While this iteration of the product seems like it is not ready for prime time, a couple more years of improvements and you could have a machine able to reliably do the work of 2- 3 employees. There are many other industries where the current version of automation is clumsy such as in trucking or fast food but could look completely different in 5 years.
We are rapidly approaching a future where the entry level jobs like fast food / picking crops / warehouses can be automated away and even well paying ones such as truck driving / taxis could disappear or scale down drastically. We really should be having an urgent conversation on how to handle this and what the responsibility of those that own the means of production will be (if any) when this occurs.
Yang is obviously speaking about this and suggesting a BI as a band-aid / solution but I do think this should be getting a lot more attention that it is.
Exactly. In the 1980s people in America mostly got coffee from machines, in their home or office. Few humans were employed making coffee ("barista" was not really a job title in the 1980s and if you told people it would be in the future, they would probably think you were crazy).
I know what you mean, not sure why others are brining in spurious arguments.
Coffee shops in the 1980s were donut shops, mostly with truckers sitting around.
Well it's gone full circle now, as Starbucks (and a lot of other chains) use machines to make the espresso part of the drink at the touch of a button.
The barrister is still responsible for steaming milk, but I only assume that is because the sounds and smells of that are what make the coffee shop atmosphere.
Starbucks is selling an experience and it's central to this experience that there be a human putting a little TLC into what happens, whether it's a friendly greeting, a product recommendation, remembering your name and writing it on the cup, etc. The key is literally for there to be a human involved, because we get emotional satisfaction from a positive interaction with other humans.
There's nothing new in this discussion -- we've been automating away manual labor since the Industrial Revolution. We end up creating these human touch jobs with the resulting productivity gains (this is the transition from a manufacturing economy to a service economy, the value isn't in the muscle anymore, it's in the relationship).
In fact, thanks in no small part to capitalism's ruthless efficiency, loneliness and isolation are at record highs. The demand for a little TLC has never been higher.
We'll have fewer pizza makers but more people whose job is essentially to make you feel good.
From my experiences with coffee machines that produce cappacinos, the act of frothing milk seems to be one of those tasks that is much more difficult to automate than I would have thought. I've had a few that were passable, but none that are remotely as good as a decent human made one. The actual coffee on the other hand seems to be easeir to automate.
Fair enough, though a well adjusted Super-Automatic Espresso machine can reliably and reproducibly get you 90% of the way there hundreds of times a day. The majority of customers wouldn't be able to distinguish that last 10% anyway. Have a barista on hand to adjust the Super-Automatic daily, to account for changing bean freshness, and you're pretty much good to go.
How can a barista accomodate the variance in taste of a customer when pulling an espresso shot? The barista can dial in the pressure profiling, how long to pull the shot, and the grind size. For a given batch of beans, based on the type, level of roast, and age, there likely exists a sweet spot for those parameters that most baristas would agree upon. My understanding is that this is what baristas do anyway - adjust the machine for the beans available today, and then just crank them out all day long. A super automatic (or future versions) that's dialed in daily could achieve pretty much the same thing.
AFAIK, the barista will try to pull the best espresso out of the beans that they have. I'm not sure variance in customer tastes can be accomadated at this level, beyond a ristretto or an americano.
I never had a coffee from a Costa machine, but purely from the looks of it, and my experience with similar looking machines, I wonder: does this really produce espresso (as in say 16grams of fresh ground beans in, water at proper temperature and pressure of like 6 or more bars, yielding something like 32grams of beverage) or rather something which can best be described as 'strong coffee'?
Yes that's pretty much spot on. Making espresso with a manual lever machine isn't too hard too mess up, especially not in comparison with pressing one button or filter coffee. Yet it's apparently not that trivial to make such a machine, and keeping it working for hundreds of cups a day: prices of those are easily well over 10k$.
Sure but espresso varies, change the bean even the age of the bean and the recipe needs to adjust. Not to mention not everyone wants the same experience every single time, and not every person wants the same exact espresso that the last person had.
I think the key to this is to hide the robots from public view. Crank out pizzas, but not with the machine in human view. Have someone visibly handing them out. Let buyers assume the pizzas are made by artisans out the back.
Of course it can. One of the best machines for making coffee is the Technivorm Moccamaster (the thermo pot version). It's been in the market for decades, lasts for decades, it's easy to clean and together with a good grinder and fresh good beans, it produces a perfect mug of coffee every morning.
They were a thing in the 50s, 60s, 70s. Maybe more of an urban or hippie thing.
I know the term "coffeehouse" was a cultural touchstone in the US because it was used in titles of books published before the 80s. Sorry I don't remember specifically, but this "nothing existed until Starbucks" is ridiculous. Starbucks came out of somewhere. Coffeehouses came from Europe to the US, before and after the American revolution.
I'm tempted to just call out the absurdity of arguing this point, but instead I'll just air-drop a factoid: if you Google "https://www.statista.com/statistics/196590/total-number-of-s..., you'll find a graph showing approximately 1,650 specialty coffee shops in the US in 1991, and 31,490 in 2015.
I'd link directly to it but the site forces you to pay to see the graph unless you clicked on it from a Google search results page, which is bad and evil. I can't find a better source.
Starbucks has something like 15,000 locations in the US, half your latter figure. They could have a million stores and it still wouldn't mean they invented coffee.
Even in NYC, a coffee shop was mostly a Greek diner. While it may reveal my Irish working class roots, I never saw espresso made or consumed outside of an ethnic Italian restaurant in Arthur Ave until I was like 17.
Last week, I had a mediocre latte from a Starbucks at an I95 truck stop adjacent to a cotton field in North Carolina.
The major difference here is the economies of scale. When you have a centralized facility, you can afford to invest in expensive, high-throughput machinery because you make up for those investments with volume. However, centralizing necessitates some distribution costs, which in the case of many foods imposes some harsh constraints on ingredients and quality; there’s a reason those pizzas are all frozen, and don’t reaaaaaallly taste all that fresh. If the cost of these small-scale production robots drops low enough (achieved through smarter control rather than high-precision parts), you can remove the ingredient constraints and make a better product while capturing savings of automation.
If you want to model a city as a circle, you should probably instead use a radius where half of the area of the circle lies within that radius. When you halve the radius, you divide the area by four!
For example, the area of a circle with radius 15 km is about 707 km², while the area of a circle with radius 7.5 km is only about 177 km². If the population density by unit land area is evenly distributed, this circle will only enclose a quarter of the city's population, rather than half.
I think an interesting future limiting factor is air traffic from delivery drones. Right now it seems absurd to worry about that because there are no delivery drones flying, but if this method becomes popular, the presence of other drones in the airspace could limit delivery drones' mobility.
I'd like to add that the pi r^2 circle we know and love isn't the only one. It depends on the "metric" which in that case is Euclidean ("as the crow flies"). If you have have traffic rules, like only flying certain routes you get a different shape circle and a different area as a function of radius. A famous example is the "Manhattan" metric where you drive around on a square grid. In this case the "circle" looks just like a square and has area 4 r^2.
I feel like there's a lot fewer SKUs involved in store frozen pizza than varieties available from a takeaway pizza store, and the volumes are also a lot higher.
So the trick might not be "cheaply and reliably automate making a fifty thousand a day of the same pizza" , but in the detail of "cheaply and reliably automate making a pizza, composed of any combination of the basic ingredients, but only amortised over a hundred an hour for a few hours each day".
That's an interesting point. Does that mean you think a chain like Little Caesars, that offers about 4 types of "Hot N Ready" pizzas could basically automate away most of its workforce? If frozen pizza is possible, it seems really weird that no one has pursued this vigorously already.
> Does that mean you think a chain like Little Caesars, that offers about 4 types of "Hot N Ready" pizzas could basically automate away most of its workforce? If frozen pizza is possible, it seems really weird that no one has pursued this vigorously already.
Little Caesar's entire business model is "the lowest quality pizza at the lowest possible price". I assume they already have automated away most of their workforce; given a minimum wage, that's a key part of low prices.
I've only been to Little Caesars a few times recently, but in none of those experiences did I see more than a single young (hourly) employee running the entire operation.
“Machines have been making frozen pizzas for years, but Picnic’s robot differs in a few respects. It’s small enough to fit in most restaurant kitchens, the recipes can be easily tweaked to suit the whims of the restaurants, and — most importantly — the ingredients are fresh.“
From my extensive experience watching episodes of "How It's Made", frozen foods still often are hand-made. There are enormous assembly lines, heavily automated but with some steps done by hand. Especially those that involve irregularly-shaped, delicate, or sticky objects[0].
They did do an episode on frozen pizzas[1], and it is fully automated. But the techniques are difficult to adapt to making individual pizzas. Sauce, cheese, and toppings are sprayed all over the assembly line; the stuff that falls off is dumped back into the hopper.
