The latest Freakonomics podcast episode is about the guaranteed minimum income, and starts off with the premise that for instance, driverless cars will put 3.5 million people out of work who drive taxis, trucks, etc.
In looking for a link to the podcast, I discovered that they apparently put up transcripts that are nicely formatted and have all kinds of informative hyperlinks in them. I may actually have to start reading this podcast rather than listening to it, although I do enjoy the host.
I posted the link to the episode before finishing it, and then as I listened I came across a pretty remarkable part that referred back to my point about how capitalists may well see the value in a minimum income to prop up consumer demand, Keynesian style.
They talked to Sam Altman, the head of a venture capital firm in Silicon Valley. This section of his comments goes right to the heart of the topic of this thread:
His company’s blog has a post about the five-year project they intend to do, giving people a guaranteed income and seeing what results.
Well, the people who left the farms and went to the cities DID kinda die like rats, if you include greatly diminished life expectancy as “dying”:
Sure, I’ll agree that we’re not going to see any time soon fabricators that require a laundry list of materials. I’m more envisioning something like what happened with electronics with the invention of the silicon chip. Back in the old days every electronic component was manufactured individually, out of optimal materials. So you had carbon resistors with copper wires, for instance. But then they realized that you could make a resistor out of silicon–it wouldn’t be as good as an off the shelf carbon resistor, but it could be done. And so every electronic component that used to require a bunch of custom built glass and carbon and copper and ceramic could be replicated in silicon.
And so I’m imagining something similar eventually happening, along with virtualization of functions. Does your joystick really need all those different feedstocks? Maybe it does, but maybe you could substitute carbon for almost all of them. And for those few pieces that really need exotic materials you wouldn’t have a fabricator that can extrude steel springs on demand, instead your recipe would say “buy 8 30mm springs from the hardware store for $0.25 and toss them in the hopper with the shredded soda bottles.” Your fabricator probably isn’t making silicon chips, but you buy those as commodity items, or rip them out of old musical greeting cards and Happy Meal toys, and the fabricator knows how to reprogram them. And so on.
And of course there’s still the likelihood that ordering a joystick for $2.00 and having it delivered by drone 30 minutes later is going to be a better option. But that’s only because manufacturing the joystick is so cheap that they are practically being given away. Maybe you can sell 100 million joysticks for $2 and they only costs you $1 each, and so you make $100,000,000. But how are you the only person in the world who knows how to make joysticks that cheap?
Even if it really does require a fairly expensive factory to make these things, something that is out of reach for a guy with a garage, it’s still going to be easier and easier to set up a factory that churns out arbitrary widgets. It’s going to require lower and lower capital investment, and lower and lower operating costs. A factory producing headphones could literally switch overnight to making joysticks if joysticks were more profitable. So how can you sell 100 million joysticks and make a dollar on each one if there are millions of arbitrary widget factories all over the world who would be happy to make $0.50 per unit, or $0.25 per unit, or $0.10 per unit, or $0.01 per unit? $0.01 profit per unit is still $1,000,000 if you sell 100 million of them. But a million dollars is a lot less than 100 million dollars. A million dollars is barely enough to buy a decent house in San Francisco in 2016. I mean, you’re rich compared to the guy who’s living in a cardboard box and eating out of dumpsters, but you’re not RICH.
So what happens when every moneymaking scheme is sped up so fast that within a few weeks of profitability you have a million copycats jumping in and driving your profits down to nothing? The only way to really make money is through something that other people can’t just copy and churn out.
So there will be fabulously wealthy people in 2116, but they won’t be wealthy because they own factories or software or intellectual property. Those are the ways to get rich in 2016 but will be obsolete in 2116.
Here’s a Slate article about basic income. There’re going to try giving 6000 Kenyans a basic icnome for 10 years and see what happens. Doing the experiment in Kenya makes it a lot cheaper than it would be in the US/EU.
You need to decide what century we’re dealing with. Your previous post mentioned the last 100 years, which is what I was referring to. Now you’re in the 1840’s. Yes, before the advent of vaccines and hygiene there were times when life was short in the city. That had more to do with overcrowding than the jobs, however.
[QUOTE=Lemur866]
Does your joystick really need all those different feedstocks? Maybe it does, but maybe you could substitute carbon for almost all of them.
