Year: 2018

Effectiveness of search methods for your own tweets (from best to worst):

1. Request your Twitter data, wait for notification, download the archive, load it in Excel/Google Sheets and issue a “find”
2. Search in Google with your @handle in the query
3. Native Twitter search

Apparently, I could even finish the sentence…: https://t.co/RoCpi5CnLF

Long-term success of a project concerns me way more than its immediate popularity.
3 key q I always ask:

1. How long have you been exposed to the problem you’re trying to solve?
2. How others, today + past, tried to solve the problem?
3. How your product will evolve in next 5y?

@Dicedotcom ??

Maybe the drug of the future will be a better version of your life. Courtesy of #VR, #AI, and neural interfaces.

If reality can be altered in a lifelike way via GPUs and #AI, who wouldn’t want something better than empty skies? But once you get used to it, who’d want reality anymore? @kschwabable > Windowless planes are coming, and they’re not as bad as they sound https://t.co/Gd2KPscREW

Over 4.5M apps in the iOS store means that your chances to be noticed, especially if you develop a niche app, are very slim. Unless you have massive marketing budget (independent devs don’t) or massive network (& you are highly influential) or your app is game-changing. https://t.co/C8Cyr1HxRR

I’m a huge fan of @appannie. And this is a very interesting report > The Data Behind 10 Years of the iOS App Store https://t.co/GI4Npiin4M

Fascinatingly simple and yet profound observation. A lot of aspects we take for granted on how jobs work actually are solutions for challenges that don’t exist anymore or are fading away > Synchronize your watches https://t.co/JHP8ywMjif

Now this is something really interesting. I wonder how it compares to Google AutoML (which is still in alpha). Any data scientist wants to comment? #AI > Amazon SageMaker Automatic Model Tuning: Using Machine Learning for Machine Learning https://t.co/Ss7ZwmlZgu