They also did an episode on building pizza-making vending machines, and it does seem as if that was technology that was ready years ago.
> I see frozen uncooked pizza in stores. I feel like it existed 20-30 years ago too although I'm not sure exactly. Surely these are not hand made.
When looking into manufacturing of things I'm often amazed how much manual labor there often is involved as humans still are cheaper than construction and maintenance of machines doing complicated tasks. Also humans can be replaced simpler on failure than a machine and can simpler be adopted to varying products.
Specifically on pizza I only found this video: https://youtu.be/OMPFlbGXdFA where humans at least fix the salami slices. It doesn't show how they portion and form the dough, which I assume to be the complicated part.
> We really should be having an urgent conversation on how to handle this and what the responsibility of those that own the means of production will be (if any) when this occurs.
The problem with that is the people who would be in a position to make change don't feel the constriction of the immediacy of it and those who do are rarely invited to the table to talk.
Just change the ownership rules, the old concept of private property is clearly nonsensical for the same reason that the game of Monopoly was meant to be a warning, not a tutorial.
I don't mind people downvoting but you're not gonna tell me that people with the 18th century had timeless wisdom that allowed them to extrapolate to today's economy in the 21st century and appreciate all the problems of resource allocation, externalities, and financial markets. If your principles are implicitly predicated on infinite land and labor availability then its possible your understanding of economics is kinda primitive.
Our ancestors in the 18th century didn't have a time machine, but neither were they stupid. Our view of economics is going to look similarly primitive in 300-400 years. The notion that jobs, and thus work, was going to disappear has been a theme since the Industrial Revolution began, in the late 18th century. A treatise from that time period on how to avoid a depression is certainly suspect, but the crash in 2008 wasn't so long ago, so it's not our understanding of market forces is 100% either. (To be fair, our understanding did allow us to recover from it, in an imperfect way.)
We still don't have a better answer as to what the displaced will do for work than we did in the 18th century, other than "oh, we'll create new jobs". Not comforting, especially to anyone that simply can't afford to go to a coding bootcamp, or those that have gone and dropped out. They, too, deserve not to starve or die in the gutter. Thus, there is wisdom from the 1800's that is relevant, even today.
Sure, perhaps we will solve by shrinking ourselves down like the ants did to effectively expand our territory while reducing our footprint. It's clear that our current large size is economically inefficient.
"entry level jobs like fast food / picking crops / warehouses"
These things are very hard to predict. We have a long history of such "automations" and the outcomes have been often unpredictable. Effectively, "automation" is interchangeable with any labour saving tech. Production lines, tractors, etc.
Agriculture basically did shed jobs due to the green revolution and mechanisation. Fewer farmers grew the food and people moved to towns. You could argue that towns pulled agricultural workers in rather than automation pushing, but machines replaced people regardless.
Industrial manufacturing, OTOH, increased output and added jobs to the sector for hundreds of years despite labour saving "automation" progressing constantly.
It's only recently that manufacturing employment plateaued. Global trade and other shifts have caused some severe job shedding locally, but even that is only in the last 30-40 years. The long term trend was more automation and more jobs.
Computers landed on every desk very fast. Typists, mailrooms, secretaries and other such "entry" jobs became mostly obsolete. Still, the "white collar" and administration sectors experienced massive employment growth since the 80s.
Think of colleges, with their ever rising admins/academics ratio even as "administration automation" technologies became available.
When it comes to something like automating restaurants it's so dependant on "consumer preferences" and food culture that I wouldn't hazard any guesses.
Prepared food isn't a commodity and restaurants aren't driven by efficiency. There's no hard, purely rational reason why one burger is worth €3 and another is worth €20.
McDonald's us extremely efficient, in terms of labour cost per bite, but most newer competitors compete by being less "efficient" and more "artisinal."
Couldn't agree more. It is highly unpredictable. We tend to get antsy about the entry level jobs and quickly jumping to some conclusion where if you're currently doing the toppings at Domino's pizza by hand and Domino's replaces you with this machine, that you suddenly will have nothing to do but go home and lay on the couch all day while being unemployed.
There are many things someone in such a situation may decide to do next and that "next" thing for that person is what could be unpredictable. And if tens of thousands of these people were replaced over a several year period, each of them making individual decisions about what to do next, it is hard to say where they all end up. We simply don't know.
On the other hand, thinking creatively is often discouraged at the majority of bigco white collar jobs and executives would gladly replace an expensive person with cheaper software doing the same task if they could do so without tripping on labor laws.
With the amount of mind numbing easily automated work I’ve done in my career I wonder if they don’t realize or there is some kind of tacit social contract not to cause mass unemployment. Because companies could replace like 50% of office workers if they were willing to redesign their processes.
Mmm nope, they could replace the workers with more expensive programmers. So it gets into a question of how many office workers can you afford for the price of one programmer. Often not at all obvious, especially as programmers can often trend towards huge inefficiency in enterprise/corporate environments. ROI of business IT projects is not always positive you know.
Pft, please. Back when I was trying self employment in shareware games, the first thing I wrote was an AI to compose music for me. I can’t write music.
In Japan, robots to aid with elderly care are being developed[1]. Probably won’t eliminate all elderly care jobs but likely will reduce the amount required. Japan is facing worker shortages and doesn’t like immigration, hence robots. I personally wouldn’t like to be looked after 100% when I’m old by a robot, but it will be interesting to see how it plays out in Japan. The rest of the rich world will face similar problems in the future.
I've heard it said like this: AI will take over repetitive manual and repetitive cognitive tasks. The only thing required is that the task be repetitive (so plumbers are safe (every house is different), but truck drivers and radiologists maybe not).
>While this iteration of the product seems like it is not ready for prime time
Yeah just looking at it I couldn't stop myself from imagining how nasty it's bound to get after getting put to work full time. Too many nooks and crannies and it would need to be easy to tear down and put together again for cleaning.
you do know we've been replacing humans with machines since the Industrial revolution, right? These people that were replaced didn't just go off and die. They found other more advanced work, allowing our civilization to produce more wealth allowing all of us to be richer. The same will continue.
Most people who use this argument forget about the scale and duration of this transformation back then and now. Back in 1800s there were merely ~1 billion people, then in 1900s it was ~1.6 billion. Lets say even if 50% of people back then lost jobs you had to find new jobs for 'only' ~0.8 billion people and you had more than 100 years for that transition. WWI and WWII at the beginning of 20 century 'helped' to reduce population and keep other busy to rebuild what got destroyed during the war. I guess we would like to avoid that scenario.
Now we have 7.7 billion people and it is estimated we will have 10 billion by 2050. Pace of current technical development is much faster than in 1800s and education for acquiring new skills probably takes even longer than in 1800s.
Will we be able to hypothetically reskill 5 billion of people in the next 50 years? Personally I doubt it.
Absolute numbers are a distraction. Retraining 1 person in a population of 10 has the same net demands as retraining 100,000 in a population of 1,000,000. Human society is massively parallel by default.
I think most people don't imagine properly the world as automation increases at all levels.
The first expectation is perhaps that mass unemployment would abide; clearly we don't have mass unemployment in most places (most of the world has a relatively high employment), so the phenomenon must not be occurring or there's the oft-mentioned effect of workers moving to high level tasks.
Since Quantum Supremacy was en vogue, let me coin another term: Robotic Supremacy. That is when a robot (i.e. some kind of automatic machine, not necessarily a humanoid robot) will be able to perform each and every task a human can do more economically. Humans will become unemployable. My hypothesis is that Robotic Supremacy is still far away (maybe a century); and also that this milestone doesn't matter as much as it appears.
The scenario of unemployability is highly unlikely; there's almost always something you can get a human to do that'll pay their food and shelter (we're fantastic machines evolved for billions of years; robots are still far away from domination in many niches). The question is how much this person will earn -- perhaps increasingly less, not more; because it has more competition and the tasks are not as essential as before. So a natural manifestation would be rise in income/wealth inequality. That's precisely what we're seeing. If left unchecked, most places will see a spiral of a tiny elite concentrating all wealth.
Except good chunks did go off and die because they couldn't adapt, or because they were easy to exploit. (E.g. child workers). Or lived in subpar housing due to urbanization, which made perfect breeding grounds for typhoid, cholera and assorted fun.
You can't just handwave those changes away by saying "ultimately, they found other jobs". They didn't. There were entire generations that suffered. At some point there was an equilibrium, but it was far from instant, and "advanced" work doesn't help the person displaced.
They didn't die out, but they died in much larger numbers, because the industrial revolution lead to an exploitation of child labor. Combined with non-existent protections, a lot of them did die on the job.
Children started to be children for longer because labor unions helped end the practice. That was a good chunk of time after those jobs changed, and it wouldn't have happened without active pushback. (The AFL pushed for an end for child labor under 14 in 1881, and it took until 1938 to get the Fair Labor Standards Act)
The idea that paradigm-shifting transitions somehow don't affect anybody because "there will be new jobs" shows a stunning unawareness of historical precedent, or how capitalism in general works.
"because the industrial revolution lead to an exploitation of child labor"
I'm skeptical of the idea that before factories, farms didn't use child labor. Isn't the practice of closing school during the summer kind of suggestive?
Anyway, you seem to be arguing the opposite of the previous post, that there were more jobs and this was bad. I thought the issue was there were fewer jobs and that was bad.
What does "education wasn't a given for the poor" mean? That's such a vague statement I don't see how you can link it to anything. Education can mean grammar school, high school, college...people in the past, who were farmers, were generally poorer and less educated. However, less educated doesn't mean no education. My grandfather was a farmer in the 19th (and early 20th) century, and no, he didn't have a college degree, but that doesn't mean he never went to school or was illiterate.
Linking to a google query like that is a step below even Wikipedia. If you have a source you consider authoritative, why don't you provide that. I'm not going to read the clickbait garbage in those hits.
"Journal of Inquiry & Action in Education, 5(1), 2012" was a top link for me, apologies for not picking that, but I figured Google has something for every bias.
> Except good chunks did go off and die because they couldn't adapt
Maybe that's a feature, not a bug. It's worthwhile imagining how life might be impacted in the present if a sizable portion of the homo genus has not died off over the past ~2.5 million years because they couldn't adapt. How would humans today handle the continued co-existence of homo austalopithecus, habilis, erectus, etc over those ~2.5 million years up until the present?
> You'll sing a different tune if you find yourself on the wrong end of that equation.
(We all, eventually, end up on the other side of the equation.)
Given that resources are indeed not infinite, there must be a limit to this planet's capacity to sustain human beings. Most governments don't seem to want to limit childbirth and instead _encourage_ it (more babies == more taxpayers).
Once that limit is exceeded, barring technological advances, what other outcome can there be?
Right, that's historically been true for humans because we are pattern recognition workhorses.
We have some other examples worth considering though, like literal workhorses. Once the engine came around, horses became inefficient and were relegated to shows and races. With AI able to do superhuman pattern recognition, it's only a matter of time before humans also become a luxury, as they're no longer needed for traditional work.
In a lot of data science focused places, like mine, people are more or less researchers. There is no repetitive work, and everyone must show creativity and autonomy. You can't automate that.
Uh, yes you can. Robots are already being used for battlefield simulations coming up with moves humans couldn't think of. They beat humans in Go by finding better moves that's entirely creative. They write news articles, paint, create music, and help in diagnosis of disease. Humans are all going to be replaced soon enough. We'll be serving AI masters or wiped out within 50 years tops.
AI's permutate artwork. They don't feel, they don't assign or derive meaning. They don't feel, they aren't inspired, they aren't moved.
They don't derive joy from the beautiful, despair and disgust from the loathsome, awe at the sublime, and dissolution and insignificance in the face of the all encompassing.
They don't dance with glee at a bright and bouncy tune, they aren't struck to the edge of tears by the melody of wind through a forest of bottles. They know and can reproduce patterns that are labelled for them, but they don't get it, ya dig?
They don't relate or understand the soul of Jazz, the message of Blues/Rock and Roll, the Struggle and perseverance evoked by a good fight song.
They can't appreciate the byproducts of their work, or "make 10 more, but different."
They can't appreciate, display, or develop technique; they cannot perform. When composing, they generate content based off of higher-dimensional correlations between words as encoded through syntax and grammar, but a poem generated by a machine is not but a permutation of words ejaculated forth, with no rhyme, reason, or correlation to the world at the time; even being removed from the whimsy of it's programmer.
Art is a tricky thing. It is what it is because a human found that at that time in their life, in the details of their personal situation was the right time for that work to be born forth into the world, in all it's symbolism, ugliness, beauty, sublimity, and to affect all those who gazed upon it.
Think hard about the significance of that. That man's mortality factors into the fruits of his labor; something a machine, deriveable from a prescription can never truly know or imitate.
It isn't humanistic chauvinism...merely that as the clock is not the first cause of time, so the AI is not the creator of the Art it produces, if what it produces can even be truly called Art.
It may not always remain that way. Right now though, it is.
Art isn't born in the mind of the creator. It's born in the mind of the consumer.
People are moved everyday by music created by machines. That's enough to put artists out of business because the output is the same even if the input is different.
So you see art as an emotional response in a bottle?
I never really accepted that. Otherwise, the response I've gotten to a hypothetical man under a rock perfectly replicating the Mona Lisa without having witnessed or heard of it before would be as great and worthy of celebration of artistic work as the original; many of more artistic inclinations I've spoken to balk at the very suggestion. Man that was a fun day in Philosophy of Art class.
Most of the artistic I've gone back and forth with do not see the end product alone as Art, but also the process, from ideation, to execution, and finally display. The reason behind the creation of that particular work and not another even has a place in the Art-Ness of the work-of-art.
Just because something is moving to someone, somewhere can be said to be necessary, but not quite sufficient to bestow the quality of being Art. It's a rather perplexing problem to discuss. I tend to approach it like linguists do language. Descriptive, not prescriptive. Though I haven't mingled in artistic circles recently to reconduct a census with regard to generative music/art. Most of those I do run into though tend to be non-committal on the subject and just treat it as just what I've described. A pleasant, and surprisingly novel sensation from an unexpected origin.
Given the fact that the economy is at full employment and that there are more job openings than people looking for a job, I'm not too worried at the moment.
> Given the fact that the economy is at full employment
The economy might be at full employment, but people are often underemployed, getting few hours, and shit wages. Most people are employed but many don't make enough to support even just themselves on their earnings let alone a family. Real wages have barely increased since the 1970s while costs have risen. We've had the first generations of Americans who are worse off than their parents. I wouldn't say our current situation was at all encouraging.
I don't get why we use this term. You're either employed or you're not. Just because you got an education in something that is no longer valued by society doesn't mean you're entitled to a job doing whatever you're educated to do.
Society rewards those that provide some good or service that is wanted/needed.
Furthermore, employment by others is not the only option. You can also work for yourself (except when the state forbids it through licensing and permits)
We have a term, "employed", for people who work and earn a living wage.
We have another term, "unemployed", for people who have no work.
We need a third term, "underemployed", to describe people in between those two situations, who have some work (so they're not "unemployed") but nonetheless do not earn a living wage (so counting them as "employed" is disingenuous).
Correct me if I am wrong but doesn't under employed often also refer to when you have part time / casual work but want full time employment? You might not be able to survive of what your earning but your not part of the Unemployment statistics?
Where are you getting this? From the govt statistics that don't include people who have dropped out of the workforce entirely? The same stats that don't differentiate between a salaried career and scraping by in the transient gig economy?
I'm extremely skeptical. Can we afford to just dismiss all these legitimate concerns because, "the govt told me everything is okay"?
Seems like a poor way to rationalize avoiding planning ahead.
The number of open positions says nothing about the nature of the work. I can open a position for a doctor to work for me for $100 a month and while it's an "open position" it will not be and could never be filled and yet it's still a job opening.
In addition some people are working multiple jobs to barely survive so again there is no defined relationship of number of jobs to number of workers or whether that's a good or a bad thing.
I would agree with you if wages had risen enough that workers could work one job and have their needs met but that won't happen any time soon, especially with robots on the horizon. The only thing that makes sense is to plan ahead.
I think that caused by social separation. Today people have hundreds or even thousands of "facebook friends" but not one where they can sleep on the couch when they are down.
Also consumerism. Lots of people have the latest and greatest iPhone and OLED TV but don't have 6 months worth of savings(I think that's the minimum you should have) in the bank.
It's a lot harder to find your next job when you stressing because you are going to be evicted.
> These people that were replaced didn't just go off and die
This is exactly what's happening in every large metro area today. Larger amounts of work are being by fewer, highly paid individuals using technology. This drives up rents and fuels the homeless crisis everywhere.
You do know that the rate we're replacing humans with machines far outstrips what we've previously seen, right? As in, this time around it's happening much quicker and is much more widespread compared to the first time around. That people and legislation are already struggling to adapt which will ultimately cause far greater chaos, market disruption and unrest than we've ever seen.
We as technologists know better than anyone how quickly new tech can sweep across the industry. Once the tech/product/process is figured out, scaling it is trivial in comparison. This is the same song and dance but 100x as loud and 100x as fast.
I don't know that actually. Can you please prove it with data?
I studied the industrial revolution quite extensively at school and I follow AI research today, reading individual research papers.
It's not clear that we're going through anything even close to as disruptive as the industrial revolution. In fact I'm sure we're not. Absolute numbers may be higher because the population is so much larger, but as astutely pointed out elsewhere in this thread, civilisation is massively parallel by default so it's proportions that matter.
And when we get to that, we're not seeing massive job displacements caused by automation. If anything we should be asking why not: why is the software industry apparently so unexpectedly bad at reducing employment? This isn't just me waffling by the way. Economists say (caveat emptor) the same thing:
Productivity growth i.e. the rate at which machines replace people, has been falling over the decades. The spike you might expect from massive automation is not only not present, the data shows the opposite - we are getting less bang for our technological buck, it's a long term trend, and economists are worried/puzzled by it. If productivity growth continues trending towards zero, what it means is that our society is getting no automation benefits whatsoever. It means the only way to produce more stuff is to use more people!
We can join economists in speculating as to the reasons, but either the data is flawed, or we're in a time of unprecedented job stability, despite the froth and hype around AI (whose impact is so far extremely limited).
This is automation on a scale that we have not seen before where machines are actually capable of decision making and precision work. This is combined with a massive concentration of wealth at the top. Also there are a whole lot more people now. Potential long term could be colonization of the moon / mars. Plenty of unskilled labor required and opens a new frontier.
It's not at all obvious to me that colonization of another planet, if we're ever in a position to actually undertake it, will involve "plenty of unskilled labor".
Agreed. Pretty much the opposite, actually - shipping, housing, feeding, etc. a bunch of humans is obscenely expensive in the context of space exploration. You're far better off sending a bunch of robots with a couple mechanics to keep them running. Additionally, I'm pretty sure that by the time we're in any kind of position to do actual colonization efforts, automation effects will have long since taken their toll.
Consider that universal high school, labor union-friendly laws, and labor day were government responses to the last industrial revolution. How will our society change for the next one?
> These people that were replaced didn't just go off and die. They found other more advanced work ...
This almost never happens in real world. The next generations do benefit that. But the current generation is fucked.
You wont retrain 40> years old truck drivers into anything that will pay even remotely the same. Most of them will be stuck doing minimal wage jobs somewhere, and be happy that they have them.
(of course there are exceptions)
I have seen that play out in ex communist countries, where whole swaths of jobs became obsolete over few years.
It's been almost 30 years, newer generations have moved on somewhat, but there are still a lot of people that were better off before.
Industrial revolution was great in general, but its effect on people also gave rise to communism (i am not defending it).
I mean we already have way better automation for pizza making and at huge scale and efficiency -- that's what frozen pizzas are with few exceptions. And some of these are great; safeway brand rising crust pizza is pretty damn good and can be had for as little as $3 a pie.
The machine in the video kind of sucks. Slow, one pizza at a time, poor control of amount and placement of toppings, poor consistency, lots of waste with toppings flying off everywhere.
I don't think machines are going to be putting pizza places out of business. The margins are already quite high, and paying one guy to sling pies at a restaurant and get high quality product isn't a huge expense.
Maybe a place like little Caesars would use this tech but local pizza joints? nah
Have you ever actually picked a crop? It's an extremely complex problem. For the bulky crops (s.a corn) we've had machines for decades. For fruits... we're decades away from automation.
Fruit picking robots are being worked on, strawberries if I remember correctly. Search for strawberry picking robot. It’s possible thanks to advances in computer vision.
Yes, but there will be fewer jobs. That's a problem.
The industrial revolution removed a large number of jobs (basic manufacturing skills) and created a load more (advanced manufacturing skills / logistics / etc.).
There is no expectation or even hope of a new class of jobs coming out of the high-power automation that we are liable to see in the next couple of decades.
Last wave of automation took out millions of manufacturing workers. Instead of retraining to become software engineers they largely went on disability and then as a cohort started dying of drug overdoses and suicides at higher than normal rates.
Most of the new jobs created in the past 10 years are temporary / gig jobs without benefits, nothing like the manufacturing jobs lost.
So, great opportunities for some, but many will suffer.
Just as a side note, but I wouldn't in any way idealize the manufacturing jobs that were lost. A lot of them were so horrible that the new hires quit in the first day (often just not coming back from a break). I'd probably prefer to be an Uber driver than to work in a car factory.
The difference is that automation is now targeting skilled, social, and technical work. Traditionally they've been aspirational fields of work due to their resilience. Limiting capacity to enter these fields is only going accelerate the destruction of the middle class.
No I disagree. This is not "progress", this is "technological progress". We need to make sure that we don't leave older people behind, and we don't pull the rug from under our childrens feet. Of course we should not stop technological progress, but we all (big businesses, governments, everyone) need to think and plan ahead on what will happen to the people. I believe that there is a statistic that 1/2 or 1/3 of USA citizens have at some point worked on Macdonald's. I remember a piece of news about a robot burger flipper and discussion here on HN a few months back, and people were writing how working on Macdonald's paid for college or what have you. Eliminating fast food jobs is depriving important resources from the youth. Who will be paying for that?
We cannot just move forward and leave the people behind. Capitalism should try to care for more than jist the few.
I do state that we should never halt technological progress. All I was saying is that while progress moves forward, we (as a species) should ensure that we don't just consider the shareholder value of decreasing Payroll costs, but we should also consider that the hell will happen to those thousands/millions.
This is not a threat, this is the reality. If we take the food from the table, and exclude them from retraining etc, then at some point they will go hungry and will seek food that is in other tables.
This is already happening to a small degree in some countries. Let's make sure it won't happen to a higher degree and in more countries.
> We need to make sure that we don't leave older people behind
old people can now summon rides on their phone and be out and about instead of being home-bound.
> and we don't pull the rug from under our childrens feet
Everything we build, our children will inherit. It will be their responsibility to improve upon what we leave them. It's not like older people end up taking these technological improvements with them to their grave.
There are more jobs than just fast food. People did jobs that weren't fast food before McDonald's existed and people will do jobs that aren't fast food once McDonald's is fully automated. Right now, on the Internet, you can learn so many things for free. The necessity to learn a skill that pays the bills will drive people to learn skills that can be used to provide goods and services others are willing to pay for. All that is necessary is to open your eyes and look around and see what people are willing to pay for.
When I moved to the SF Bay Area in 2011, I spent ~8 months living on about $32 a day for all living costs including housing. That's under the US poverty line and well below the SF Bay Area poverty line. I bought rice, beans and other dry goods in bulk and had my daily meal costs averaged around $2 a day. People need to learn to budget better.
Also, old people are not necessarily only use-less consumers. Giving them Uber is nice, but if they can and want to offer/produce it would be a waste to put them in the sidelines just because they can't use the latest technology.
Four year degrees need a re-evaluation. A lot of alternative entry level jobs that are hard to automate exist but they have a reasonable degree requirement. A two year job specific degree and a restructuring of higher education is needed.
Why only re-evaluate non-high school degrees? Career oriented education could start significantly early. There's no real reason why 16-18 year olds couldn't already be appropriately skilled to work in many fields save individual maturity.
My big priority would be ensuring continuing access to education. Schooling that pushes certain students toward skilled blue collar work seems potentially a little classist if folk aren't given the appropriate opportunities to achieve something else, if they want it, throughout life.
How is the AI revolution different from the industrial revolution(s)?
A farmer in 1700 would have seen the 1880 machine and said « We’ll all lose our jobs »... and he’d be right. Also « and money will be centralized in the hands of the factory owner ». He’d be right too. « We’ll never be able to make everyone learn ti read and work in a desk job ». Right again. We just raised everyone’s education, and also raised everyone’s IQ in levels that would seem impossible to a 1700 farmer. We’ve also have the underside: poverty for those who can’t get office jobs and socialism to try level the field.
Similarly it seems to us like we’ll all lose our jobs to AI, wealth will be concentrated by 99% more and we’ll never be clever enough to make each of us an AI scientist. That’s what it seems today. You can call to regulate but it’s hard to know in which direction. After all, we had to spend 2 centuries of wars and famines till we figured a merely stable system, and even today, we wouldn’t stand upright if we didn’t rely on Chinese workers being poor. So all the regulation in the world didn’t suppress any problem, we just put it out of sight.
I see no difference between the scale of the AI revolution and the scale of the industrial revolution.
1) Speed. The industrial revolution took place relatively slowly, over the course of 60-80 years (2-3 entire generations). Once a tipping point in automation is reached, it will very likely happen extremely quickly in our connected, easy-capital, globally competitive economy.
2) Reach. The industrial revolution primarily impacted manufacturing and to some extent farming. Today's automation is attacking almost every single major industry on the planet at the same time. Farming, manufacturing, finance, construction, food service, you name it. If your warehouse picker job gets automated away it's unlikely you'll be able to go get a job at Burger King because those jobs will be automating at about the same time.
3) New economies. The Industrial Revolution led to the industrial economy, which enabled the service/information economy. It's unclear what new economy is waiting beyond automation for the workers who lose out. Entertainment economy? Personal servant economy? What's left if most of the manual labor jobs and a good chunk of white collar jobs all go away within a generation?
> How is the AI revolution different from the industrial revolution(s)?
Just to add to what everyone else is saying, the first industrial revolution led to one tiny island conquering a 24% of the planet while simultaneously their historical enemy and immediate neighbours controlled an additional 7.7%, and then a newly unified and industrialised nation [1] managed to drag both of those two into a war that killed 16 million people and which stalemated against those two superpower empires for a few years until America joined in, so even if the AI revolution is basically equivalent to the industrial revolution, I’d still say it’s not going to be fun for quite a lot of people.
I assume the counter-argument is that this time change is happening faster since it's driven by intangibles like code. It took a long time for the effects of the first industrial revolution to spread out because transportation was slower, the world's communication infrastructure was not as developed, etc.
But honestly, I could see both sides of the argument here.
The best thing we can do is to allow the labor market to be more dynamic. This means reducing barriers to entry, including licensing, employer-mandated benefits, minimum wage, etc.
Placing minimum employee costs upon an employer or minimum qualifications upon an employee both have the same effect: reducing the ability of entry-level employees to get work. Either the jobs don't exist due to minimum costs, or the labor that qualifies demands too high a wage.
A highly-restricted labor market doesn't make any sense in an ever-changing global economy.
First, I get there's a lot of people supposedly looking for work but I've spent a solid 20 years now in a "we are constrained on labor" environment across a few industries. I think this is a problem more of what work people are able or willing to do rather than the existence of jobs themselves.
Second, the concept of "means of production" is now not just about the factory the workers work in but the factory that makes things on it's own. We have to accept that if someone else owns something, if they have property rights over it, it just isn't mine no matter how much I want it. If I own a breadmaker machine and it makes me bread anytime I want, I still don't owe anyone my bread no matter how much they want or need it. They can ask. They can't take.
A person might ask "why can't they take? what goes wrong when they have a right to your bread?" The answer is that without strong property rights, we fight over who gets to own eachother's stuff instead of making new stuff for ourselves. Making new stuff instead of trying to take someone else's stuff is how we have more stuff. And before anyone says we don't need any more stuff; If you don't need any more stuff, don't take anyone else's. I disagree; I want more stuff. I want a space ship. But I'm not going to try and take bread from another or a space ship from another. I'll have to find a way to buy or build my bread or space ship.
Except the strawman of abject simple property rights falls apart when you acknowledge that taxes, especially sales / capital gains / etc, are the absolute and certain forfeiture of your property - your money - to the government, to almost exclusively be given to others. Societies are erected on a foundation of the seizure of private asset for public benefit. If you own the breadmaker machine, what is the difference between the seizure of bread and a 100% tax on your revenues of said bread?
Back in the 1930s there was pancake-making-robot outside the diner across the street from the apartment where my dad grew up on 72nd St in Manhattan. My dad loved to watch the thing work. He'd put in a penny and a mechanical arm would come out and pour some batter on a hot grill, then a few moments later another arm came out and flipped the pancake, then a few moments later an arm would come out and move the pancake to a paper plate, which then got shunted out to my dad. Sometimes my dad would put in a penny when he wasn't even hungry, he just loved the elegance of watching the robot move.
I often wonder why there are no such robots in Manhattan now? I have to assume this is simply a matter of fashion. Robots seemed very futuristic in the 1930s, and even something as simple as a pancake making robot helped a little boy feel in touch with the future.
I often wonder why there are no such robots in Manhattan now?
As someone in the food automation space, I would suppose this is probably some combination of: (1) Cost to custom-manufacture in small volumes, where small < 1000s. (2) Patent encumbered. (3) Food safety regulations. (4) Low output volume. (5) Reliability not 100% (eg. fails in winter temperatures) (6) Cost of site-call engineers to resolve semi-frequent issues too high. (7) Electrical consumption rather high. (8) Potential issues connecting to public water system, limiting viable sites and potentially introducing issues of pest ingress. (9) General novelty / wow-factor wears off. (10) You can't eat it every day.
Your dad's story reminds of my experience with the Krispy Kreme machine [0]. They have one at Daly City [1]. I remember going to see the first Krispy Kreme that opened in Canada since I lived within driving distance from it at the time. It seemed to me to be remaining part of fully automating Homer's Donut Hell [2].
Thank you for reminding me about this! I remember being mesmerized watching a few donuts go all the way through every time we stopped by when I was young. Sadly that location has been closed down.
Interesting theory in the link that they died off in the 70s because coins were no longer worth enough to buy a decent amount of food and bills couldn't be accepted by machine yet.
That is interesting, but IMO what's interesting is that USD coinage doesn't go as high in value as anywhere else I'm aware of.
(Continental) Europe has €2 coins, worth just slightly more than two dollar 'bills', here in the UK we have £2 coins (worth slightly more again) (and commemorative £5s), Canada has 'twonies' too.
The 50¢ top-out seems surprisingly low now that I think about it. Is there any particular reason it's developed that way?
A decade ago I remember being surprised that they were becoming a "thing" -- getting them as change, might have been from vending machines too -- but now I realize I haven't seen one in years.
the ticket machines for the patco train in NJ to philadelphia still give them as change i believe, although i havent used them in years.
Before they added a refillable card, i hated getting a heavy pocket full of golden "pirate money" (sacajawea's or presidential dollar coins) buying a $3 ticket with a $20 bill, if i forgot to bring smaller bills.
Lots of hotels have pancake 'robots' ('machines' is probably a better word, they don't have arms or anything, just look like a big toaster) at their breakfast buffets.
Some have waffle irons, but they still require the user to put the batter in them. The pancake machines that OP is talking about automate the entire process of making a pancake (and don't require flipping the pancake). Here is an example: https://uncrate.com/chefstack-automatic-pancake-machine/
Not saying this is the reason NYC doesn't have pancake robots anymore, but it seems like New York in general has banned certain types of automation in order to preserve jobs. For example, that's why you still see - in 2019 - manual toll booth workers sitting in glass boxes in NY instead of automated tolls like every other state has.
Something about that seems wrong. I definitely respect the fact that someone will do the job, but isn't that basically the equivalent of a universal basic income except you're forcing the person to waste half their day doing trivial work?
As someone from a city covered in tolls, Richmond, the workers serve what may seem like an insignificant but very important role. They provide change and do so quickly.
Atleast in Virginia, the attendants are usually off to the right side with the other slower traffic. They have the change available right away, usually in several different sets for the most common ways of paying. I believe the big bucket machines used to provide change back when tolls were well under a dollar, but they would just overfill since most people didn't even know they could get change back and were just mad they didn't have correct change. Some kind of vending machine style system of people trying to insert dollars on the highway would just cause more problems.
Its been especially awful recently when they've tried to remove them. EZ-Pass has started to become fairly common, but so has the ubiquitness of debit cards and credit cards so people are lacking cash and change more often. The automation on tickets has increased to near perfect levels, so when someone doesn't have change on hand they end up trying to back up and cross over several lanes to the single one or they root around in their car for minutes looking for any change when there are no attendants.
Some California toll roads still have manned booths (others have been 100% automated for many years). It's not a matter of providing jobs; it just costs a lot to put the infrastructure into place so it doesn't happen overnight or all at once.
> For example, that's why you still see - in 2019 - manual toll booth workers sitting in glass boxes in NY instead of automated tolls like every other state has.
Does Kansas have the same protections? Plenty of humans in booths on the Turnpike.
I think its more likely they just don’t fire a few hundred people on the few bridges and tunnels in from NJ that probably pull in millions a day. All new bride tolls I’ve seen have cameras and toll processed do not stop signs.
Not sure if this law exists in relation to toll booths but there are some laws in place that do seem to be geared towards what you describe such as being unable to pump your own gas.
I'm not a NY resident so I don't know. If not a law, it's possible labor unions push back against automation, allowing their union members to continue to be toll booth collectors
> I often wonder why there are no such robots in Manhattan now?
It sounds like it only made one pancake at a time? A person stood at a griddle can make a dozen pancakes at once. And we don't know how often that robot broke down. I somehow suspect the novelty was one of the reasons to have such a device, not necessarily the efficiency of the thing.
Nit: unless the video is running at a reduced speed, the claim is false. You can clearly see that the pizza takes 18 seconds to pass over a single location. If you were to run then pizzas through edge-to-edge, then the fastest production rate would be (3,600 seconds per hour) / 18 seconds = 200 pizzas per hour. But that would risk having the dough from one pizza get stuck to the dough of the next. Adding a sensible 2 second gap between each pizza would reduce the throughput to 150 pizzas per hour.
