Elon Musk’s physics mindset has aided him in many of his businesses, but might it be blindsiding him about Tesla being able to soon achieve true autonomous self-driving capabilities.
AFP/Getty Images
Leaders often use specific mental models to guide their approach to leadership (see my book on business pivots and leadership for further explanation). Allow me to provide a quick example.
I was brought into an up-and-coming high-tech firm to do some management consulting due to a recent organizational change.
The board of directors had decided to replace the CEO with someone recruited straight out of the military. As a senior officer with a lengthy successful career in the Army, the new CEO proceeded to try and run the company in a military style. This move was quite counter to the firm’s culture and ultimately caused a number of difficulties and issues.
Had the new CEO leveraged his military background and been willing to adapt it to the commercial sector, the result might have been different and possibly a boost to the company. Instead, he got bogged down in what he already knew, serving as a kind of mental anchor, shaping all of his actions as the top leader of the firm.
In that sense, the mental model that guides leadership behavior can be a dual-edged sword. On the one hand, the mental model provides a handy framework for being able to make decisions and take executive actions, doing so with the belief that if it has worked well before, it will certainly work well again.
Unfortunately, it can also be a subtle and sometimes unrealized trap, causing the leader to see the world through the lens of the mental model even when the model might be misapplied or misappropriated to a new situation.
Tesla And The Leadership Of Elon Musk
Shifting to another example of leadership mental models, consider the case of Tesla and what’s taking place there.
Elon Musk is a celebrated leader that has often touted the importance of physics underlying his way of thinking, shaping how and why he takes various actions and approach to problem solving.
You can readily find published quotes of his such as “physics is a good framework for thinking” and “I tend to approach things from a physics framework,” which dovetails into one of his bachelor’s degrees being in physics and his initial pursuit of a PhD in physics (he left after two days, opting to pursue other opportunities).
Let’s consider how this physics mindset has and continues to shape his leadership efforts.
And, in particular, consider too that perhaps his physics mindset might be stoking his hopes and views on soon achieving full self-driving cars at Tesla, of which few in the self-driving car industry believe is likely and express qualms about a stated overreach in a moonshot-like achievement.
Status Of Tesla Toward True Self-Driving Cars
There is a myriad of speculative reasons being given as to why Elon Musk believes so fervently that Tesla is on the cusp of presumably achieving true self-driving cars.
As clarification, most would agree that a true self-driving car is one that can operate fully autonomously, meaning that the AI is driving the car and there is no human driving involved at all, often referred to as a Level 4 or Level 5 on the autonomous car scale.
The Level 2, which is generally the classification for the current Tesla models, even when including today’s AutoPilot, and the Level 3, which is gradually emerging, consist of semi-autonomous capabilities that require a human driver be at the wheel of the car.
Any such co-sharing arrangement requiring a human driver and AI is not properly classified as being autonomous or fully autonomous, and instead is at best semi-autonomous.
According to numerous comments made by Elon Musk, there is an implication that Tesla is on the verge of having a truly autonomous or fully autonomous self-driving capability soon, perhaps by the end of this year or sometime into the first part of next year.
Many have tried to parse his words to figure out if that’s what he is indicating since there is ambiguity in how he has expressed the matter. There are numerous heated and acrimonious debates about what he has stated in his actual words versus what those words imply.
Assuming for the moment that indeed the implication is that fully autonomous car capability is within the grasp of Tesla, it would be a rather remarkable and unexpected miraculous turn of events if that was actually true. There has been sparse publicly presented evidentiary indication by Tesla that they are really so close to achieving true self-driving cars, and the Autonomy Investor Day event in April did little to shore-up such a promise (it was merely Machine Learning 101 kinds of presentations).
Why then does it seem that Elon Musk believes Tesla is nearing true self-driving car realization, perhaps these reasons might apply:
· Because he knows it to be true and indeed it is going to happen as he has predicted it will,
· Because he wants it to happen, though there might be a gap between what he wants and what the firm is actually going to be able to achieve,
· Because he is using marketing parlance to gain attention to Tesla in the desire to sell more Tesla cars and keep the company underway,
· Because as a strident entrepreneur he has beat the odds before and in this case though the odds might be steep he is using the belief as a motivator for Tesla to strive mightily,
· Because his physics mindset is spurring him to underestimate the difficulty of achieving true self-driving cars and causing him to equate the AI of self-driving car to being akin to any other kind of thorny physics problem.
