image1 image2 image3


Assisted decision making

Scott Adams (the man behind the Dilbert comics) has a blog where he posts some wacky ideas, some of which are about the future of technology and humans. Today, I read this one titled 'The Era of Humans - Ending Soon'.

He goes on to argue how decisions are made by weighing imagined pay-offs for different decisions and the one with the highest imagined pay-off is the decision made. Many gadgets and apps (fitness trackers, for example) are already helping us with these decisions by helping compute imagined pay-offs in a better way.

My biggest gripe with all these gadgets and apps in providing assistance to my decision-making is that the algorithms are centrally applied and are based on my (and those like me) previous behaviour. In most things (books, movies, music, sports, friends, dates, clothes, etc, etc), I don't think monitoring previous behaviour (of me and others like me) is enough to introduce me to new things I might appreciate.

For every set of users, there are leaders (early adopters) who like to actively try on new things and followers (laggards) who rely on trends and the 'in-things' to decide what they use and buy. All the technology is focused on the followers and there is nobody offering anything to the leaders.

Since the assisted decision making happens based on algorithms taking in several parameters, I would love to have a place where I can define these algorithms with parameters coming in from various apps that I use. A lot of apps allow this today, but only within their own ecosystem (like Facebook allowing to customise your newsfeed), but nobody provides a cross-platform solution.

If I'm still sounding vague, let me give you an example. Instead of relying on 8tracks or Spotify or Youtube for discovering new music, I should be able to get a list of 10 songs computed in the following manner every Sunday. 40% weightage if the song is in Billboard top 20, 30% weightage if any of the twenty friends I have selected have shared/liked it on any of their social networks, 30% weightage if it was released in the last one month.

I should be allowed any number of parameters measured by any of the existing apps and use the social graph of any of my connections across the app/web ecosystem.

The audience for such a tool might be low (likely only the early adopters) as most people won't want to define this (because that would then move them from laggards to early adopters). But I think it's worth the effort to prevent everyone from being a follower.

Share this: