The machines are coming – 2049?

It is notoriously difficult to predict the future. I lived my life at IBM, following Alan Kays 1982 aphorism

The best way to predict the future is to invent it

In my career I got many things right, and many things wrong. While Amazon was still a small time bookseller, and Youtube for the most part didn’t exist, it was obvious both business models would thrive. While I couldn’t convince IBM to pursue either of these opportunities with ny success, we demonstrated the technology perfectly. My “Wired for Life
Presentation contains some of my wins, and many of my losses.

It was much easier to build on these, especially the societal impact in my 2003 “Trends and Directions” presentation.  Societal impact is much easier to predict as you can demographic data, current trends and it’s pretty easy to extrapolate. Technology adoption is much harder.

Many of these predictions are not useful, after all who needs  a robot to write high school essays? Many though will continue to fundementally change work and life as we know it.

What they are though is a signal in the way the World Economic Forum predicts the technology will develop, and to some degree it’s a self fulfilling prophecy. Watching this and reading many of the “machines are coming” articles that have been published over the past 5-years, it’s easy to become depressed about the rise of automation, AI, and robots. In a year when the sequel for Blade Runner will finally appear on our screens, there are some key things to remember.

  1. There is no magic, no silver bullet – If they can’t explain it, or worse don’t understand it, they have not invented it. Machine learning is great, but the machines can only learn with the machine learning constraints they have.
  2. Listen to the doubters – Doubt is very different to dismissal. People who dismiss possibility out-of-hand either don’t understand the opportunity and the potential, or are afraid of the change. It’s the doubters who have thought things through and understand the problems and the weaknesses.
  3. Don’t fear automation – If you do, you will be left behind. Learn, adapt, change; if possible work to invent the future By all means be a doubter, don’t be a dismisser.
  4. Find a problem, don’t start with a solution – AI, Robotics, Big Data, Machine Learning, Algorithms, Neural Networks are all speciality fields, grabbing onto them and asking how can we use them isn’t useful. The more specific you can be about a problem that needs solving, the easier it will be to identify the correct technology.
  5. Be Human – the more we automate, the more important human interaction becomes. The more empathy you have for someone who has a problem, the more likely you are to be able to understand how to solve it. Empathy, the arts, sports and human interaction are all fields where robotics and automation are least likely to take over.

More Human than Human – Dr Eldon Tyrell, The Tyrell Corporation

Giant fleet of small scheduling nightmares

In tenuous link between my recent posts on automation, here and especially here, where back in November I discussed autonomous vehicles and their impact on employment. I also said:

While many cities are salivating over the ability of self-driving, autonomous vehicles to fix their broken road and transport infrastructure, that’s missing the point.

Sometime between then and last weekend I became a weekend subscriber to the (Boulder) Daily Camera, a great local paper for the Boulder/Denver metro area. Right on queue, my first Sunday paper was laying in the snow on the drive this weekend and I opened it up and parsed it during the day. One item that particularly caught my eye was Dave Kriegers main editorial entitled “Imagine a giant fleet of tiny buses“.

serveimageI grabbed a pen, marked the editorial up, scribbled in the margins and sat down on Monday morning and wrote an open forum letter. It didn’t get published, I have no idea if it’s policy not to publish corrections on staff written op-ed pieces, or they just didn’t think it interesting enough to include?

Since a big fuss has blown up about an Uber self driving car today, running a red light yesterday(in full transparency, Uber self driving development has a big office here in Louisville that is a build-up from the Uber acquisition of Microsofts Bing mapping service.) I thought I’d turn Mondays open forum letter into a blog post.

This also lets me correct one misstatement. Self-driving cars will help with congestion theoretically. In heavy traffic, they will drive at a regulation speed, a safe distance from the vehicle in front, thus avoid the hard braking and the impact that can have on several miles of traffic.

It is hard to respond to Dave Kriegers editorial imaging “a giant fleet of tiny buses” in 300-words, but I’d like to have a try.

First, I completely agree with his sentiment that if you keep trying the same old thing, you’ll keep failing. However, when it comes to his “giant fleet of small buses” he falls into the same trap most transport ‘imagineers’ do when the come to self-driving vehicles. For the sake of brevity, let’s assume they’ll be electric; let’s assume they can dock themselves; let’s assume they have a slightly better range than current electric cars.

Dave jumps to the conclusion that less space will be needed for parking. Sort of, except the cars have to be charged somewhere. But yes, they could be charged in either fields or reclaimed parking garages outfitted with self-docking chargers. Dave then makes the confusing jump to the conclusion that “[they] could reduce congestion because fewer cars would serve more people”.

