I have decided to take down mean tweets on twitter one by one - https://www.twitter.com/MeanAlert. When somebody tweets “I hate talking to drivers on Uber”, they get an auto-tweet back.
I was a bit annoyed when I searched I hate Uber on Twitter, lots of people were hating on drivers,
People are sharing the same physical space yet tweet out their frustration, for me this culture of not having uncomfortable conversations is a downward trajectory.
Why I did the “I Hate Uber” search
- The use of the word “Hate” is used too freely.
Some problems are just not problems, bitching about a chauffer you’ve summoned at the press of a button is not a problem.
Look around, your life is amazing, don’t be a jack-ass to somebody not as well off as you.
So I built the bot
This was the first tweet that went out,
I was surprised when I got retweeted, I’d like to think I made a point and changed somebody.
Not everybody took it well, infact just the second tweet was replied to by a person who made an uncomfortable (and if I can say so), poorly articulated, comeback.
I lol’d. I’d like to think I made a point there too.
I lol’d, this person looked really tough, but I was happy hetook it in stride and didn’t say anything back.
Under the hood
So let’s take a look at the code:
Some Problems that need solving:
I was manually starting/stopping script (did 10 tweets)
- Some responses weren’t accurate, especially when the script was run late at night (when people aren’t taking or talking about Uber)
- Response tweet (harcoded) was not dynamic. Solution I was working on involved scraping tweets from a gist.github.com file, I never got there:
If I make my bot dynamic it can extend from this uber driver experiment and taking on more mean tweets.
** Updates to code for extensibility **
With the help of Sam Lavigne here are some updates to the code for extensibility. I’ve made tweets dynamic (they’re pulled from an editable Gist file and not hardcoded anymore).
Here’s what the code looks like now: