Automation we will live to see

// Over the last few months I’ve been back on a plane quite a bit. Its given me time to do some reading. Specifically, Cybernation, Who owns the future, and But, What if We’re Wrong. //

You’d be right to expect this has something to do with financial services, but doesn’t! A lot of the discussions floating around on this topic have very little to do with FinTech.

Most of the discussed automation and where it’s expected to occur is fairly obvious. In financial services, the automation is focused on taking processes that involve paper+people, where people create a compounding error. The automation reduces that cost center by reducing the number of errors and through that automation is valuable.

An easy example where a company like Dwolla can help customers using automation is in the ACH returns and corrections process. An analyst reviewing lines in a batch file and clicking buttons doesn’t really help a business operate more efficiently and it introduces a number of things that could go wrong. Automating the ACH return process through an API saves everyone time.

By and large, automation has allowed financial businesses to reinvest in more people where they can generate greater value for the company. Handing paper between two people is not typically*1 a high value task.

Back office automation isn’t really obvious to individuals in organizations who don’t have access to, or read, the P&L. It might show up as increased margins for the accounting folks and leadership, but otherwise its success fades into the background. It also seems to be the focus of angst for those holding jobs that could be displaced by future automation, as well as those among us who believe they are important/creative/human enough not to be replaced by an automated task.

The fact that the automation fades into the background makes it incredibly easy for it to be forgotten or appreciated. When done correctly, automation silently makes everyone more money and reduces errors while increasing transparency. So great!

There are other applications for automation, not AI*2, that I think are far more in our face than any automated task that benefits FinTech companies, that are just over the horizon. I do think they will all effect us in our lifetimes in a substantial way.

There’s lots of talk of bots and AI but in the near-term and I can’t help but feel strongly that automation is the more important thing to focus on, not AI. I say that because automation, and even self-driving cars, don’t really have the ability to self-decide. These technologies take actions based on information given to them by their programming. The information is provided by humans who told the program what the decisions were to the problems in front of them. Automation, is really fast application of supplied rules. AI, is self-awareness and self-creation of ideas. A consciousness, so to speak.

Automation is here and it’s really amazing. AI, is something that I don’t really think is here but will be at some point. Much like quantum computing, there’s a lot of talk but not a lot of reality, yet.

Computers don’t have consciousness yet, but we do have automation coming out our ears. The three areas my brain keeps coming back to are Trucking, IOT in big manufacturing, and Farming.

Effectively, how our stuff gets from to A to B, how our stuff is produced, and how we’re fed. All things that scale that will or already do effect virtually everyone in some way.

Trucking

Trucking employs an astounding number of people in America and lately seems to have been getting a lot of attention because employee turnover is high and efficiencies through automation seem obvious. There are entrepreneurs and VCs trying to solve this problem because this industry is in need of solutions.

If trucks only have to stop to recharge and otherwise don’t sleep, the cost of the operation goes down dramatically as does the human error. It’s not to say that humans aren’t good drivers but humans do tend to get tired, make phone calls, or maybe drift on in thought while trying to spot all 50 state license plates on a long road trip…

The roots of this transformation probably are in consumer vehicles and as we’ve seen with Tesla even slight abuses can slow down automation (self-driving) initiatives. The social comfort with these technologies is falling far behind the technical innovation and that’s perfectly fine. Other countries seem to be laying the groundwork for us to learn from.

A completely unsolved problem as far as I’m aware of, is actually the number of sensors and sheer number of variables dealing with a vehicle this large, moving through town to town. The computing would almost certainly have to be done on the truck and not remotely because of the crazy amount of data that would have to be consumed by the vehicle’s computers and the speed at which the truck would have to react. Internet latency or availability isn’t really something a truck carrying freight can wait for.

A self driving 18-wheeler essentially needs to be a supercomputer that can be trusted by humans not to kill everyone around it. It prompts crazy ideas of a chase scene where the truck has gone awry and just crashes through us mortals because the truck is trying to make a deadline to get a load of rice to Texas by 9AM. Unfortunately, as we’ve seen lately we can’t always trust humans to make good decisions behind the wheel of a truck either, so maybe automation here isn’t something to be scared of.

I don’t think the technology sector has cracked this but I’m willing to bet someone will in my lifetime will.

Farming

Iowa has been home to self-driving vehicles for quite some time. John Deere has been building self-driving combines and it’s nothing new. I’m surprised sometimes when the media talks about the upheaval in farming, as if all of the farmers are being crushed by it. Modern farming is nothing new but knowledge of it may be.

I’m not a farmer but I’d certainly love to hear from those that are. It feels as though reducing errors, increasing yield through better data, and having predictable machines that you can operate is likely good for business and for safety.