Depends on the style, the consistency of the sauce, and the dough, how much you can get use without making a mess. That looks way on the high side to me, though their sauce might be very dry (thick) which might let them get away with it.
Normally on my thin-crust pizzas with a somewhat watery sauce (what you get when you use fresh tomatoes, or canned crushed tomatoes, or anything along those lines without mixing it with lots of paste) you can see about 50% of the dough through the sauce, it's that thin, but if I make it much thicker the toppings get very... slippy-slidey.
Making them round is so inefficient for the machines. Make 'em long -- that way you merely need to pass dough through a conveyor. Then, use the money you saved on the simpler machinery to do an advertizing campaign to change the public's perception of what shape a pizza should be.
The boxes are already square. Two long rectangle pizzas in a box. As a bonus, it's easy to pack two DIFFERENT pizzas with different toppings into a mixed box, for group orders. And variety.
I'm with you on the love of square pizzas. You get about 12% more crust on a square pizza versus circular (for the same area.) But unlike circular pizzas, a square pizza gives you at least 3 different crust configurations for different "kinds" of slices in the box. So both those who want more and less crust ratio can be satisfied.
Actually, in Italy the style of pizza people eat most is baked as a big square piece and cut into rectangles. These are typically eaten at lunch, not dinner. Italians don't really eat pizza at dinner (as I've been told by an Italian colleague).
As for the pizza in the machine video: it looks downright horrible, and only acceptable when compared to what goes for the average 'US-style pizza'. I'd much rather eat a frozen pizza from the supermarket, or something else entirely. Don't see the point of have 'more customization options' compared to a frozen pizza created in a factory, if the quality threshold of the pizza is so low in the first place. But maybe I'm just being a food snob ;-P
Detroit-style and some others (I think there's a thin crust square shape that gets claimed by various other midwestern cities?) are already rectangular.
It's kind of a ridiculous metric anyway. There's probably only a handful of pizza shops in the world, at their peak times, that sell 300 pizzas an hour.
I make lots of pizza for fun with some friends at hacker events. We've managed to do ~120 portions (not individual pizzas) an hour with an entry-level pizza oven (2 levels, 60x90cm² area), the kind you see at kebab shops, and the oven was quite far from being the limiting factor (rolling the dough is hard work). From a quick web search, conveyor ovens can do up to 150 pizzas per hour, and you can typically stack them three high. So shouldn't be a problem.
The trick is that pizza is baked at a very high temperature. Our oven goes up to 500°C (900°F). Baking time is 90s-2min if you put the pizza on the fire stones directly, or ~3mins if you use a screen (tray).
I'm sure you could have said the same thing about cars before the production was heavily automated, or even cell phone manufacturers before they went all-out mass production.
i.e. if they can make more per hour, they'll sell more per hour.
If I call up a pizza joint and they say it will take one hour to get my order to me, I might call someone else instead. Improving their throughput could absolutely generate more business.
It also doesn't include the dough preparation or the cooking, although the latter can be done by connecting the output of this to the input of a conveyor oven.
Perhaps it takes into account the % of pizzas which have a certain number of additional toppings. Presumably Margherita pizzas can take much less time?
I feel a bit stupid, but does the product look really bad? Why are they hyping it? I thought it would compete with restaurants, but it really looks like a frozen pizza. Weird distribution of ingredients, weird dough without a real crust and not properly baked (compared to for example a pizza baked in real a wood oven or from from a electrical pizza oven). It certainly doesn't look like a normal pizza.
You touched on a pretty good point there. It's highly likely that this startup just reinvented the wheel, and people like DiGiorno already operate a superior version of this machine.
I remember seeing several videos from different companies with their own version of a "pizza robot". Just look at this ad for a pizza vending machine from 2012 https://www.youtube.com/watch?v=Pyrav_9Pbsc
That vending machine pizza maker doesn't get it right either. It mixes the dough "fresh"... but that's actually not ideal. The dough needs time to absorb all the water and time to sit out and be proofed before made into pizza.
Pizza seems simple... until you want something that tastes really good...
If the output is frozen pizza quality, wouldn't it be cheaper to just buy frozen pizzas directly from the pizza factory, and then install ovens instead of a full pizza making machine?
Even though it looks the same, it probably won't taste the same. A lot of the issue with frozen pizza is that it is frozen, which means the ingredients are not fresh. These ones could theoretically have fresher ingredients. That will matter a lot when it comes to the dough.
I mean if the ingredients are brought directly from the farm to the factory, put on dough, and frozen. The only difference would be the dough.
The actual ingredients may be more fresh than if a random pizza joint bought them from a grocery store that picked them too early, shipped them half way across the planet, let them ripen in the back, brought them out, got bought a day later, left them hanging around for 5 days at the pizza store fridge before cutting them and putting them on your pizza 8 hours later.
If you are going to make a decent pizza, you are going to need proper dough, not a preformed crust. The dough will need to be worked and tossed or rolled out. The time it takes to put on sauce cheese and toppings is inconsequential wrt the time it takes to work the dough.
I am with you. I cannot for the life of me imagine who would want this machine. Companies that mass produce frozen pizzas have industrial processes that would put this gizmo to shame in terms of consistency and output volume. For restaurants, this machine looks like it would require so much maintenance to keep it working and sufficiently sanitary for use in a commercial kitchen that I am not even sure that it would offer any labor savings over a traditional pizza operation, and that is before you consider the fact that those pizzas look completely unappetizing.
It's just adding toppings to pizza (and too much sauce!)... I was hoping to see how the dough preparation would be automated. I'm sure it could be done, but I would prefer a hand-made pizza. I enjoy making my own pizzas, and the dough/crust is what makes the biggest difference in a pie, and is the most interesting/fun to learn and experiment with.
this robot to me is just the next level up from the skittles sorting robots.
Yeah am I the only one who thinks less is more when it comes to sausage on pizza? And smaller crispy bits not big grey chunks that are barely warm in the middle please. This pizza looks gross.
The line between robot and "regular factory machine" has always been fuzzy, and gets even fuzzier as time passes - and more automation from robots back-propagates to things that are "just" machines.
Pizza on the blockchain! Tired of your orders getting made wrong? Just put them in an immutable, distributed database, and you're guaranteed perfect pizza every time! But don't eat that flawless pie, because our upcoming SausageCoin™️ ICO guarantees its value will only go up, up, up!
Consumer (and I assume commercial) bread machines often have a "dough only" cycle that you can use for automating your dough process. But I'd like one with a water filter and more granular control over the germ time, temperature, rise time, and knead strength/time all in one widget. Something with multiple wet/dry hampers and a programmable cycle would be really cool too.
Domino's (yes maybe not everyone's idea of perfect pizza) has a good video explaining how their dough is made. They have a production line that mixes, kneads and balls the dough, then it just needs to be placed into trays by a human. It is shipped fresh to restaurants, so proofs in transit.
> More than a third of restaurant owners are having trouble filling jobs... And more than 80 percent of workers will change jobs each year, requiring employers to constantly train recruits.
Maybe pay people more and treat them better?
But I can see the appeal of restaurant automation. Machines can work 24/7, don't need health insurance, don't fight for higher wages, don't need to be trusted around cash registers, etc.
One thing that always staggers me in endless discussion is that nobody ever, ever talks about decreasing profits as a way to fix societies problems.
I mean, in 2018 McDonald's returned $8.5 billion to shareholders through share repurchases and dividends, it announced a 15% increase in its quarterly dividend to $1.16 per share. [1]
I was initially skeptical, so I went looking for some numbers. If the numbers I pulled off of Google are right (and I did the math right), that's something like $850 per employee every quarter. I've got to say, that's not nothing, especially considering many employees are in countries where with lower prevailing wages than the US.
Numbers:
Shares outstanding 768,000,000
Employees: 210,000
Percentage of McDonald's that are corporate owned: 20%
Dividend per employee: ($1.16 * 768,000,000 * 0.2) / 210,000 ~= $850
What have the workers done to "earn" this? They literally signed an agreement to work X hours for Y amount of dollars. They said yes to that. If they wanted a piece of the equity -- the profits, they should have negotiated for that in their contract.
McDonald's and other similar companies are having a hard time convincing potential employees to sign the existing agreement. If they want to convince more people to agree to work for them, one option is to offer more money, which McDonald's at least can certainly afford.
Or, more snarkily, what have the investors done to "earn" this?
About $4,400 per year per employee globally. A large amount of money to most McDonald's workers in the US and an life changing amount of money for a worker in say India.
