Do you want to work at an artificial intelligence (AI) company? I definitely did back in 2010, when I was hired as one of the first ten employees at Narrative Science. I’ve been here now for almost eight years, which is forever in the startup world. (The average tenure at a startup is only 10.8 months!)
One of the most gratifying things about being at a small company for so long is that you get to see friends and comrades-in-arms grow and progress in their skills and careers. When I think back on the people I’ve seen find the most success here, a few consistent traits stand out to me. Here are some of the things I’m looking for when considering whether someone would be a good fit at Narrative Science or an AI company in general.
They keep things simple.
AI is moving incredibly quickly, and new applications and techniques are being shared and published every day. This can make it tempting to use overly complex approaches or apply them where they’re just not necessary. For example, regression analysis is over 200 years old– it’s also completely sufficient for many prediction tasks. Making any application of AI really valuable is hard enough already. There’s no need to overcomplicate it.
They understand that A.I. is a component of the system, not a magic wand.
Unless your entire organization is based purely around building backend AI services, then AI is just one piece of your system or product. Like other components of your system, it has defined inputs and outputs, a set of requirements that it is expected to meet, etc.
At their core, AI approaches are based on algorithms and data structures, just like everything else in software engineering. Therefore, you need to treat AI components like you do the other components in your system as much as possible. Don’t let yourself fall into the trap of believing that AI somehow permeates the entire system, or that it will magically perform well just because you really, really hope it will.
They care about people.
AI is an extremely powerful set of techniques and ideas, and like anything with that much power, it needs to be deployed thoughtfully and in a spirit of empathy towards its users and anyone else affected. We’ve seen first-hand some of the benefits of AI applications like Natural Language Generation (NLG) here at Narrative Science, as our product, Quill, has automated away some of the uninteresting reports that people were required to write regularly.
But the news has also recently been full of instances where real harm was caused to people, organizations, and institutions by AI operating over large, and often personal, data. I believe it’s incumbent on everyone working at an AI company to consider the effects their product will have on people– both positive and negative. I’m proud of Narrative Science’s impact on the world through AI. and that’s because the people I’ve seen be most successful at Narrative Science all care deeply about our users and people more broadly.