My car can tell the dealership something is wrong with it and an appointment can be scheduled. Imagine all of the farming accidents that won’t happen in the future because no one will pick up a wrench to fix a part that’s too dangerous for them to be touching.

At some point, hopefully farmers won’t have to be in the machines tending to the land and can instead focus on the health of the soil, irrigation, distribution, and other efficiencies that hopefully result in better crops (things we eat and make us all healthier).

Automation in farming*3 has a lot less to do with a machine doing the work a person could do and a lot more to do with getting the farmer in a position to focus on what a machine can’t.

If technology is able to put food producers in a position to spend more time thinking and being critical about the chemicals and processes to produce what we consume rather than just exhaustingly producing output, I can’t help but think everyone will benefit.

We don’t treat engineers by the number of lines of code they write and I imagine the world will not judge farmers by the sheer volume of what they produce, in my lifetime.

IOT in big manufacturing

Large machines need to interoperate for manufacturing plants to run correctly. They have to interact with one another and the less a person takes hold of the parts the safer it is, and likely the tighter together tolerances can be. We’re already starting to see this take shape.

The term internet of things is constantly applied to consumer products where I wonder if security more than the devices themselves are the great opportunity.

The same could be said for interconnected devices that produce hundreds of parts per hour that are far too complex or heavy for a human hand to create quickly or at all. Imagine if the machines just start making parts or products of this size and scale incorrectly. What if a variation in the electrical supply caused the press to produce a variation/change between parts produced and those bad parts got into a car and someone died as a result?

What if the machine was tampered with and the code that was supposed to shut down if the machine didn’t get adequate voltage, was hacked and someone did this intentionally? Unfortunately, that’s a very real concern and the complexity of this is something manufactures have to deal with. The automation and machine learning to detect anomalies and alert the relevant technology (or humans) has to exist so the right things come out the other side of a production line.

Changes caused by input variables of any type, which bear in mind are the things that define the output and can not always be controlled but must be measured, can drastically change the output of what eventually ends up in our cars, homes, food and electronics.

There is an unbelievably large business opportunity in the automation around these machines and processes. Protecting these devices from intrusion and analyzing requests between the machines for integrity feels like a place where automation can have profound impacts.

I found myself wondering if there is room here for proof of work products designed to run on the machines themselves to confirm each request is legitimate. If the messages passed between machines is known to come from the machine sending the information and can only be consumed by the machine intended to receive it, massive changes will follow.

Ramifications and Concerns

Some of these industries are going to displace large numbers of people and that’s something I’m hopeful the federal and local governments are thinking about — how taxation can play a role in the shift between industrial changes and the re-education of the workforce. As a part of the workforce myself, I think it’s something we need to pre-plan for as opposed to try to deal with it as it happens and rely on post-change regulation to force things into normality as a committee.

There is some incredible thinking about this, and Basic Income Guarantees, that are worth reading if this interests you. At the moment, I tend to agree that some level of basic income guarantee should be available at some point with baseline requirements as a part of education into new job roles.

As a society, we should be perfectly well equipped to figure out how long re-education takes, how much life costs, where re-education should take place geographically and skill.

As a society, my concern is that while we have the data and likely the resources to solve the problem, we get caught up in the politics of squabbling and position over details that ultimately don’t matter.

In closing

I shared with my wife last week that I don’t know if our kids will be allowed to drive at some point in their life because it’s so dangerous to drive a car. (For the record, she disagreed.) It’s also the same argument I used in my head when getting a bigger engine in the last car we bought. If gas engines are going away, I might as well enjoy what a gas engine can give me in the meantime so I have the story to tell my kids about the roar of the engine that wasn’t produced by a speaker.

I’m excited about the innovations in automation. I’m not sure quite yet about AI. All of this should also probably be prefaced by the fact that I am neither an expert or participant as a producer in any of these fields. This is me just thinking out loud as a curious bystander. If you have thoughts and feedback, I’d love to hear it.

I’ve also started to think more about how to ask my parents about the changes they saw thematically in the world as an indicator of what’s next as opposed to asking my kids what apps they like. That’s probably because the only answer I get right now at home is Pokemon.

*1 — There are times when a piece of paper between two people is the only time a piece of information exists in the world. It’s an easy argument to make that in this case, passing paper is indeed a very high value task.
 *2 — AI, Artificial Intelligence, a digital consciousness that we don’t control or teach rules to. Something that has its own ideas and views.
 *3 — Ironically, a rise in Ag automation may give way to a market that doesn’t exist today. Food grown, cared for, and produced entirely by human hands. I’d probably give a restaurant that pitched something like that a try but equally as ironic as I’m looking at the noodles I’m eating as I type this, I have no idea where they came from nor would I have thought to ask.


Originally published at benmilne.com on August 6, 2016.