Uber is like your brain : “Inefficient” for a reason

I was recently reading an article on the 10 percent myth of the brain and how each neuron fires 200 times a second on average. One particular section stood out:

But, as LeDoux put it to me in an email, “the brain could be one hundred percent active during a task with only a small percentage of brain activity unique to the task.”

It got me thinking on the concept of “efficiency”. We define efficiency as the ratio of output to input for the particular task, and the brain is particularly “inefficient”. Most neurons tend to fire “meaninglessly” because there is a strong possibility that one of these neurons firing will trigger a domino reaction that results in “useful” output or end goal (a good thought, maybe?).

The end result of a neuron firing is not known when it is fired, but due to the nature of the brain being a network, it benefits the system as a whole in the long run. You are able to think, respond, behave and perform other such activities because a lot of background activity is happening in your brain - and it is not always necessary that this activity results in the actual task at hand.

What does this have to do with Uber?

I took this little thought experiment further and one neural activity led to another which made me think of Uber because it is essentially a massive network. Uber is a classic example of two sided network effects, which is explained very clearly in this short post.

Uber continues to add more drivers and users to its already massive network, and it has always stressed on the fact that it wants to get you transportation at “the touch of a button”. Uber and the on demand startups in general have been built on the fact that they increase asset utilization.

My instinct on the “efficiency” (or utilization) of the network being low in absolute terms has been beautifully worked out to show that the average Uber cab utilization is close to 31% - lower than that of the yellow cab taxis. This is close to my anecdotal understanding of the average taxi driver doing 10 trips a day. On an average cab trip of 45 minutes it works out to a ~40% utilization of the car for the day. I can’t comment on whether Uber has increased the utilization of these assets from where it was before, or decreased it - but the number isn’t close to 100%.

Is it in the network’s benefit to keep things “inefficient”?

Uber manages the network and subsists on commissions it generates from every trip. Just like the brain has an end goal, this network has the end goal of being able to provide you transportation ASAP.

Due to its commission, Uber is incentivized to convert any possible “deal” (a potential passenger trip) and this means that there should always be the other party (the driver) available to make this “deal” happen. It is therefore the best case scenario to flood the market with cabs - so that no “deal” is missed at all - great examples of “inefficiency”.

Just like your neurons are always firing dopamine with the hope that something would happen, drivers remain active in this network in a similar manner but over a longer period of time - and they are incentivized to do so. A driver moving around is like dopamine in the system - with the neurons (Uber nodes) firing with the hope of conversion. Sometimes this random firing clicks. We can probably draw many parallels - I see some fascinating possibilities. This “inefficiency” allows for near perfect on demand service - the passenger is always served.

But is this really an “inefficient” activity or something done that actually makes the broader scheme of things work?

Uber is not going to drive the driver to 100% utilization

… because it’s gunning for the passenger always being served. Just like we’re the fortunate recipients of the brain always firing and are able to function, passengers will always be the recipients of a driver being available and be transported.

The network will possibly discover some optimal utilization (that will be low) which has presently been driven down (or up?) by the massive amount of money Uber has raised. The market is definitely artificial and it will remain to be seen how it evolves over time.

Does this mean that Uber is truly inefficient for drivers? Or the brain is for the neurons? I do not know. What I do know that both are going to be very active to achieve the respective end goal of each network.

As an end note, this also brings me to a related point on how macro measures (such as the GDP) are not capturing true value. Efficiency, as measured by the mechanical input and output may not be capturing the true value add in each step in the case of complicated large scale networks. How each step ends up functioning and contributing to the network is a fairly complex task and definitely adds value in one way or the other. We just need to figure out how.

I blog about venture capital in India at Life of a Junior VC