Anyone that’s given any serious thought to scheduling and transportation would understand implicitly that that isn’t true. It’s implied because it fits the paradigm of autonomous vehicles. If 20,000 people want to get into Boulder today between 7:15am and 9:00am in their own unshared transportation, and the demand is the same in the era of self-driving cars, then, you’ll have the same number of journeys. Add in the recharging trips, the fact that using Daves logic, there will be less self-driving cars, then some of those cars will have to drive in and out and back into Boulder, actually increasing the number of journeys and therefore contributing to the congestion.

If we take Daves “less parking space” claim at face value, then what will the space formerly used by parking garages be used for? Green space… err no, more offices/accommodation, with the potential to further increase the number of journeys and congestion.

Don’t get me wrong, self-driving cars are great, but until we have flying cars they will only help indirectly with congestion won’t help with congestion. The only way is shared transport. Bus Rapid Transport isn’t it either. Trams, street cars, metro-rail are the only real fix.

Creeping automation

Automation is everywhere, but most of us don’t notice it. Every product we buy, every service we use has been touched by automation, some more than others. Think about the products you buy about the grocery store? Come in a package? Packed by machine!

I’ve had some interesting emails from regular followers/readers about automation. I don’t think people quite understand how invasive and creeping automation is.

Here is a perfectly simple example. I live on a new development in Colorado. Some 70 single family homes, and now they are moving into the multi-family condo and town homes. They’ve built two condo buildings, and a 3rd 12-plex is going up now, literally right across the street.

Pressed Siding, Floor and Ceiling Boards from Canada
Crane lifts pre-constructed parts into place

First, let’s be clear, this isn’t a pre-fabricated building. It’s a unique design to this location, that most would consider “traditional construction”. Only it isn’t, it’s massively labor free, and largely “skilled-labor” free.

The construction is typical timber frame, in the old days, there would have been an army of craftsmen working on site, the continuous sound of saws and hammers. While there are some professional craftsmen on site, the bulk of the construction is being done by “nail gun jockeys” using pre-cut, assembled panels and components. They just nailed them into place.

Of course, walking pass this, you’d never normally take a second look. Since it’s directly opposite my home office, everytime I look out the window I see it. What do I see?

Creeping factory assembly and automation. All the major parts arrive pre-cut to size; the joists and all the boards for the side that need holes for windows and door arrive pre-cut. The original boards come from Canada and Mexico. Anyone who has seen the home improvement shows knows that the boards are not cut by craftsmen and craftswomen, but simply cut by operator assisted laser cutting machines. What do all these things have in common? Automation.

It would be easy to simple, let’s at least bring back the board creation, prep and the component assembly to the USA. Indeed easy to say. That assumes we have the raw material, and that the factories exist that can manage the increased workload. If they can’t then let’s assume they can be built.

What happens then? Well, the businesses that do the manufacturing either produce the same goods at the same price as they are available overseas, which will be hard. The US no longer has the same wealth of natural resources. Those that we do have are harder to extract, or come with environmental, planning or development restrictions. Even if these were lifted, they would still come with a price tag.

Those costs, plus any for plant construction, or increased raw material cost would be passed onto us. Effectively doing a “Carrier” and raising prices to cover their increased costs.

Automation is everywhere, but most of us don’t notice it. Every product we buy, every service we use has been touched by automation, some more than others. Think about the products you buy about the grocery store? Come in a package? Packed by machine! Ready made meals, the whole production line from animal slaughter to food prep and cooking are all now largely automated. It’s invisible, invasive and all encompassing.

Bring back what jobs?

Look, the jobs that are lost, are not coming back, get over it. When Trump claims he’ll bring back jobs, he either has no idea what he is talking about, or he envisions some dystopian future where Americans are more like slaves than they’ve been since, well, slaves.

China and Mexico are not the problem, automation is. Even if Trump were able to force companies to bring manufacturing back to ‘Merica, through punitive tax and trade barriers, the manufacturing won’t be the same as it was, ever.

Listen to this recent extract from NPR’s All Things Considered. Bertram de Souza of The Vindicator talks about steel mills following a recent visit of Trump to Youngstown Ohio.

The next wave of automation is coming, it’s in driverless vehicles, it will have a dramatic impact on employment. Automated delivery trucks, automated taxi’s, autonomous vehicles will make a large dent in the current employment of some 3-million in America. While many cities are salivating over the ability of self-driving, autonomous vehicles to fix their broken road and transport infrastructure, that’s missing the point.

