The Many Flavors of Bot

Artificial Intelligence February 6, 2018

The Many Flavors of Bot

Lots of bot talk going around, isn’t there? Don’t get lost in the robotic shuffle. Here’s a breakdown of all the breeds of AI helping us navigate our gadgets, jobs, and entire lives:

Platform bots

Some bots exist only on closed platforms— they live in their own little worlds. And what great worlds they are: on platforms, bots can leverage several of their properties.


  • UI. Bots make great user interfaces. Messaging platforms like Slack, Hipchat, or Telegram all allow the user to interact with bots through simple text messages.
  • Distribution. Automating elements of platform distribution channels is incredibly handy: it takes the legwork out of communication. A great example is Nightbot, a moderator chatbot for the webcasting platform Twitch which went viral.
  • Data input. If you hook up a bot directly to a platform and feed data through it, you’re unstoppable. We’ve all heard of Twitter and Instagram bots: they access info from these social media platforms and automatically favorite posts containing keywords. A more dignified example is perhaps Youtube’s copyright bot that scrubs through videos looking for copyright infringements.

Standalone bots

Unlike platform bots, some bots roam free. While integration can be nice, they’re not restricted by a platform and aren’t subject to its whims. These bots are defined by what they can do. So, what can they do? Check it out:

  • Collection. Some bots crawl around the web just collecting data. This is how competitor price monitoring stays on-point.
  • Processing and Parsing. Bots can parse huge amounts of publicly-available data  (à la RossIntelligence, which specializes in legal knowledge) or process your own self-supplied data (à la resume parsing services)
  • Personal use. Let’s not forget the humble script bots that simply complete predefined tasks once triggered. You can set up for yourself using services like Zapier, Hoist, Dexter, and

From script to AI-powered bots

Not all bots involve fancy stuff like machine learning or deep distributed learning. It’s a spectrum of sophistication! There’s actually a continuum stretching from script bots (no intelligence required) to cutting-edge software like Tesla’s self-driving AI (pretty damn smart.) In the middle of these extremes, we have things like Birdly, a mildly sophisticated bot that processes receipts you send on Slack and Clara, a virtual assistant that partly relies on AI.

Where are we headed?

Bots are everywhere, and all bots matter. Not only well-funded, majorly-innovative technology is changing the world, but also the tiniest script bots: it’s the rise of the DIY bot. People are creating their own little “companion bots” crafted for their precise needs, which is a huge step in the democratization of this technology. This same movement is happening even with the biggest of companies: Facebook, Google, and Airbnb are open sourcing their automation technology, making it easier to build products that use it.

Bots Abound

Maybe bots are not the backbone of all software just yet, but they’re already popping up on the sidelines as add-on features (example: Gmail’s smart reply function.) Another cool trend? Combining human intelligence with software intelligence. That means programs which are not 100% AI— they include humans who check and improve the output! And it doesn’t end there: right now, bots do work that’s deemed “repetitive.” But creative tasks are next: it’s not hard to imagine a software that can design a logo using a database of existing logos. It just needs to spend a few seconds analyzing current design trends to come up with something flashy. Innovation does not only happen at the “cutting edge” end and the future of software automation does not involve intelligent “overlord bots” only. There’s a huge number of flavors of bot to choose from, and as new platforms emerge, that number will only go up!