Tesla Android— A Future That Will Never Come

One month ago, Tesla dropped a bomb shell: they will have a working android prototype in 2022. The internet’s best and brightest quickly split to either call Elon a raving lunatic, or telling everyone that this is the most brilliant move he has ever made. My take: Neither.

Technical Challenges

Let me explain. All we know at this point is that these androids will be performing tasks that are either dangerous or tedious for humans, will be human shaped, and roughly 5'8 and 125 lbs. We also know that Tesla employs brilliant engineers that have made a very impressive self-driving car* (astrix for you people who are going to mention the current investigation). That’s it.

It’s hard to go off such a lack of information. Will the Tesla Android be reminiscent of Boston Dynamics Atlas, or will it be more mundane and simple like every robot that is build to hand out pamphlets at CES?

The complexity that goes into creating a robot like Atlas is astounding, with Boston Dynamics working for decades on this project. And yet despite decades of research, Atlas still weight 200 lbs, is 4'9 in height, and struggles with performing non-preset tasks. Sure the team has trained him to flip over from a tabletop down unto the ground, but can he clean up after a dog, or run to the grocery store for me? The hardware feat that Tesla would have to overcome build such a compact robot is already impressive, but to catch up in AI to that level of complexity is even more unlikely. I find it highly doubtful that even Elon can achieve this so quickly.

Now lets pretend that Tesla does deliver a robot that meets their specifications, then the obvious next question is why?

Target Audience

Most people can only dream of owning a Tesla car one day, but the robot seems even further out of the hands of the typical consumer. Would people pay $50,000 for such an android? How about $150,000?

Costs aside, there is the big question of practicality. What will the robot do? We already have dedicated robots who vacuum our carpets and wash our dishes, so would the Android interact with them? Will it do our grocery shopping, drive our Ubers, and work in our factories? I’ll agree that grocery shopping may be a potential use case, but the probability of the Android actually being purchased and used for consumer purposes is low.

Most likely the robot, if manufactured, would do well in exploring dangerous environments like nuclear power stations, or perhaps cover several positions on an assembly line. Large Corporations can afford the steep price tag on these Androids, and can hopefully replace humans currently performing dangerous work with Androids that can be replaced without a second thought.

Now if Tesla is to follow through and have a general purpose Android that greets people at stores, hold their bags, and cooks them dinner, then the price will have to come down significantly. The vast majority of the population typically spends $30–40k on a new car which is essential to their lives, so I will go on to wager that most won’t spend more than $10k on an Android that will be more or less useless for the next decade until Tesla develops appropriate software for it.

Taking utility and cost out of the equation, is the proposed human form factor really the best shape for a robot to take? While it’s true that the manmade world is currently built for the human form-factor in mind, let’s be honest, we are not the epitome of efficiency. Walking vertically on two legs creates unnecessary stresses on joints, while our relatively short arms are good for general tasks but not so much for grabbing items off the top shelf of a grocery store. If Tesla is taking this Android project seriously, then they should consider developing a new form factor for these Androids and not try to squeeze everything into a humanoid mold.

Conclusion

On so little information to go by, it’s really difficult to properly comment on Tesla’s proposed Android project. While in theory it may be quite useful in industrial settings (or perhaps to perform some heavy lifting on a faraway planet), it currently does not look like Tesla is seriously considering the development of an Android for everyday consumer use.

Having said that, it is very exciting to see companies start to think about products that they can offer to the public outside of what’s been the status quote over the past two decades (phones, tablets, computers, cars, etc.). Now that the proposal is out there, the clock is ticking for Tesla to work their magic and showcase their prototype by December 31, 2022.