There are many detailed, and complex reports on the impact of automation, pick your favorite organization and search their website, McKinsey and Company(2014); Stanford Business School(2015); Pew Internet(2016) and on, and on.

Equally there have been a few superficial recent reports in the news media, this one from US Today. One of the better, more recent articles is from Rex Nutting over on Marketwatch.Rex Nutting over on Marketwatch.

I’ve been horrified by the lack of actual policy discussion and examination of the context, detail and and lack of clarity even where there is policy. This is something we should have had a real debate about when, what and how we handle the future of automation.

It’s not as if the impact of automation is new. Depending on how you classify automation, it’s been going on since the invention of mills, but importantly since the computer became pervasive in business.

As far back as the late 1960’s it was a discussion topic. In the early and mid-1980’s automation had become a key issue for governments and businesses. This was a classic of it’s time.

A human teller can handle up to 200 transactions a day, works 30 hours a week, gets a salary anywhere from $8,000 to $20,000 a year plus fringe benefits, gets coffee breaks, a vacation and sick time… In contrast, an automated teller can handle 2,000 transactions a day, works 168 hours a week, costs about $22,000 a year to run, and doesn’t take coffee breaks or vacations. – Bennett, 1983

This is a well used quote from a report called “Bank Systems and Equipment” by Bennett et al 1983 and often misquoted and attributed to Nobel Prize winner, Wassily Leontief and  Faye Duchin, who used it in their seminal 1986 work “The Future Impact of Automation on Workers“.

img_20161107_125230I worked on an IBM Corporate study in 1998, following the release of Dunkerleys book, the jobless economy.  I still have the books on my home office book shelf.

Many of us were uncomfortable with what technology was capable of doing to our society, much more than our jobs. I’d seen it first hand and contributed to the loss of hundreds of jobs. When I first arrived at Chemical Bank in New York city in 1983, there were hundreds of people, mostly women, sat in large rooms, processing incoming credit card authorization phone calls. Within 3-years, they were all gone. Their positions had been eliminated. Replaced by simple automation of the repetitive tasks they did using search and a “database” lookup.

Some of the information and outlook from that IBM study found it’s way into this presentation I gave at meetings and conferences around the world at that time.

Automation was, and is unstoppable without a much bigger debate. Trump alone can’t fix it or stop it. Automation is a result of three, equally powerful trends.

One. The absolute fear and revulsion in America of Unions, their impact, power and influence. Sterns 1963 paper “Automation-End or a New Day in Unionism?” captured the potential impact of automation on Unions.

Two. Big corporations and the way the market values them, their ability to balance investment against revenue and more importantly profit. Investors and the market don’t care how business makes profit, and the tax authorities allow investments to be written off against profit. So removing expense, in the form of employees, and improving profits is always on the agenda.

Three. The continual consumer march towards ever more consumption and disposable, cheap goods. Perhaps more than the loss of jobs, if pernicious tax and trade barriers were implemented by any politician or President, we would see a revolt among the people, who more than anytime in history, want their stuff as a measure of their value.

So, we can’t stop automation, the jobs are not coming back. Where does that leave us?

I’m inclined to agree with Musk. The only way around the impact of automation is a universal basic income. That’s what we should have been debating this election cycle. Not fucking emails, walls, muslims and pussygate, let alone if somewhat left leaning Bernie Sanders proposals were socialism by the back door. Without serious discussion on these difficult topics, America will continue to into social conflict and fear.

Even if Trump gets elected today, those 5-million jobs we’ve already lost, and another 5-million are not coming back.

Will self driving cars save Austin from itself?

If you had 100 destinations you’d have removed some of the major bottlenecks, but we don’t. We have the Central Business District. It’s a major constraint and getting people in cars in and out of it no matter how automated the cars are will have the same fundamental problems and constraints. Anyone who says otherwise really doesn’t understand the problem.

There has been a lot written recently about (semi) Autonomous Vehicles (SAVs) aka Self driving cars. Especially yesterday following the Tesla announcement, see also this NPR report.

Self driving cars have long been held out as a solution. They arrive when you need them, they take you to your destination at regulated speeds, they can adjust to traffic congestion, and collisions, road work etc. Once you’ve arrived, they disappear not requiring a parking space either in a building, or on the street. Nirvana.

