https://en.wikipedia.org/wiki/Twitter#Tweets#!1,3013993710

Interesting.

Pin It… Pin It Good: Making the Most of Pinterest

Lots of copyright non-infrigement: http://www.pinterest.com/search/pins/?q=death%20note
I guess pinterest users didn’t read the part about copyright on your post.
On a more serious note, it looks like google/bing’s image search service, only you get to click the uploader name and get more images(that may or may not be related to the first one)

The Daily Post

We’ve talked Facebook, Twitter, LinkedIn, and Instagram. But for many blogs, the biggest source of traffic from a social network isn’t any of those: it’s Pinterest.

If you’ve eschewed Pinterest because you don’t care about ombre cakes or repurposing wooden pallets for home decor use, you might be missing out on a huge audience for your blog (and some delicious cake). Today, we’ll push past the inspirational quotes with beautiful typography, turn left at the green smoothies, and explore whether Pinterest is a good match for your blog.

View original post 953 more words

FFmpeg, keepvid.com and Pyglet

I think if I were to use these 3 tools my video search website would be feasible. Problem is, it’s so much work just to get to the part where I need to use my full blown creativity to create auto-taggers from scratch that I get lazy just thinking about it.

Another project would be my quote website. I could limit quote length to avoid copyright problems… however, javascript and flash text would remain a problem. This one would be a lot easier, certainly a lot easier than writing that redlink add-on. I could pull this one off in less than a week. But it sort of seems like going back to a ill-fated project.

What about the composite hedge? I got distracted half-way when I thought I had found an error on my broker part. I tried to exploit it only to realize I was confusing 1 “pip” with 1 “dollar”(the value of 1 pip varies according to currency pair being traded).

The composite hedge website would be easy too. At least if I avoid stock/commodities being valued in foreign currencies at first.

KeepVid.com vs DOwnloadHelper

I always thought the reason we needed a plugin to download yt videos was that a webpage doing just that would’ve been taken down due to copyright issues. But there it is, a webpage that renders downloadhelper obsolete.

Makes me want to code a webpage that extracts quotes like my add-on did(add-on that I couldn’t get hosted on mozilla’s official website)

January Effect

Just ran an algo that extracted the number of positive(against previous months) months in the Dow Jones Index.

January: 68(months closing higher than december)
Feb: 67 (… than january)
MArch: 65
April: 69
…

I don’t see any January effect(I used about 110 years), December seems to be a little above the rest of the pack(75). But the csv file starts halfway trough the year so slightly higher values for end year months were expected.

Also, I’m left pondering over whether wikipedia’s article about January predicting the rest of the year includes January when measuring the year. If so, it wouldn’t be surprising/profitable to see a higher number of positive years containing a positive january.

Open Analytics Protocol Comment #1

The one thing I haven’t thought about is how auto-taggers developers are going to get cash. I thought about adding the result plus an ad link but clients could not show the suggested ad and just extract the information(whether a tag applies or not)…

How do I warn the autotagger it’s ad has been shown? Seems impossible with just php and firefox. I would probably need a modified browser or browser extension.

I could centralize everything I guess but that would mean checking other people’s code for exploits… such a pain.

Minimum Wage

http://www.bloomberg.com/news/2014-01-14/seven-noble-laureates-urge-increase-in-u-s-worker-minimum-wage.html#!1,2313636928

How about this small stimulus by increased spending being matched by a small decrease in business expansion/creation?

How come people who are paid to think logically can come up with this free lunch where entrepreneurs get less money but still work with the same effort? That makes NO SENSE AT ALL!!! This is logic people. It comes before mathematics and statistical models.

Open Analytics Protocol Draft #2

A php page that outputs only ONCE the following tag:
<MAGIC_NUMBER_284537_OPEN_ANALYTICS_TAG_RESULT>

Closed by:
</MAGIC_NUMBER_284537_OPEN_ANALYTICS_TAG_RESULT>

Between those two html tags one should output one of the following:
(TRUE|FALSE|ERROR)

The website shall receive by POST METHOD the following:
Tag Name, the name of the tag being request $_POST[“tag_name”]
-If this field doesn’t match the tag a page is supposed to output, one shall output ERROR
Data Url, the link to the data to be analyzed $_POST[“data_url”]

Example:

<html>
<body>

<form action=”AnalyticsSubPage.php” method=”post”>
<input type=”text” name=”tag_name”>
<input type=”text” name=”data_url”>
<input type=”submit”>
</form>

<?php
$_RESULT=”ERROR”;

if ($_POST[“tag_name”]==”English”)
{
$raw_page= file_get_contents($_POST[“data_url”]);
$stripped_page= strip_tags($raw_page);
$freq_ar= count_chars($stripped_page);

$en_letter= $freq_ar[ord(“t”)]+$freq_ar[ord(“T”)];
$ot_letter= $freq_ar[ord(“r”)]+$freq_ar[ord(“R”)];

if ($en_letter>$ot_letter)
{
$_RESULT=”TRUE”;
}
else
$_RESULT=”FALSE”;
/*
arsort($freq_ar);
foreach ($freq_ar as $char_code=>$freq)
{
echo $char_code.”, “.chr($char_code).”: “;
echo $freq.”<br />”;
}*/

}

echo “<MAGIC_NUMBER_284537_OPEN_ANALYTICS_TAG_RESULT>”.$_RESULT.”</MAGIC_NUMBER_284537_OPEN_ANALYTICS_TAG_RESULT>”;

?>

</body>
</html>

Open Analytics Project

Now that I think about it, Distributed Analytics would probably be a better name, more true to the nature of what I’m about to do.

(listening to korn – got the life)

I’m sticking with Open Analytics, open in the sense that analytics are right now done mostly by big companies with lots of data.

Protocol for Open Analytics:

Client(requesting tag) sends to server PHP page(by POST method):
action= …/tag_name.php
type=text name=LINK value=www.example.com

Server receives http request for tag_name.php
return decision_function($_POST[“LINK”])

Client:
Shows result to browser or stores result in some database, etc.

(the most important parts of)Everything that I know

Lots of stuff missing but the most important(why hasn’t AI surpassed organic intelligence) follows.
============================
TIME FRAME #1
Physicists’ entropy seems to be the opposite of Information Theorists’ entropy.
HOT COLD COLD HOT
COLD HOT HOT COLD
AVERAGE HOT COLD COLD

Physicist: Not much entropy
Information Theorist: Lots of entropy

AVG AVG AVG AVG
AVG …

Physicist: Lots of entropy
Information Theorist: Not much entropy
===========================
TIME FRAME #?
The universe is expanding. Something located in an expanding universe needs more bits to represent it’s location.
More bits. Less entropy(information theory).
Decompressor-like universe.
===========================
TIME FRAME #2
Metabolism seems oddly similar to what Hutter(and I) understand as intelligence.
===========================
TIME FRAM #2.5
//TO DO: Biology related, will be long.
//TO DO: Economy related.
===========================
TIME FRAME #?
A decompressor with patterns accross it’s data would benefit from a global probability distribution.
===========================
TIME FRAME #3
The universe can send information faster than light speed.

….

Basically, I think organic brains have a deeper connection to a global probability distribution(that is decompressing the universe) that would give us more combinatorial power than machines.

Strong AI is based on:
-Machines getting smart faster
-Humans not getting much smarter

I think the second is false.

Also, if the universe were trying to decompress a hash(that is, trying to decompress without a map, with methods that may fail) one would see laws of physics popping in and out of existence, which could explain apparent randomness at the lowest levels of our universe