Share good ideas and conversation.   Login, Join Us, or Take a Tour!
Corvus's profile
Corvus




Hey!

I'm a mathematics student, but I have a wide range of interests ranging from anything science-related to anime and sci-fi. I also reeeeaaally like robotics and the like.

I'm proficient in Java and manage my way around C, and am currently working on a couple of projects using microcontrollers!


stats
following: 30
followed tags: 34
followed domains: 3
badges given: 0 of 1
member for: 1257 days
style: clean



comments 0

HOLY SHIT. This is amazing. Gorillaz is the goddamn best pseudo-band ever. Demon Days were amazing and there were more than one times in my life in which I listened to Plastic Beach every day. All of their work really, is amazing, in my eyes. I'm sooo excited for what this album is going to be. Every album is usually a different sound form the last, so I hope this one is going to be something new and awesome!

I didn't know that facebook's targeted ads weren't blocked by adblock. I don't use facebook much though, I only log in every now and then to check for messages and stuff, though, so I wouldn't be able to tell. I don't see what would make an extension like adblock unable to patch those out though, unless facebook was actively using some random placements and/or serving them from a lot of different servers.

Heh, facebook ads are very very aggressive. At a certain point there was a way to target one person.

Anyway, I actually don't find hubski hideous. It definitely takes some getting used to, but so does reddit (and look at it's popularity!). After that I started finding it quite lean and streamlined. These are opinions though.

Curating ads is a pretty cool idea, but I can't say I support it when there are alternatives like NPR's model. Besides, most people use adblock/ublock and I doubt that many would go through the trouble of disabling it for hubski.

Oh goddamn I thought I'd read through that thread. Oh well, thanks!

Idea #6 is sort of already in place, if you didn't know: SWAG!!

I feel like having ads on hubski would severely undermine the experience the hubski team wants us to have -- ads would make hubski look more like a conventional social aggregator, i.e. reddit, i.e. easily digestible content that panders to the lowest common denominator, instead of the awesome website with quality posts and discussions that go on in hubski. Not to say that it wouldn't work, it's just that I've noticed that the interface is very much so designed in order to encourage thoughtfullness and I feel like it ads would detract a lot from it.

That said, hubski needs money. I commend you for giving a lot of alternative funding models, but I'm interested in hearing your thoughs on the model currently being tested.

Edit: In my mind the NPR model would be the best alternative (listed here) to hubski's the model that's being tested now. A close second would be the value added model, which is pretty much the NPR model with benefits for people who donate (thus encouraging more donations).

Yeah, I was pretty psyched about it that night. Awesome, I'll probably keep writing part 2 tonight!

On another note, I'll probably be on the IRC for the next couple of days/nights so maybe I'll catch you there!

That's really cool! Can I ask what you're using as / whether you're using a framework for your neural networks? When I started my current project I started playing around with a few different frameworks so I could simplify most of my work, but I ended up starting to write my code from scratch (efficiency was a major concern).

Machine learning is a really interesting topic (with soooo much stuff to talk about) and I'll definitely talk about it.

Anyway, it seems like you've got a very cool idea and I want to know more about it!

Awesome, I'm glad you understood. I was partly afraid that I wasn't being as clear as I wanted to but it's cool that people are understanding what I meant to transmit. I'll definitely touch on google's creepy images!

Hehe, that's pretty cool. It's a really interest topic and it's getting a lot of attention right now (thanks for making creepy pictures of dogs google!) and I felt like it would be cool to explain it.

Definitely, and this is one of the most important things I want to discuss after I finish writing a reasonably comprehensive rundown of the rest of the inner working of artificial neuron networks. There are soooooo many possible applications for neural networks due to their intrinsic ability to adapt to problem, as long as you give it a reasonable input and output.

Definitely does! I'll also talk a bit about learning algorithms for these, though if part 2 gets too long I'll make it a part 3.

posts and shares 0/0