You are welcome to choose any of the aforementioned reasons, and you might even have some additional reasons that come to your mind.
For now, herein let’s take a closer look at the point about his perhaps over-reliance on a physics mindset.
Time to unpack it.
Physics As A Mental Model
Physics is great.
I wanted to start with that declarative statement so that you’ll know that I am not somehow going to argue that physics is sour or untoward.
Not only is physics great, it has pretty much done well for Elon Musk.
Want to do space travel?
If so, physics is a key underpinning. Musk has had rather remarkable success in the realm of space travel.
Want to bore big tunnels underground?
If so, physics is a key foundation. Musk has had some success in the realm of boring tunnels.
Want to capture and harness solar energy?
If so, physics is a vital aspect. Musk has had success in the realm of solar energy.
Want to make a car that uses batteries and is an exciting EV?
If so, physics counts.
For Tesla, there is no denying that their cars are a wonderment of physics and the achievement of EV into the everyday world.
Overall, the theme I’m driving at is that physics has been a boon for Elon Musk in terms of the many endeavors that he has pursued.
As such, the physics mental model has become increasingly strengthened. With each success, the outright virtue of deploying a physics approach gets further and further reinforced.
This takes us to the rub.
The rub is that the AI to achieve a true self-driving car is not seemingly a physics problem per se.
Unfortunately, if that’s indeed the case that physics is not the right tool to solve the AI driving problem, a staunch belief in physics is going to go awry and try to bat away at solving a problem that sits outside the mental model being used.
The mental model could lead to understating the problem and over-inflating the ways or chances or speed at solving the problem.
AI Of Self-Driving Cars Is Not Physics Per Se
What does it mean to suggest that the AI for self-driving cars is not a physics problem?
Currently, nobody really knows how the brain works.
Some would assert that the AI needed to drive a car has to be “equivalent” to human thinking in the sense that the AI has to be able to do the same kind of mental processing that human drivers do. This does not necessarily mean that the human brain has to be recreated into a machine, as it could also be that we are somehow able to simulate how the brain works and achieve the same end-result of being able to “think” in a relevant manner.
Suppose you believe that the brain is a physics problem.
You would possibly be searching to find the equations such as Einstein’s “e equals mc squared” kind of postulation that would universally explain how the brain functions.
You would get the best “physicists” of the brain to crack it and figure out how the darned thing works.
The reality is that today we are a far cry from being anywhere close to discovering the formulas that reveal the inner processing of the human brain. No matter how many scientists and engineers you throw at the problem, we just aren’t there yet.
Therefore, if you envision that achieving true autonomous cars requires being able to solve the problem of how the human brain functions, which could be claimed as essential to putting into AI the driving capability of a human driver, you’ve got to assuage yourself to a long journey ahead.
This is also partially why achieving Level 3 cars is somewhat easier, since you are still reliant upon having a human driver at the wheel, being able to takeover when the automation cannot handle a driving situation. Thus, a Level 3 that maybe can stretch to say 90% of what a human driver can do (this does not yet exist), would still need that human driver in the driver’s seat to provide the remaining 10%.
For a truly autonomous car, the AI must be able to do 100% of what a human driver could do.
It’s a tall order.
Now, I am not saying that the human brain defies the laws of physics. Of course, it does not. There is no question that the brain is a physical composition and abides by the physics of our times.
Yet, let’s be clear, physics has not been able to explain the mysteries of how the mass of brain in our skulls arises to being able to think. Chemistry doesn’t explain it either. Nor biology. Engineering doesn’t explain it. And so on.
Neuroscience is trying to explain it, but we are still only at the primitives of discovery.
Many researchers in many domains are all tackling the enduring secrets of how minds emerge from the matter of our brains.
Conclusion
Some would suggest that Elon Musk’s efforts to reorganize the Tesla AutoPilot team is an indicator of the physics mindset, namely attempting to reshape and replant the best “physicists” (AI developers) in the belief that doing so will solve the human driving AI problem.
Could his rock-solid physics foundation be leading him down a primrose path on solving the AI self-driving car capability problem?
As mentioned earlier, a mental model of a leader can be a dual-edged sword.
It’s like the famous indication that if you have a hammer in-hand then everything appears to be a nail.
A physics mindset might do well in lots of circumstances, but when it encounters a problem that does not seem especially suited to a physics framework, a leader can find themselves trying to use a hammer when the hammer isn’t going to get the results they desire.
We’ll have to wait and see what transpires.