Proposing self driving cars as a solution to traffic congestion, where the congestion is caused by constraints is simply a nonsense. They are a first world solution, to a first world problem. “Why can’t I text/read/sleep while I’m stuck in traffic?”

There has long been discussion among the urban transportation advocates, while they promise to reduce accidents and better manage traffic density and improve capacity through stable and reduced “headway” (the distance between cars). Anyone who proposes they can solve traffic congestion without understanding the constraints and capacity issues, just doesn’t understand the challenge. This Forbes article screams “Self-Driving Cars Would Slash Traffic, End Street Parking” yet aside from the obvious nod to headway and predictability doesn’t address the problem. This Qoura discussion covers many of the points

Here is a response I wrote yesterday on the issue to the #ATXRail mailing list. It refers to an earlier post that discussed in some technical detail the problems of scheduling and capacity.

Again, I find myself being the naysayer, Roger Cauvin made some extremely salient and possibly too technical points about any transit system design, but especially as it relates to individual transit options such as autonomous cars. The problem that most simply don’t understand is the scheduling and availability of these cars at scale.

First, whats scale? How many would be needed to make a real difference? Scale certainly isn’t a hundred, it’s not a thousand, it’s probably 10,000 maybe more. Second, lets assume these are free movement vehicles, they don’t run on track or are constrained by guide rails. Third, lets assume that they are capable of transporting 2-3 max. but the assumed capacity is one passenger. Fourth, lets also assume these are not individually owned vehicles, or at least if they are, they are available ala Uber/Lyft for other people to book. Fifth, lets assume they are electric and capable of driving themselves to charging stations where they either connect or are connected to a recharge point.

Now we have the basics of your capacity based system and you can start working on its constraints. What you don’t have is any real clue about the usage patterns, how they’ll be used, who by, for what, at what time and so on. Until you can produce even a first pass for this you cannot make any assumption that autonomous vehicles can solve anything except perhaps more consistent traffic flow, and improved emissions. Anyone who claims autonomous cars can solve anything without this data is simply blowing smoke and must be pushed back on.

They are great for private vehicle journeys. I have a Mercedes that will already follow a GPS journey, has variable cruise control that will slow down and speed up the car based on the MAX speed of the vehicle a set distance in front and will bring the car to a stop based on either breaking vehicles or an obstruction in front; it pretty much does everything except steer, but it vibrates the steering column to let you know when to turn and when you’ve are moving out of a lane etc. This type of vehicle which we’ll see for all new vehicles in the next 5-7 years, if not before, will much better regulate driving.

They’ll minimize the concertina effect of drivers speeding up, jumping lanes and breaking too hard which will make congested roads flow much more smoothly. In practice they can also stop people jumping traffic lights, travelling faster than the legal speed limit and other traffic law infringements.

This begs question will drivers allow this to happen, and can insurance companies be convinced to drive the adoption of it?

So assuming all that gets done, autonomous vehicles will become totally acceptable and usable. And then we’ll have the scheduling issues that Roger alluded to. They are real.

Having self driving parks valet park themselves is cool, the question is where? At scale, say 5,000, you need at least say 2,000 spaces, and over night, probably 4,000. Lets those spaces also need to be charging points. Someone has to invest in building and powering those points. The “free” market will(really?). Then all you have to go is schedule cars between where they are, the users, the users destinations, and the parking spaces. When the primary destination is still downtown Austin, you have some massive constraints, not magical relief.

Although I don’t work in that area now, I work on computer systems scheduling for 15-years, specializing on at-scale systems. These were airline, banking and transaction systems. That had very similar constraints in so much as they all used run one one or two mainframe systems. I was the lead architect for a system we spent more than 2-years modelling an internet banking system that eventually successfully supported 900,000 concurrent users on two servers. I helped fixed the design and scheduling for a single system that supported 23,000 concurrent users, a record at the time. Before you all say but yeah cloud computing has changed all that, it really hasn’t.

The design and use pattern considerations Roger discussed are key. If you had 100 destinations you’d have removed some of the major bottlenecks, but we don’t. We have the Central Business District. It’s a major constraint and getting people in cars in and out of it no matter how automated the cars are will have the same fundamental problems and constraints. Anyone who says otherwise really doesn’t understand the problem.

It’s true, as automated cars become standard features of our cities, it will be easier and easier to write the scheduling rules to make them work within the unique constraints each city has. Austin will have more than many cities constraints because of the sprawl, because of the access road problems, and especially if we have not viable alternatives.

On the “self driving” Mercedes observation, I had never thought of trying this.

Update: 1:23 added in Qoura link.