Jeff Bezos stepped down as Amazon CEO yesterday.

While at Amazon, he is said to have relied upon these four decision-making principles.

1. Most decisions fall into one of two categories: irreversible decisions and reversible decisions (for the latter, it's better to act quickly).
2. Make a decision once you have 70% of the info you need (...because if you wait until 90%, it’ll be too late).
3. Ask colleagues to “disagree and commit” in situations where no one (really) knows the answer, but you have a strong conviction that your idea is the best contender. (You'll ultimately get the answer(s) you need quicker this way.)
4. Escalate misalignment issues as soon as you discover them to save time. (Otherwise, you'll waste too much time.)

To my mind, I think adopting this type of thinking is useful. But I think I’d add that there’s a third category of decisions that people at organisations typically make: recurring transactional/operational decisions, which demand consistency and are irreversible. Some examples of what I mean, that might resonate with people in this community, are: should we onboard this new customer? Is this meeting in the office safe to go ahead? Or...is a certain transaction (indeed) fraudulent? (These are obviously often the best candidates for automation.)

Does anyone here use a similar (or different) set of principles for decision making? I'd be very interested to know! 

In its recently published Global AI Adoption Index 2021, IBM found that:

• 43% of firms have increased the speed of their AI adoption since the start of the pandemic
• 74% of firms have either deployed or are currently exploring the deployment of AI

But most interesting for me was the finding that: 

• "91 percent of businesses using AI say their ability to explain how it arrived at a decision is critical"

I'm keen to hear if any of these (or any other stats in the report) resonate with you!

A paper about decision fatigue was featured recently in the Economist. Its results were fascinating (and also a little terrifying). 

The paper—written by Cambridge University researchers Tobias Baer and Simone Schnall and available online here—supports the hypothesis that the quality of human decisions decreases when we become tired. The study found that, in the case of 26,501 complex restructure loan applications made over the course of one month at a major bank, a significant dip in the approval of loans was visible around midday (i.e. before the employees took lunch).

But that's not the most mind-blowing part. 

The researchers then went on to show, using additional credit collection data from the same bank, that had there been no dip the bank might have made an additional $509,023 in revenue. 

Quantifying decision fatigue like this, as the researchers have, is pretty remarkable. But it does worry you reading it, though. And not just about the cost of fatigue in the financial factor and how that impacts society, but also about cases in sectors where the result of decision fatigue is not realised (at least directly) in financial terms—for instance, patient outcomes in health care. 

Anyway, I highly recommend giving it a read: https://royalsocietypublishing.org/doi/10.1098/rsos.201059

There's also a shorter Cambridge University summary here.

And here's the Economist article (£).