I completely agree that if you're offering a minimum wage hourly job with no benefits you don't get to complain about difficulties finding staff and retaining them. They have tons of options at that price point if they want to switch jobs. If you want to avoid having to constantly deal with turnover you need to offer more.
Yeah, it's funny how the same people who say they're desperate for more/more reliable labor will also tell you that paying their workers better is laughably impossible.
In some businesses, the customers are very price sensitive, so if you can’t make a product with very discernible qualities, the customers will opt to pay less.
This is the situation in restaurants, where most purchases are not going to be willing to pay extra for food to so that the cooks can make more money, because a competing restaurant with acceptable food is almost always available at a cheaper price by not paying the staff as much.
Training any new staff in any industry is a hidden cost; difficult to quantify. We know it in the programming industry as the spin up time to get comfortable in a given code-base - at least 6 months to be properly comfortable, possibly as long as a year to be fully comfortable. The owner of Books Are Magic in Brooklyn † worked to transform his little corner of the industry, reforming as an alcoholic and his restaurant to boot. His reformation lead to a healthier staff that went to bed early and treating themselves better as well. The owner paid for his staff to go to therapy and live better lives themselves. I don't know if its replicatable, or if lower staff turnover and happier staff makes all the different there, but restaurant/bar work is, well, horrible for the soul. It doesn't have to be.
The video is missing the stretching the dough part, which IMO is the most time consuming part. Spreading the sauce / cheese and putting toppings is definitely the "easy" part of the job in my experience.
If I had infinite time to invent my ideal automatic pizza machine... I'd personally like to see the full high-quality bread cycle done on the dough. That is: ~1 hour for the bread to rise, punching down the dough, stretch the dough out onto its final form, ~1-hour for proofing the dough (maybe 20 minutes at 90F, since yeast reacts faster at that temperature), and THEN automatically spreading cheese on it.
Most of this time is just waiting by the way. But maybe some kind of "jukebox" design where a robot can work on a Pizza, automatically put it on a shelf somewhere (for the ~1 hour rise / proofing steps) and come back to it automatically would be great.
At least, when I manually make pizza, getting high-quality bread dough techniques on the pizza dough makes a difference in texture (chewiness and "evenness" of internal air bubbles).
A lot of Pizza shops don't care about the consistency of the dough however. They just pull the dough out of the frige and immediately make a pizza on it without the punchdown or proofing steps. And you can tell because of the "air bubbles" that often grow huge and break the consistency of the dough.
------------
EDIT: What is going on here, is that yeast is a living organism. Pull that dough out of storage, and the different yeast will reactivate at different times. The edges probably will have more active yeast, as it warms up faster).
When the dough enters the oven, it rises to ~110F or so first. The yeast highly accelerates and starts to produce carbon dioxide as it eats the sugars. But the "pockets" of highly-active yeast will make more CO2 than other areas of the dough.
Of course, the yeast dies by ~160F and the "puffiness" of the bread is then solidified by the heat of the oven.
That's why you punch down dough: so that all the (reactivated) yeast creates CO2 bubbles at an even rate across the dough.
------------
In practice, the Pizza is going to be limited by the speed + space of your Oven. I'm not sure if people need "faster" low quality pizzas, but improving the automation to create higher-quality pizzas should be the focus.
I worked in a Costco, making pizzas all day. The stretching part was achieved via an oiled hydraulic press. Two big ol' plates of metal squished the dough out to pizza shape.
You'd transfer it to a metal grate and then sauce, cheese, and top it. Operation was pretty quick, as long as the dough was in the right condition.
Large air bubbles come from over-proofing not a lack of it.
I've wanted to make an automated pizza factory since leaving the business in 2015. Nobody seems to think it will work in my network, but I'm certain it can work.
A real automated pizza factory would take in raw ingredients, cut wash and serve them individually for each pizza. Self-clean everyone of it's own tools/arms and allow individually addressable toppings or as close as can be whilst having a speed capacity to match the highest order count. Shops near universities or financial districts can get insane orders.
If I were in the correct position I would go whole hog on the idea simply because I like it. I think with the range of robots out there, there is a large space for wozniak-level improvement on the factory/semi-automated model that most fast food stores currently use.
I wonder under the surface if there is some hidden principle I am missing politically or something. There must be a reason it hasn't happened yet already, it seems so doable.
>There are a few details that may save Picnic’s pizzas from tasting as if a robot made them. For starters, the dough preparation, sauce making and baking — the real art of pizza — is left in the capable, five-fingered hands of people.
The dough making & stretching is still done by humans.
It looks like this robot distributes sauce, cheese & toppings (it only shows one topping, though). Then cooks it.
Oddly enough, it doesn't even do the cooking. The bit you cited says the humans do the baking. I'd have thought that'd have been an easy bit to integrate - places like Pizza Hut already bake on readily automatable conveyors.
Do you like uneven bubbles in your dough? Because that's how you get uneven bubbles in your dough.
Rise -> Proof -> rise again is the only way to evenly distribute the yeast. And you have to do it immediately before baking.
The dough could probably be mass-manufactured in a factory somewhere. But dough isn't very hard to make, and flour is WAAAAAAYYYY easier to store than live dough.
Would you rather be storing flour at your kitchens (which stores basically indefinitely at room temperature)... or would you rather be storing dough at your kitchens (which goes bad within 2 or 3 days and needs to be refrigerated). EDIT: It isn't so much that the dough "goes bad" as much as it "turns into sourdough" after a while.
---------
Hmm, a lot of commercial kitchens (ex: Auntie Annes, Costco) clearly use premade dough mass produced somewhere. Maybe they mix the dough upon receipt or something.
The main issue with bubbles is their inconsistency. One pizza may have huge bubbles, while another one may not. The "chewiness profile" across the dough changes, so its going to make quality control a bit harder.
You can make a dough chewier or crispier by adding more (or less) water, yeast, salt, sugar, etc. etc. Ideally, you figure out the "ideal chewiness" and then make it consistently throughout the dough.
>For starters, the dough preparation, sauce making and baking — the real art of pizza — is left in the capable, five-fingered hands of people.
So all of the stuff that actually takes time is still manual. All this thing does is throw down some toppings. And then takes all the time and more it "saves" to clean it properly and reload it.
You have to wonder why we really need pizza to be round in shape but then they package it into a rectangular cardboard box. Like there is a whole part of the machine that takes the leftover dough after cutting the circular shapes. Seems inefficient.
Its commercially taking subconscious advantage of the consumer who sees "big box which seems full" but actually gets "circular portion less some". If you went efficient, your product would be square and look smaller than the competition.
This reminds me of a previous post for the dishwashing robot.
It is automating the easy part. All they are doing is putting toppings and sauce on. An experienced human can do it perhaps a little slower.
Essentially this big piece of equipment saves the business owner 2 minutes of an employee making $13 an hour, about 50 cents. At full capacity, 60 an hour, theoretically it might save $30 an hour. (Practically that won’t happen for lots of logistical reasons).
It better be quite inexpensive to justify that minor savings.
A pizza employee makes much more than 13 hr. There are sick days, holidays, taxes, fees and lawsuits. For 13/hr employee you're probably looking at 20/hr at least in costs.
There might be some fancy pizza restaraunts where the staff isn't hourly, but I don't know of any. Taxes and fees are pretty minimal on a minimum wage employee, so that leaves lawsuits which I have no idea how to quantize.
One big downside to this is that every topping you add requires yet another sliding conveyor belt. The demo here can only make cheese, pepperoni, ham/sausage, and meat lovers pizza.
And just my personal opinion that the pizza that comes out looks pretty damn sloppy. Those sausage chunks roll around too much. If this were my machine I'd probably have it only make cheese and pepperoni and have a person on hand to make other kinds of pizza. Assuming I use this contraption at all.
I would bet you're right - even just doing sauce/cheese/pepperoni is probably 70% of pizzas though.
And you can throw fresh toppings on top of that as a manual last step in 30 seconds before it goes in the oven.
Still seems kinda ridiculous. Watching people in pizza places... seems like most of the time is spent stretching the dough to be pizza shaped and appropriate thickness without tears, which this contraption doesn't solve.
Once that's done, people sauce, cheese and top pizzas about as fast as their ovens seem to work.
Judging from https://youtu.be/Do7qLGiVbwY?t=164, the advantage is that they have 3 pizzas in each row rather than just one, and that the pizzas move much faster along the conveyor. It's also not clear from the article but the Seattle system might only support 1 pizza in the system at a time.
The dough is still hand-made, which is an advantage over frozen pizza dough which pretty much sucks, so I'd guess they're primarily interested in selling to delivery pizza chains as a more upscale pizza. Domino's peak is 160 pizzas/hour and they open stores until the average is <100. (https://www.quora.com/How-many-pizzas-does-a-Manhattan-pizze...)
