HBX Core

http://hbx.hbs.edu/hbx-core

Its a paid online business course by Harvard Business School with 3 modules - Business Analytics, Economics for Managers, Financial Accounting

Material is not super challenging (maybe except for Accounting), but its still a lot of work and very rewarding. There is a strong social element to the course because they incentivize students to ask and answer each other's questions. At the end of it, you have to go to a testing centre and give a 3 hour exam on everything they have taught you. I finished this course a few months ago and really enjoyed the material and all the people I met through it. Highly recommend!

:( Been using this service for 4 years.. have grown so much thanks to it. Everybody seems to be recommending Day One as an alternative, but that looks way more complicated than a simple daily email.

Edit: I take that back.. day one looks pretty awesome Found https://twitter.com/kylesethgray/status/513476424955482112 which let me import all my OhLife data into this.. and this is backed by my DropBox account which makes me feel safe about my data

I made an improved export tool that handles the line endings from OhLife correctly: https://github.com/kevin1/OhLife2DayOne

You can use Day One as an OhLife replacement by setting a notification at your email time. Then just write and ignore all the other features.

I had the same problem a year ago. For the apartments you really like, send in a cover letter with your application explaining your situation and offering more references and a larger deposit. I was really desperate and even put in my resume into the application (which had a bunch of fancy sounding NLP/CS terms). Ended up getting a studio in Pacific Heights.

Thanking you for showing me the 'dis' module. Dis is going to be fun to play with!!

I learnt Go by watching this anime on Netflix called 'Hikaru no Go'. Its pretty kiddish, but fun, and really inspires you to learn the game!

Don't worry too much. Life is like that. Had you taken care of this problem beforehand, I am pretty sure some other major problem would have turned up. When you create something new, you have to deal with 100s of problems. No one gets it right in version 1 itself. What you need to do is to be able to calmly work through all the issues that come up and think about what you can do to recover the traction your app initially seemed to be getting.

Your app is tricky because I can derive value from not when I install it, but when all my friends install it as well. This implies you guys should make it really really really easy for me to get my friends on it as well. Perhaps you can take my phone contact list and show all those users as my friends within your app the moment I install it, and send the other users an SMS (using twilio) or email if I try to interact with them using your app. This way you can disguise your cold start from the user and make it more useful to him from day 1.

I have been to 9 grubwithus dinners so far and loved every one of them.. this is a really awesome concept!! Eventually they should come up with 'Grubwithus Now' where people can join impromptu organized meals with a bunch of random strangers whenever they want to, instead of having to plan the thing in advance.

They are talking about Latent Dirichlet Allocation which was made by Blei in 2002 ( http://en.wikipedia.org/wiki/Latent_Dirichlet_allocation )

Here is video by him where he explains the basic model as well timeline-related hacks the article talks about - http://video.google.com/videoplay?docid=3077213787166426672#

You can very easily play with LDA yourself with this toolkit - http://mallet.cs.umass.edu/

I think they're talking about this one:

http://projecteuclid.org/euclid.aoas/1183143727

This work uses the correlated topic model, which extends LDA to model correlations among the extracted topics.

Edit: and this one [pdf]:

http://www.icml2010.org/papers/384.pdf

Was I just looking at formulae or some pages taken from the Odyssey?

I guess we could get the computer to categorise it ...

1) People had fewer distractions earlier

2) Thanks to the Flynn effect, people are growing smarter, so need less time for the same content

3) Thanks to the internet, its a lot easier and faster to learn something than it was earlier (more cliffnotes type resources)

4)People today can probably make much better assessments to what things will actually help them in their professional lives, and what things are random rites of passage i.e. hoops they have to jump through to get their degree

it seems like to much of a risk, in terms of the controversies that might get raised every time the recognition software screws up.. imagine the system tagging somebody's ex girlfriend on their current partner's face, or interpreting a picture of Saddam as some random friend of the poster

If you read the article,

"...the system will automatically detect faces in the photos and prompt you to select the friend who’s face it is."

It seems that the software is (so far) only recognising the presence of a face, not identity.

Also, there's been a lot of concerns over privacy with Facebook, and Facebook being able to "recognize you" has a pretty high creepiness factor.