Even though we live in an era of “big data” and huge amounts of our internet usage and content consumption are governed by algorithms (Facebook’s newsfeed, YouTube’s related videos, Google’s predictive search, the advertising we’re served online, etc.), many people don’t trust algorithms when they’re presented with the opportunity to use them in their own decision-making.
Berkeley Dietvorst thinks this results in people making a lot of very foolish decisions, and wasting a lot of time, money, and effort.
So, he’s been researching the concept of “algorithm aversion” for several years and he’s published several highly illuminating papers on the topic.
Berkeley has developed a theory of why humans don’t like to use algorithms (they’re probably chasing perfection in their predictions and they excessively punish algorithms for making visible errors) and he continues to work on understanding ways in which we can increase the trust that human decision-makers place in algorithms.
Check out this conversation with Berkeley to hear:
- Why humans avoid using algorithms to make decisions – and what Berkeley has discovered about how to make people more comfortable with algorithms
- What – if any – are good reasons to avoid using an algorithm to make a decision?
- How our cognitive bias can cause us to make bad decisions (about where to invest, what route to take to get to work, etc.) – and how basic algorithms can make all of our lives easier
- Google Play
- Or stream here:
- If you’re enjoying the show, why not a leave a review? It makes a difference in terms of other people finding the show.
Check out more from Berkeley here:
- [1:28] Berkeley is a marketing professor – yet studies algorithm aversion
- [4:22] Humans are algorithmically averse – what’s our problem?
- [12:10] Humans are risk-seeking so will choose not to use algorithms in order to seek outsized reward
- [19:02] Humans err by regularly changing the weighting they give things based upon emotions
- [26:22] Humans are more likely to use algorithms when they’re allowed to modify an algorithm
- [35:20] Increasing human adherence to using superior algorithms to make predictions
- [40:58] Are there ever good reasons for humans to distrust algorithms?
- [1:04:17] How do we optimize the decision-making for individual decision-makers? And what would Berkeley like to know about how large tech companies get humans to use algorithms?
- [1:11:15] How can people learn more about Berkeley’s research? And what research projects is he currently working on?