[/quote]
Exactly. Differently assembled carbon can be a conductor, a semiconductor or an insulator, for example.
And carbon can be sucked out of the air using artificial photosynthesis. By 2116, we’ll be wishing for the good old days of high CO2, when the feedstock for our assemblers was easier to extract. The atmosphere makes a good, free, method of transporting carbon for free to the point of use.
Artificial photosynthesis could be competing against genetically engineered plants to suck carbon dioxide out of the air, and both processes could producing hydrocarbon fuels to burn in any vehicles that still need fuel. Sure the oceans will still be warm, because of their vast specific heat capacity, and the ice will still be melting - but carbon dioxide will become a commodity rather than a waste product.
Not a utopia, by any means - but very different to today.
FYI, the phrase “for themselves and their families” returns more than two million Google hits, so that might be a clue…
Doesn’t everything (any reasonable phrase) return two million hits?
“any reasonable phrase” returns 25 hits 
Heh. OK then.
The On Point radio show tackled AI and the professions. Here’s the podcast of the episode, plus some relevant links, and below is the teaser:
What you are talking about isn’t foreseeable robotics, it’s fairyland. It may happen but so may time travel and wormholes through space.
You are ignoring the reality of now, even though it’s right in front of your eyes. There are any number of things that use cheap materials and publically known processes in cheap factories that still sell massive amounts of stuff and the owners of which still makes heaps of money. Efficiency is everything. Unless transport costs are high (and they aren’t, even now, let alone in the la la land you are imagining) it’s far, far more efficient and hence ultimately cheaper to make a hundred factories that each make a bazillion of something and sell those bazillion things just above cost than to build a bazillion plants that can each make one of the same hundred things.
Here’s a fascinating piece. It’s arguing that the age of fiscal insecurity that would make basic income appealing isn’t coming in the future: it’s already here. It citesa J P Morgan Chase survey (you know JP Morgan, such wild-eyed progressive futurists) which shows that income volatility is increasing while median income remains stagnant or decreases.
The middle class is being hollowed out, slowly and in non-obvious ways, as wealth and income in America continues to shift to the oligarchs. The future is here and now, and it isn’t pretty.
I could imagine a hardware store in the not so distant future selling raw materials like this. Not to say it would be a huge part of the overall market, just filling a niche.
During the 18th, 19th and 20th centuries most people would have gone to local cobblers for their footwear until the rise of factories for production for the masses; whilst the upper gentry in Britain went to the craft makers in London and patronised shops like Lockes. Which cost enormously and made up for it by paying the very highly skilled craftsmen who lived in very wretched conditions as little as possible.
Still, they were good shoes and boots — and oddly people in the past may have had more comfortable and longer lasting footwear than mass produced stuff today, if they could afford it, simply because the boot was made for the foot ( and most of my pairs of trainers survive about 4 years rather than 40 ). Lockes and others made wooden lasts of each customer’s foot, moulds which were replicas to their exact measurements, kept in vaults past the customer’s death, many of which are still extant.
Now, with home 3D printing, with the capacity to pre-pattern exact measurements of bodies and feet, how long will it be until materials and current fashion designs, the latter by internet, are delivered to the home for composition — home dressmaking without effort — wiping out entire industries across the globe ?
And that will definitely be an improvement!
Aaand the rot starts on upper middle class jobs …
Gonna be working on bankruptcies at first. How fitting!
Investment managers, watch out!
It’s a lot easier to create an AI that can pick investment opportunities than one that can do a decent job mopping the floor.
Some thoughts by Dana Blankenhorn, the computing columnist, reprinted in The ( British ) Register, linking all this to the Passion of the Trump.
Why everyone hates Salesforce’s Marc Benioff*
***Trump followers demand an enemy. Silicon Valley is that enemy
**
Benioff and other captains of Silicon Valley industry look around them and see waste. They see market friction. They use the Salesforce.com cloud to eliminate that friction, replacing salespeople with apps and middle managers with clouds.
They never consider the price. Since 2010, US counties with populations under 100,000 have lost more businesses than they have created.
