programming, teams, and the limits of automation

It is always worth pointing out that (as a general rule) human beings are terrible at predicting the future. This isn’t a harsh condemnation so much as recognition of the human condition.

It wouldn’t take much effort to fill up a small book of quotes/proverbs/sayings that all boil down to “it is out of our hands – we can’t guarantee what will happen.” A personal favorite:

“If the good Lord’s willing and the creeks don’t rise”

Jerry Reed

Of course that doesn’t stop folks from making predictions – which is what I’m getting ready to do …

Career Training

Historically the entire concept of “career training” is something that most folks didn’t have to worry about. For most of human history “subsistence farming” has been the “career” for most of humanity.

The entire idea of needing to be specifically trained for a profession probably only goes back a couple hundred years. What often gets called the “oldest profession” didn’t require any “training” at all.

Even what modern folks would call “professionals” – people like doctors and lawyers – didn’t require a great deal of formal schooling/training until the 20th Century.

The reasons “why” this is true becomes a lesson in the development of human civilization – so I’ll just say that the AMOUNT of human knowledge has grown at a great rate due in large part to improved technologies.

Obviously first you need a writing system – then you need materials to write on and create “tomes of knowledge” – then you need a way to reproduce those “tomes”, etc.

So if we did a poll on the “most influential invention in the history of humanity” – the “moveable type printing press” would easily be near the top – metal working, gun powder, domestication of animals, fishing/boat construction would probably make the list – but being able to record and transmit knowledge over time and space (i.e. what books allow) was obviously kind of a big deal.

Of course none of these “inventions” developed independently of the others – that isn’t the point. At some point the accumulation of “knowledge”/”skills”/”expertise” made the concept of “job training” a reality.

Upward mobility/Job satisfaction

MAYBE in an ideal society individuals would be able to choose the work they perform.

Of course there aren’t many “ideal societies” – so I’ll just point out that the worker makes the work “honorable.” How a society tends to reward different professions says a great deal about that society not the workers/profession itself.

In general people can put up with almost any “how” as long as there is a good enough “why.” (I’ll mention/recommend “Man’s Search for Meaning” by Viktor Frankl for the curious).

ANYWAY – for better or worse, most folks don’t “choose” a profession so much as “wander into one.” Which isn’t necessarily a bad thing – e.g. once again, the unexamined life is not worth living. Yes, there is a larger plan being worked out here on earth – but one way or the other:

 Whatsoever thy hand findeth to do, do it with thy might; for there is no work, nor device, nor knowledge, nor wisdom, in the grave, whither thou goest.

Ecclesiastes 9:10

ANYWAY – wrong predictions

I’ve spent a lifetime in “technology” almost by accident. I picked up some “personal computer repair” skills when the ‘personal computer” was a new thing – then moved into “computer networking” partially because it paid more AND an opportunity presented itself.

During that 20 year period I tended to “buy a book” and read up on the latest technology. While I’m “naturally curious” and enjoy learning – most of the time my motivation for learning was a larger paycheck/$$.

There was a point in the early 1990’s when the expectation was that “computer programming” would be outsourced to lower cost workers “overseas.” Which did happen – but then those jobs ended up coming back when the “theoretical cost savings” were consumed by “real world communication issues.”

Now, I don’t think anyone ever officially stood at a podium and said “don’t go into computer programming” – but the trendy concept at the time was “globalization” – and if you wanted a job “safe from globalization/outsourcing” then maybe you should avoid “computer programming” as a career choice.

The end result? Well, there are a LOT of “coding boot camps” out there trying to fill the the need for trained “computer programmers.” Ok, the “Interweb” kinda happened in the same time period – which changed a lot of things, not just the computer programming field …

More predictions

Now (written in 2021) everyone is expecting driverless cars and trucks to make “truck drivers” obsolete.

Maybe. Maybe not. Either way – the immediate need for “truck drivers” isn’t going away anytime soon.

Would I suggest “truck driving” to someone looking for a “lifelong career?” Probably not – but that requires context.

In general I try to stay away from “giving advice” – particularly “career advice.” The problem is that there are just too many variables – e.g. “should someone pursue THIS field or THAT field?” – I have no idea, it depends on the person.

I tend to say “know yourself” in those situations. HOWEVER – just for fun – “should someone become a truck driver in 2021?”

Well, relatively short training time, relatively high wages, in demand skill set that isn’t likely to go away anytime soon – might make the job of “truck driver” a good choice for a large number of people.

If we divide the “supply line” into “long haul”, “medium routes”, and “the last mile” – the sweet spot for HUMAN truck drivers (as in not easily replaceable by automation or crowdsourcing type apps) is the “medium routes.”

e.g. a “shipment” might come in on a big boat in a large container, get unloaded and placed on a train where “X” miles and then gets “dispatched” to a truck that takes it to a local warehouse, where it gets unloaded and (maybe) delivered to “customers” by someone using “crowdsource delivery app.”

