Competition vs. Collaboration: Which Creates More Value for the Company?

Deepmind’s AI defeated human champions in the game of chess and Go. Open AI won Dota 2. Facebook’s AI won the game of 6-player Texas Holdem. What’s the next frontier in AI research?

Deep Mind published a paper in 2019 about a game called Hanabi. This is particularly interesting, because while the best AIs can win games of competition, these AIs are not winning the games of collaboration yet. 

What is Hanabi? Hanabi is Japanese, and it means fireworks. The goal of this 2-5 player game is not to beat the other players. Instead, the game is won when all the players collaborate to build 5 different card sequences, like setting off fireworks in the right order. Only very little information exchange is allowed. In this game, the AI can play with a copy of itself which it knows quite a lot about, or play and collaborate with other AI agents which it knows very little about. 

Even in the easiest situation, which is the AI playing with itself, it still severely underperforms compared to the best human players.

In other words, AI knows how to compete, but not how to collaborate, yet. 

What kind of person are you: are you more open for collaboration or competition? 

I’ve never seen job ads from companies stating “we’re looking for people who are relentlessly competitive to join the company”. These companies would most probably say, “we are looking for team-players.” So of course I always say, yeah, I’m a great team player.  

And you know very well the reality. Sure, they hire team players, but sometimes the higher you rise in the company, the uglier it gets. The competition among high-level executives becomes cut-throat and toxic. Maybe your managers don’t personally want to participate in these political games, but they probably also feel they don’t have a choice. 

It’s a dog eat dog world.

But think about this: what would create more value for the company in the long run? Collaboration or competition? 

In the game of competition there will be resources spent on bringing other people down. Sure, not all the time, because some competition could be healthy, they said, but if there is competition in the company, it can’t be 100% healthy. Which means, some energy is lost on destruction, instead of construction. 

In the case for collaboration, no energy is spent on bringing others down. As a whole, simply having more concentrated resources overall for constructive behaviors, gives the company a better chance to grow and survive in the market. 

So in theory, collaboration should create more value than competition, and why doesn’t it seem true sometimes? One problem could be that management doesn’t know how to work with this energy, because it’s easier to just tell employees what to do.

“Why don’t you do what I tell you to do? You’re not collaborating with me!” (That’s probably how the management sees it). 

The other problem could be collaborating but in the wrong way. Having long and unnecessary meetings going over powerpoint is not collaboration.

The reason why we see more competitive behaviors in organizations is because it’s much easier not having to consider the strengths and weaknesses of everyone and then figure out how to work with them. 

But hey, AI knows how to compete better than you already.