**
…
***Marc Benioff insists he just wants equality, as he campaigns against discrimination in North Carolina and elsewhere. But alongside the fight against HB2, Benioff and other tech titans have enjoined the battle for H1B, a demand that tech be able to get the talent they need from India, or China, wherever it happens to be. Trumpistani kids are already losing the fight for elite college places to the wealthy of overseas, and now these kids want to cut in line – what will happen to US? They’ve already lost jobs to workers at overseas manufacturers - jobs the leader of Trumpistan says he’ll bring back on shore.
…
Next, they will come for the truck drivers as they have destroyed the cab business. They will come for the lawyers, the doctors’ office help, the ambulance drivers. They will come for the auto repair yards, for the actual car dealerships, for the school bus drivers, for everyone Trumpistan knows. What will happen to US, they ask, when all the strings that held our lives together are held by men like Benioff, and all the jobs are done by bots, by robots?
Silicon Valley shrugs its shoulders at this. New jobs will appear, they insist. There will be better jobs, easier jobs. Trust us. It has always been that way. Think of it as evolution in action.
The wide-eyed soothing Panglossism of those useful idiots has it’s own amusement.*…
*
Can anyone suggest a stock that will take off if the robot apocalypse doesn’t happen? I’d like to invest in that, because I think this subject is one of the most over-hyped, sensationalist pieces of ‘conventional wisdom’ I’ve ever seen. It’s up there with flying cars as one of those things that everyone knows is right around the corner, but which is not likely to happen for many decades, if ever.
Instead of asking academics for their opinions on the robopocalypse, I would suggest talking to factory engineers and other people actually on the ground making things work. Because this is one of those ideas that sounds very obvious in theory, or when viewed from 10,000 feet, but breaks down when it meets the complexity of the real world at the lowest levels.
Let me give you an example. I just read an article that went like this: Driverless vehicles are coming. Therefore, there will soon be no need for long-haul truck drivers. Hundreds of thousands of people will be tossed out of work.
You can only believe this if you believe that truck driving can be reduced to the act of steering a vehicle down the road. In other words, from the viewpoint of an outsider who knows nothing about what truck drivers do as part of their daily activities, this can make sense. But as you drill down into the job, you start to realize that truck drivers do an awful lot that isn’t ‘truck driving’. They act as agents for the company. They monitor loading to make sure it’s safe and balanced. They act as monitors of vehicle condition. They handle emergencies. They take payments and do paperwork. They spot opportunities for sales. They can be negotiators. They do a million things that add value to the job that have nothing to do with steering a truck down a mapped road.
And when they are just ‘truck driving’, it’s sometimes around a loading dock, each of which is unique and has specific hazards. They have to read hand signals from others when backing blind in a tight space. Sometimes they can’t get into the exact position they wanted, and have to improvize loading by rigging ramps or assembling human conga-lines to move cargo. If the vehicle breaks down they have to direct traffic, call authorities (different ones depending on the load), set out warning reflectors and signs, and on and on.
Could all of this be done by machines? Perhaps one day, if we redesign the roads, the loading docks, enforce paperless transactions everywhere, yada yada. That won’t happen in my lifetime. So what you may see at first are very specialized forms of autonomous truck driving - say, from one specially built loading dock to another within the same company. You can see the beginnings of that with autonomous material movers inside factories, like Amazon’s automated warehouses. But the general job of ‘truck driver’ is safe, even though it will be whittled down on the margins by automation.
Humans bring value to organizations not just because of the strict definition of the tasks they do, but because human brains are incredibly powerful, generalized computers. We have this vision of people as cogs in an organization driven by top-down decisions of management. But in well-run organizations, the information flows in two directions. No plan is perfect, and we rely on the people at the bottom to use judgement to work around the flaws of a process, to spot opportunities for optimization that upper management can’t see, etc. This is critical to the efficient functioning of any large enterprise. If you don’t believe me, go look at what happens to companies when their employees go on a ‘work to rule’ strike.
We are a long, long way from robots that have a generalized intelligence - the kind that can can employ lateral thinking to spot problems and opportunities never considered by a high level planner. We can currently give them some level of judgement within certain boundaries, such as an autonomous car that could spot a road hazard and avoid it. But that software wouldn’t be able to conceive of the concept of spotting a robbery taking place on a sidewalk and calling the police, or take the initiative to note that there might be a better way to deliver its goods if the company just changed the way it packaged a product, or whatever.
It’s always the details that get you. And the world is full of critical little details.