I was confused about this too. I think the win here is that the machine is small enough to actually fit in a fast-food restaurant. Plus, they're planning to branch out from pizza.
I assume such factories employ machinery that isn't suitable for use elsewhere. Additionally, it seems unlikely that such machinery would have been designed to produce individual customized pizzas on an as-needed basis, whereas this seems built specifically to address both of those points.
I'm no pizza connoisseur, but the pizzas in that video don't look particularly appetizing.
> That speed comes in handy in places where large numbers of orders come in during a rush, such as at a stadium or in large cafeterias
In other words, their pitch isn't to be appetizing. It's to crank out a lot of pizzas for buyers who aren't being picky at peak hours when you can sell just about anything.
> There are a few details that may save Picnic’s pizzas from tasting as if a robot made them. For starters, the dough preparation, sauce making and baking — the real art of pizza — is left in the capable, five-fingered hands of people.
So it's fundamentally just a sauce and topping sprinkler?
This seems slower than the manual speeds I observe at the place I go for lunch slices (Big Mario's, on Pike in Seattle), cooks there just use the back of a ladle to spread the sauce in a couple seconds. Per the video the robot takes 10 seconds, plus it leaves the sauce in lumps.
Video of Mario's process (they don't show spreading the whole pizza but they do at least half in the second or so they do show: https://youtu.be/tXMz-DxC_dk?t=163
I might have missed it from the article but it doesn't seem like the machine "spins" the dough to get air bubbles into the sides and make a proper crust.
The product on the photo looks like a frozen pizza from a factory tbh..
What is the innovation here? The fact that it is inside a restaurant?
I wonder what are the social implications of these types of machinery.
On the one hand, cheaper means more people can afford it, thus more inclusion.
On the other hand, pizza making is not such a harmful job, comparing to let's say mining in which a robot actually saves lives, to justify substituting hand labor. Is it actually leaving people unemployed?
I guess places where cheap food is sold will continue to sell cheap food, so instead of underpaying someone to do the job, a robot does this, leaving people free to do more intellectually demanding jobs.
I guess the real question is then, how can we reallocate people and give them better, more meaningful jobs and not leave them helpless.
Don't get me wrong. Any form of job is respectable. There is always going to be demand for higher quality versions of the same product, where, hopefully people are not underpaid. I mean better job in comparing to poor working conditions.
In the end all the "manual/hand labour" will/should be replaced by robots. Which one are automated first depends by the difficulty of the job and economics. Unfortunately there will always be low skilled and high skilled jobs. We already have low skill jobs in the IT(i.e data entry)
crappier pizza, fewer jobs, fewer pizza chefs. boo. The past few years I have been making my own pizzas and think back to when I would eat frozen pizzas and just shake my head.
At first I scoffed, because making good pizza is essentially the same as making good bread, which is more of an art than a science.
After reading through, it became apparent that this isn't geared at good pizza, it's focused on bad pizza where none of that matters, like cafeterias, gas stations, and football stadiums.
You know, I completely disagree with the your first sentence. Of all cooking techniques, baking seems to me to be the most scientific. You must carefully control temperature and moisture in order to get a decent loaf. It's practically an exercise in chemical engineering once you start adding things like chemical leaveners.
I thought about addressing that, as there is a huge scientific component behind baking, and if you could control every variable perfectly, you would get exactly the result you wanted.
However, reality is there are wayyy more variables than what you can find in a recipe, even a good one. And that's to say nothing of the techniques of master bakers. Compare a what a master does and what they're able to write down or tell someone and it's an enormous gap. That's where the art is.
Many arts use science. But when people say "it's more an art than a science", they mean something like "the production cannot be reduced to a simple algorithm", typically because the space of outcomes is large, has many dimensions, and doesn't have clear success criteria.
Chemical engineering? Isn't mixing chemicals just chemistry?
Regardless, there is no standard for baking, each little extra bit of yeast, sugar, etc. you add in can make it an entirely different product. This gives it its own personal craft aspect, hence why it can be an art.
Those are different recipes, but I'm not sure that's where the "art" comes in.
The challenge of baking is that the components aren't very standardized. You won't get a consistently good result by robotically mixing 1000.0g flour with 7.00 g yeast and 650 mL distilled 35˚C water, and mixing for 400 seconds. The flour varies from bag to bag, even of the same brand. Yeast is alive and responds differently depending on the temperature and other conditions; sourdough cultures are even more finicky.
As a result, there's an "art" to knowing when to add more water/flour or deciding when the bread is ready to shape and bake. I would bet you could get this more under control in a commercial/industrial setting, where you've got proofing cabinets that control the ambient temp/humidity and other specialized equipment. At home though, you've more or less got to do it by feel.
Clay works must carefully control kiln temperature, baking time, glaze production/composition, cooling time, clay consistency, room humidity, drying time etc. to prevent a clay work from going wrong. This doesn't make sculpting a science- you still must shape your object, in any shape you desire, with your hands, and express based off structure and pattern and paint and color theory or lack of colors. The end creation is still a piece of art.
When it comes to baking, you are still shaping, structuring, patterning, coloring(or not coloring) your breads.
I wish someone automated electricians and plumbers instead. Most of those guys charge an arm and a leg and are booked months in advance. I spent a month trying to get someone to come in and install a NEMA 14-50 outlet in my garage. Automating $15/hr jobs is just weird. The people who work in those jobs usually also man the cash registers, make the dough, clean up, take calls, etc. You aren't going to replace something like that with a glorified frozen pizza maker, even if your investors think otherwise.
I was in the SF airport the other day and I tried an automated coffee dispensing machine that actually tasted really good. Way better than the old machines. I think it was called Briggo. You connected to it with your phone, and you made your own custom "barista" style coffee profile.
I can imagine this sort of thing replacing drive-through coffee shops, especially here in Canada where drive-through coffee is ridiculously popular, and the lines are notoriously long.
300 pizzas an hour sounds like quite an achievement. But are they any good?
The quality of pizza varies so much from one restaurant to another that it doesn't matter how many pizzas it makes of they're not any better than human-made pizzas.
Sure, robots may lower the cost of a pizza. But the majority of pizza consumers will choose to pay a little more for better pizza. Little Caesar's tried the uber-cheap but garbage pizza model years ago and eventually had to improve its product.
The challenge is not to build a machine that makes pizzas but more:
- individual toppings
- small enough to fit in a pizza shop
- not too expensive
everything else seems solved by frozen pizza manufacturers (though there is a suprising ammount of hand-tossing ingridients involved, morzarella seems to be especially hard :) ) https://youtu.be/DTnevo7XPqA?t=90
I want a robot that can make arbitrary meals, not one which can only make one type of dish (these already exist in factories in various forms and are nothing new).
And when you think about it, a robot cook would save more time per day than a self-driving car.
One cool thing you can do here (old idea of mine) is that you can allow customers to fully customize their pizza's toppings. Imagine an app where you can "draw" your pizza and then get exactly what you ordered.
> We were using frozen dough and an electric oven for the demonstration
The main reason I go to a pizza restaurant is for the fresh ingredients. A lot of the reason I hate frozen pizza is because the crust tastes like cardboard.
Is there a place where people discuss societal theories about (in an ideal future) entirely workless life ? Not that I'm for it or even convinced it will hold but I'm curious about the subject.
in the video, does each little conveyor do one ingredient?
what about food allergies and special dietary requirements where you can't use tools for e.g. dairy/meat to process vegetables
I hate to delve into this controversial area in such a throwaway story -- but...
The cost of labor (and living) in this country (California in particular), is making everything terrible. And at the root of it, it's driven by people with good intentions piling on restrictive regulation after restrictive regulation about housing, safety, liability, etc.
Everything that you might choose to complain about stems from the high cost of employing anyone -- and related to that (I genuinely lament), how much it costs to barely live around here.
BART and buses suck? It's because we cannot hire enough station agents, police officers, maintenance staff, etc to properly man the system.
Your commute sucks? Because you have to live so far away to afford anything reasonable in rent or mortgage.
Restaurants suck? It's because again, labor is expensive, land is artificially expensive because of restricted supply.
Pizza sucks? It's because you can't open a pizza place (except staffed by underpaid immigrants) at any reasonable cost, and instead have to resort to robots that can't get a crust right.
This is a major problem -- our country is not getting so much older like some countries, yet our costs are in runaway mode. Something is wrong with this picture.
We are rapidly approaching a future where the entry level jobs like fast food / picking crops / warehouses can be automated away and even well paying ones such as truck driving / taxis could disappear or scale down drastically. We really should be having an urgent conversation on how to handle this and what the responsibility of those that own the means of production will be (if any) when this occurs.
Yang is obviously speaking about this and suggesting a BI as a band-aid / solution but I do think this should be getting a lot more attention that it is.