to pack in more results and experiments. (I know some conferences have submission guidelines with page limits)

There might be "shadier" reasons, but if that explanation makes sense, I for one would like it a lot if some researchers got in the habit of having longer explanations somewhere, say on their website. I don't know if that'd somehow break implicit rules in the academic field, but it'd lead to less head-banging-in-the-wall-trying-to-understand-what-that-damn-symbol-means.

He had me at "code/history" division. I really don't like the way I must usually enter functions in the basic interpreter (line by line, can't go back... unless I missed something?).

To me that's an interesting cross between an IDE (with completions, snippets etc.) and an interpreter.

As much as this would make for a nice concert/show gizmo, I'm also interested in those tiny helicopters themselves. To control them with such precision would be impressive (well from a non specialist point of view). Yet I can't seem to find info about them on either of the labs' sites (this one and the ARES lab). There could be many other use for such swarms (I'm thinking controlling a local set of cameras to get different angles, but I'm sure there are so much more).

To me it really depends on the type of material. Obviously if I'm reading a math paper, I may re-read the same sentence half a dozen times, so it doesn't apply.

But there are clearly cases where I want to read as fast as possible, without necessarily skipping. E.g. scanning the news, or reading a novel in a shallow fashion.

(But yeah the blog post talks about learning techniques, so I don't know how relevant it is)

To add to the list, for those who don't know about it, there's spaced repetition software. There are some situations where rote memorization is indeed useful, e.g. learning vocabulary (well _using_ the vocabulary is better, but memorizing is a good first step). I personally use Mnemosyne, but there are lots of such programs (the most well-known being Supermemo), many web apps too.

Here's a database of reviews of flashcard-type software: http://learn.how.to/

Makes sense (esp. these last few years where device come with the least physical switch possible, e.g. the iPhone), but given the benefit, I personally wouldn't mind the added clutter.

One good first step would be to cut the link to the microphone on lid close (for a laptop or a cellphone, say). Or have a global "privacy switch" which would still allow apps to run to but electrically cut off the privacy-invading sensors (mostly gps, camera, microphone... other?).

Just dreaming out loud here...

So that'd mean some schools (or is it the same one?) did such spying without even trying to conceal it. That it hadn't occurred to students to cover the camera with something is a bit troubling too (either they don't realize the possibility/risk or they don't care).

I may sound paranoid, but having integrated cameras in almost every electronic device around, with no easy physical (vs. software) way of blocking it somehow, is a recipe for such situations happening one day or the other (of course that doesn't make it any more OK).

Oh, and having tried it a bit with a twitter search feed, maybe you could have some options for formatting too (with replacements like %t for title, %u for url etc.).

(But of course both these suggestions break the simplicity of it, which has its appeal.)

EDIT: also, I'd add the basic instructions (STOP, START etc.) at the bottom of the front page.

Overall good idea, but I'd say: obviously I wouldn't use this for my regular feeds, only for some specific ones for which I want live notifications. So maybe some basic filtering options would be good (by keyword, by author...).

(Or maybe it's implied people would use existing services for filtering?)

The technology looks very nice, if a bit slow to react at times (from the video). But maybe that's just a voluntary kind of "momentum" introduced in their software.

(And getting to make choices for a movie based on his research is one heck of an opportunity to imprint minds with the idea "this is the way of the future" if the technology ever does come out commercially)