Confirmed: Google has Acquired Aardvark

Why aardvark? u guys had something great going on here.. selling out so soon, with so much potential to make you much richer than the $50M google gave u… just like smule who had a fantastic head start with the ocarina and other apps.. they raised $3.9M dollars thereafter. the aardvark community grew strongly quickly and created a niche that no other search engine does nor did. real time answering from real people. sure, u have yahoo answers et tal.. in those u hope that someone knowledgeable will pick up the glove and provide you with the info you need. by using vark, u can get good answers in minutes, and the more you ask, the more the system learns about how to route the question to the right person and figure out which answer was a good one. there was an attempt at the beginning of the interent revolution. at the time i was at my 2nd year in college studying computer scient. one of our teachers started a small venture where people answer each other’s questions. the implementation was not at it’s best, but it kinda worked. after two years of operations the money had run out and the creative ideas of how to make money dried as well, so they shut it down (for the life of me, i cannot remember the name of that company). what vark is doing differently is several things. one, it uses jabber/xmmp protocol to be present right there alongside gmail contacts with google talk, or any aggregator (like adium) that supports this open source protocol. the heart of the system is the algorithms to analyze a given question, suggest the right tag (ie topic) and route it to whoever is a good authority in the field (usually more than one person). those people are online (naturally) and can either pass or answer. you can then continue discussing the issue with them, rank it as a good answer or flag it as an inappropriate one. if u did not get the answer you want, you can resubmit it, change the tags and hope for the best. i started testing aardvark since the day it launched it’s beta and later on the iphone app. at the beginning there were not many users, and not so many questions to answer, but quite slowly, over a period of two months it was working really well. best case scenario, i got a solid answer in less than a minute. nothing i could do with google or any other search engine. the social aspect is interesting as well. people love giving answers to what they know about. there is an innate desire in humans to help each other out… u may say it is an ego thing (and to some extent u are probably right), however, people get our of their way to provide answers sometimes, doing basic research, giving good links and more. so why is google buying aardvark? good question. two answers: one, the small team of aardvark are ex google employees with top notch management with yale and stanford grads. these are serious people with the creativity and leadership ability to makes things work. they understand google and it’s weaknesses and have found a way to monetize. two, real time person to person (rtp2p) search does not exist. and the aardvark team has shown us that it can be done, and done well. so where is the money? meaning… how does the aardvark makes it’s profit. good question again. and the answer is simple: just like google’s adwords, the aardvark places results that are paid for. for example, let’s say someone is asking the community if they can recommend or know a computer technician in the upper west side of manhaatan. some people will respond giving their input, but this question can be routed directly into a paying customer, who wants to be notified immediately when someone is looking for a technician in the metropolitan area. long short – a lead engine. i have answered several questions and provided links to products on amazon, using my amazon referral id, so why not others in a streamlined maner, where the vark directs all of those potential leads to the user. aardvark also plants automatic answers. i was questioned by the devel team about the quality of that feature a couple of months back, when i was looking for someone to build a loft bed in my studio. i got an answer from service magic, a well known lead generator for all type of construction work. if haven’t tried aardvark u definitely should. they are at, they have a free iphone app and they can be added to your fav jabber client as [email protected] Confirmed: Google has Acquired Aardvark.

why we need the semantic web: or how google can do a better job

google provides me with email alerts whenever the word “pengas” pops up on the news/blogsphere. we are a small family and i am interested in learning if there are members of us out there in the world (like greece, israel, US, canada and australia).

today i got an email from this service which, supposedly, had found a mentioning of “pengas” in it. when i opened the link (Lucky in Love: A Flower for My Hair), i did not find any pengas. what i did find is penga. more specifically: “…I basically followed the tutorial Penga from Convention of a…”.

so what is going on here? my alert is for “pengas”, and here i am with a result of “penga”…
my guess is google is using a rule engine to better understand a word. gramatically, pengas can also be penga’s. for example, think of the game “pengas’s peril” which is everywhere. i get notifications about it as well.

so for google’s engine penga and penga’s are related, hence, i get both results if i omit the apostrophe. but what google doesn’t get, is that i am not making any abbreviations. pengas is my fully qualified last name.

here is where i believe the semantic web will make a difference. by using metadata to describe the essence of the word (or context) i will be able to instruct google that pengas is NOT penga’s. it CAN be pengas’ (for singular or plural), but please don’t waist my time with redundant results.

unfortunately google is lagging behind it’s services in other areas as well…
take a look at this page

this is has 54 map locations that i have created on a google docs spreadsheet. using the map gadget, i’ve put them all on a map and published it as a web page and then embedded it into the site.

this page take over a minute to load on certain browsers. on chrome (mac version) it does not load at all.

absolutely ridiculous. i actually published this issue in the google docs help forum

as u can probably see i got not answer since november of last year.

long-short. google needs to stay on it’s tows and get things together. it is these little things that will separate them from their competition. search sites like wolfram alpha.

animoto review

Animoto is online service to mashup your photos and music into a cool video. the 60sec intro got me fired up and i immedietly opened an account, pulled out some flickr photos off my photostream, uploaded a fantastic brazillian song and waited for the magic to happen.

Animoto allows 15 sec of videos for the free accounts. for anything more you will need credit. makes sense i guess. i would pay for well put together videos that take less than 5 minutes to put together.

Animoto claim their backend algorithms analyze the music and photos to create the right transitions that emphasis the music, colors and all the glory of my sound and images. that is just fantastic on paper, and literally means no two videos are the same.

aaah what an awful video guys.. i was excited to try this new service but i guess that this style of a video i wanted to created simply does not work. a couple of issues:

1. the photos animoto grabbed from my flickr account were reduced to low-res yucky looking ones. that is probably the worst move here. are you trying to save disk space? if the video is not pleasant to look at, it does not matter how fancy the algorithms are for transitioning with the music.

2. i chose a slow song yet the photos rotated every two seconds – what is the idea behind the “we’ll match the movement to the music”? clearly there is no correlation between the two, and if there is it is done poorly.

3. the effects used to move in between the photos just does not match the tempo of the song.

4. a 15 second movie is simply not enough. how about a single credit per new user to make a full length video?

as you can probably gather by now i am not excited at all to try animoto again, though i was really excited after watching the 60 second intro priorly.

however, i do see how this service can be used for fast paced videos, and i can only assume i have way more control with the full-length videos (which costs $$). maybe i will try to create another presentation to a fast techno beat instead of trying to recreate the tranquility and calmness of brasil.