Automation isn’t free …

The problem with trying to automate any process is that then the “automation” needs to be maintained.

Which means the “processes” best suited for “automation” have a few simple steps.

e.g. if the process is “place a box (that is always the same size) in the same spot (that is also always the same size)” – automate away. BUT if the process is “make hundreds of smaller decisions along the way to an uncertain destination” then automating (and maintaining) the process will probably be extremely expensive and error prone

… which means that training human drivers is gonna be cheaper than automation for the “foreseeable future.”

Yes, the technology for “driverless cars” is available in 2021 – the problem is the variability of the deliveries and routes and the cost of constantly updating/maintaining those routes.

e.g. GPS is great – but if it goes down, then what? Well, the automated system probably completely ceases to function – while the “human driver” might be unaffected or just slowed down.

I know the “automated workforce” always sounds great to “upper management” types – and I’m not saying that drones/driverless cars aren’t going to change/improve the supply chain EVENTUALLY. Just not in the next couple years …

Teams

Imagine that you have “human like devices” that can be trained to play “sports.” Which sport would be easiest to “automate?”

Well, we would want the sport that has the least amount of interaction between “devices.” Maybe an individual sport like golf? then maybe something like tennis?

When the “devices” have to work together as a team – then things get complicated. So then the sports with the fewer “players” would be easier to automate – basketball? hockey? Both fast paced, hockey (usually) has a dedicated goalie – I’m not sure if that would make it easier or harder to automate – not important at the moment.

Both hockey and basketball lack unique “offense” and “defense” players – sure, some players are expected to perform different functions but in theory all of the players are on both “offense” AND “defense”

Baseball has a kinda complex set of 9 players on defense that we might be able to subdivide into “outfielders”, “infielders”, “pitchers”, and then “catchers.” Maybe the “hitter” and “baserunner” functionality can be 1 generic “offensive player.” still MORE complex in theory (harder to automate) than either basketball or hockey.

What gets called “soccer” in the U.S. has 11 players on each side (including a goalie) – but from a “theoretical automation” point of view (POV), those 11 players are NOT supposed to run into each other. Coordinating passing of the ball would be a challenge – but from a “team complexity” POV probably somewhere between basketball/hockey and baseball.

(remember I’m not rating the “sports” – I’m estimating the complexity of automating)

Then we get to “football” – or more precisely “American football.” We have 11 distinct players on offense AND defense with “collisions” between players/devices on every play. Maybe the players on defense don’t require as much “unified precision” as those on offense but a very high level of “interaction” between all of the devices.

Obviously the devices on “defense” are working together to try to stop the devices on “offense” which are also working together in a coordinated manner … and don’t forget the “kicking” game … ANYWAY

SO “in theory” American football would be the hardest sport to “automate” because it requires the most “teamwork.”

Again, this is about one sport being “superior” to another. In the real world this difference in “required team coordination” is seen in pro-sports on a regular basis.

Every once in a while some wealthy owner will try to “buy” a championship by spending a lot of money acquiring great individual players.

This seems to happen in the NBA on a regular basis.

The NY Yankees have been doing it in MLB since 1923.

Hockey has had great NHL franchises win a lot of championships in a row – but I don’t know enough about the history of the sport to comment on ownership $$ spend. The Montreal Canadiens have won 24 Stanley Cups – but I don’t know if I would compare them to the “NY Yankees” from a player acquisition POV …

I’m told that the successful “European football” clubs tend to be the same each year because of $$ spent on players – again, I don’t know the sport well enough to make comparisons.

HOWEVER – anytime a franchise is extremely successful over a long period of time – they are doing something “right” besides spending a lot of money. I certainly don’t want to imply that just because the owners spend a lot of money, AND win championships that they are “cheating” in some form – and moving on …

… and then we have the NFL. Forbes estimates the “team value” of the Dallas Cowboys (in 2021) at $5.7 billion. The Cowboys are always near the top of the list – and last I checked EVERY NFL franchise had a $billion+ valuation.

Then as I was writing this an article on the Tampa Bay Buccaneers ownership came up – they guaranteed Tom Brady $50 million to come to Tampa Bay. Does that constitute “buying a championship?”

Ok, obviously you need great players to win a Super Bowl – and those players are gonna cost $$ – but then those players need to work together as a team to win. You can’t just go out and spend a lot of money on free agents and expect them to automatically win championships.

I’m guessing that Tom Brady saw a potential championship in Tampa Bay, or he would have gone somewhere else.

The Cowboys and the “whatever we call the franchise in Washington DC” (last Super Bowl win 1992) both have owners that are willing to spend a massive amount of cash to win a championship – and are still searching for that right combination of talent and then something that money can’t buy “team chemistry.”

ANYWAY – What makes football fun to watch is that the team with the “best players” doesn’t always win –

As always – I’m just making observations – if I actually knew the secret ingredient to “winning” in the NFL/pro sports, I’d be making a lot more money 😉


Posted

in

, , ,

by

Tags: