And it's not just blue collar workers that are a target of this idea. We are already starting to see automated programming moving out of the research lab and into the commercial realm, e.g. GitHub copilot.
One of the great improvements in the second edition of FORTRAN, FORTRAN II, is that an application program can be written not as the output of a single compilation, but of many separate compilations.
[The above is liberally quoted and reorganized from the IEEE Annals of the History of Computing, 6(1), 01984.]
As soon as GitHub was acquired by Microsoft I knew their intentions were for automated coding tools. I wasn't concerned about this affecting my livelihood in the short to medium term, because state-of-the-art ANNs won't be able to grasp the context of business requirements without developing adult human level intelligence. Thus, even a full program built by such a system would need to be verified by a human to be sure it will behave as desired, obviating any gain in terms of automation. Even the rudimentary boilerplate that copilot spits out suffers from this problem.
I feel like the end result is just what we already have, game engines with visual programming and drag and drop editors. You don't add much value by automating the programming since you still have to define the input and outcome.
The real win was building the complex logic and state editor into a good UX.
I think we'll get to genuinely open ended sand box games.
Have a look at the video, it's quite impressive.
Even just watching the video, I came up with several possible improvements to try out. Eg adversarial training, that would really hone in on the situations and aspects where the model is weak so far, like edge conditions; instead of just using normal gameplay as input.
The other part about that GAN Theft Auto example is that it doesn't actually know what's going on, like there's no game state. All it knows is that "When I have a frame that looks like this, and they press that button, I think the next frame would usually look like this". So it's got no internal game logic, it's just really good at painting what games look like.
Even going about this very naively, you could at least use it to train a model against a supercomputer running the game, and then run the inference on much more modest end-user machines.
But you can be much more ambitious: have you seen eg style transfer? So you could probably do a bit of ML black magic to train your model on GTA, and then point it at the Google Earth data to get a GTA-like set in real-life London.
Or you could use something like style transfer to go for a cartoony look, or add ray-tracing like effects, even if you didn't have these effects in your original engine.
Or you can use a pre-trained model (eg on GTA), and then spend a relatively modest amount of extra training to get a different kind of game, eg one that has magic or so.
About the latter part: I do think their model is already running with some state. But even if it ain't, that's a relatively small thing to add with already known standard techniques (or you can come up with new techniques.)
I realise that you could define everything as data - laying a brick, you take the inputs of where to position the brick, etc. However I think we can make the distinction between chess where the data is "Pawn is on e4" and the much greater complexity of the real world where we are dealing with billions of atoms. Perhaps not everyone agrees with me.
It's fairly sucessful. We can simulate for example driving a car really well.
Simulating human behaviour is harder, but simulating brick laying is not that hard, we have the technology to do so already.
Simulating brick laying might be able to be done in a controlled environment, is it possible to make it low cost enough and accurate enough for all general purpose brick laying situations? Probably, given enough investment we could get closer. Is ironing out all the nuances cost effective? I don't know.
We can definitely simulate a driving environment but I given the recent struggles of self driving cars I don't think I would say that we are at the point where we've solved the problem of actually driving them in everyday situations.
Can you qualify this more specifically? In many domains (particularly safety critical ones) “reasonably well” may not be sufficient
But what do you mean by 'across the board'?
As far sources, a web search gives many articles about safety of self driving cars. See eg https://www.wsj.com/articles/self-driving-cars-could-save-ma...
Anyways:
> But what do you mean by 'across the board'?
By that I mean across all of the people driving in the US and their rate of incidence. For example, average miles driven, the amount of drivers and the rate of accidents. I think the most intriguing detail could be drawn from the rate of fatal accidents, since that's the most concerning, ignoring accidents that cause a casualty as I don't know the method for gathering that data off hand. One could glean a lot of info from that. Here's some rough numbers I gathered, and please forgive the naive approach to my data gathering to express a point:
Average miles driven/person[0]: 13,000 Average fatalities/year[1]: 37,000 Approximate number of licensed drivers[2]: 231,652,000
I don't have numbers for self-driving cars and the number of accidents, but regardless, would it perform the same with the same number of miles driven per car. Keep in mind that self-driving cars currently aren't navigating in all circumstances and will beep to make the human take control again. At least with Tesla.
[0] In 2019, there were almost 229 million licensed drivers in the United States. (Source: https://www.asirt.org/safe-travel/road-safety-facts/) [1] Over 37,000 Americans die in automobile crashes per year. More than 90 people die in accidents every day. (Source: https://www.thewanderingrv.com/car-accident-statistics/) [2] https://hedgescompany.com/blog/2018/10/number-of-licensed-dr...
Looking at fatalities would make the analysis easier, but when I last checked, Waymo hadn't driven enough miles to make a good comparison possible on that metric.
(They haven't killed anyone yet, but neither would the average human driver have done so, yet.)
So we would need to look at less dramatic accidents.
See https://www.forbes.com/sites/bradtempleton/2020/10/30/waymo-... which tries somewhat to correct for conditions.
[1] https://www.automotivetestingtechnologyinternational.com/ind...
Every step in programming tooling since programs stopped being input as manual hardware configuration has been automation of programming, and its just made more work, and higher paying work, for programmers.
The details of job changed, as more low level pieces of it were automated leaving the higher-level, more abstract bits, but for most of the changes, while some either bailed for other work or rode the dying embers of the old way to retirement, programmers generally adapted.
The funny thing about automation in programming - it has always opened up more doors - and led to more employment in programming.
We are so far from anything that resembles real automated coding - that coding automation should continue to be celebrated by engineers for a long time.
GitHub Co-Pilot and VS Code aren't going to replace you - they're just going to let your company offer a better product, release a new version sooner, test different versions, etc.
The closest a product that ever did this was Excel.
Domain experts are all using Excel, not any other automated tool.
I just hope we could improve on it.
See eg https://www.microsoft.com/en-us/research/podcast/advancing-e...
Any language any human would touch these days automates lots and lots of things for you.
(Even Assemblers do a lot for you nowadays.)
And on a casual level, nothing changes if the best in the world is a human or computer, so the casual player isn't particularly affected.
Not sure how important Netflix is here? There are lots of chess players on YouTube, too.
Go also get much more popular in the West for a while when Alphago came along.
By comparison, brick laying is trivial.
Brick laying robots already exist. This isn't a problem that can't be solved, it's a question of making it economically viable.
But the future is pretty obviously going to be skilled machinery operators overseeing the automation.
Even more interesting its once you automate like this the constraints start changing: i.e. it's easier to have a robot lay bricks with epoxy then cement, whereas a non automated work flow would struggle.
There's a Perth based company which has a prototype which basically will layout an entire house on a concrete slab via a boom arm that uses this approach: pallets go in, structured bricks come out.
I agree. It will happen eventually, but the bigger point here that is more related to to discussions on HN is that despite a lot of enthusiasm, machine learning and AI are still in the amino acid phase of evolution, and everything still sucks.
It works poorly, therefore bricklaying has been automated. The programmers won. The bricklayers won too. It's only Grakel with his weird bet that lost.
For robot bricklaying, you can depend on the electrical supply. Everything else is a toss-up. I do think it's possible, but you don't choose your working environment or the weather conditions so everything is gonna be a lot more hassle.
1. According to the BLS link below it's actually closer to 50,000.
There are over 50,000 bricklayers in the US according to the BLS[1].
There's tons of "no code" solutions - but these are all just different versions of coding and programming languages - usually with GUIs and pictures instead of words.
There's more than 60,000 SWEs that work in my company. Amazon, Google, Apple, and MSFT could make >$10Bn/year each if they could automate coding.
The fact that none of them are even trying - when it sort of goes with their core businesses - should give you some indication that this is something unlikely to be automated any time soon.
The reason bricklaying ISN'T automated is probably because there's ONLY 59,000 bricklayers in the US (a $3B market), it's not generalizable, and even if it was - the cost to move a machine, set it up, and maintain it - it's hard to imagine massive cost savings - 50% seems generous.
If there were >1M bricklayers - and/or they made >$400k/year on average, the work was generalizable, and automating would bring huge cost-savings that could be captured - there would be a LOT more effort into automating bricklaying.
But none of these are true. How much of brick laying is generalizable (building a new house) vs custom (repairing some old wall with non-standard bricks)? I have no idea - but I'm guessing not a lot more than 50%. That's a $1.5B market. You'd be luck to cut the costs by 50% - that's maybe $750M.
It could easily cost more than than to automate bricklaying! Why even try?? No one is interested in investing in that risk / reward.
On the contrary - most of Radiology could be automated and most of it is generalizable. Since hospitals are monopolies and healthcare is a mess - you might be able to capture all the cost savings - which would be close to 100%!
Since there's ~35k Radiologists, and they're some of the highest paid workers in the world - there are a lot of efforts to actually automate this (and they're doing quite well).
If you think automating radiology is easier than building a hamburger-cooking robot, you're naive. If you think AGI is easier to achieve than building a tomato-harvesting machine, you're clueless.
There's just simply not hundreds of billions to be made automating cooking hamburgers and harvesting tomatoes more efficiently. And it's not easy to generalize and automate cooking or harvesting EVERYTHING. And even if there was, restaurants and groceries are commodities - not monopolies. You couldn't capture all the savings. A race to the bottom on prices would eventually just pass the savings on to the customer - not juice profits. That's not something you want to spend massive, risky R&D on. That's why we haven't automated cooking hamburgers and harvesting tomatoes. Not because it's harder than protein folding, fusion energy, or true AGI...
From another point of view - we've had machines that mop floors for decades - and there's still a lot of people employed to mop floors. Is this because mopping floors is incredibly complex and creative? No - it's because people who mop floors get paid minimum wage, and they do a lot of other things, too.
Perhaps a company that wants to display the current Bitcoin price on a screen and let you do currency conversions you can do that in "no code", but then again, a programmer can also do that with code in 15 minutes...
As a tradesperson myself, working in a highly automated part of the metal fabrication process, I feel an urge to say something like:
The only thing standing between a programmer and unemployment is a sufficiently advanced compiler
But only because I great contempt for the hubris on display in these sorts of threads.
At the end of the day though, it's a-little-bit-form-column-A-and-a-little-bit-from-column-B.
We went from not-flying to landing robots on another planet in a handful of decades, who knows what a little bit more processing power and a few more layers of abstraction could bring.
Automating programming definitely seems like an easier target in comparison.
I've never had a printer that didn't work: I've had plenty that wanted the correct offering to the HP website to be made and several hundred megs of adware installed before a postscript file made it to the actual hardware.
Well, car assembling plants are older than Deep Blue, IIRC and they are very complex and "very robotic" in a physical sense.
I would also say that the approaches taken by something like AlphaGo are more interesting, too. Because unlike chess the solution space for Go is too large to simply look at all possible moves.
Is that the same?
There's lots of them on YouTube. Here's one with two different robots.
Quick test: replace any pawn with a queen. See if the bot notices.
What are we if we aren't trying to move forward--solve problems, invent, and improve lives? A few weeks back I came across Hadrian, one of the machines profiled in this piece. The idea of building with higher accuracy, lower cost, greater speed, all while allowing people doing backbreaking labor to do something else, excited me. There was no contempt for anyone.
I like woodworking. I'm currently thinking of purchasing an industrial-grade CNC router. I think it's analogous to robotic bricklayers in many ways. I have to design a piece before I make it whether I cut it with a hand saw, use a table saw, or have the CNC router do it. The reason I use a table saw instead of a hand saw is the same for using a CNC router instead of a table saw--it's faster and provides me with a more consistent and accurate outcome.
If that's not a good enough justification, you should look into how much skill is required to operate a CNC router. You can use different bits, rpms of the router, speed the machine is moving at, entry speeds, depth of cuts, etc. There is a lot of skill involved. And when a lot of that is taken away so I can focus on what matters most--creating beautiful furniture--I'll be so very happy and won't feel the least bit bad about it.
> Some jobs have been automated, but many more have not and most likely never will be.
To read:
> Some jobs have been automated, but many more have not.
Or
> Some jobs have been automated, but many more have not and most likely will take a while.
You would have nothing to complain about.
As often with AI / automation, the things that seem simplest to humans are the hardest to automate and vice versa.
As another example, look at translation: take an average English speaker and ask them to translate Japanese texts. They will take lots and lots of training to get good.
But provide them with a literal word-for-word translation of the Japanese text into English as cheat sheet, and after a few hours of practice, she will become pretty good at rewriting those into fine English translations. Given long enough text snippets, she'll mostly be able to figure out from context how to resolve ambiguities.
(She won't be perfect, of course, and professional translators bring additional skills to the game. But she'll become pretty proficient pretty quickly.)
Try to do a machine to translate between human language, and you'll notice that doing the literal translation part is what's really easy for the computer, but all the context is hard.
(Incidentally, that's why in the stone age of AI, people thought translation would be easy, but playing chess well would be hard.)
Yes, there is nothing dumber than a robot that doesn't know what it is doing. It is merely following a preprogrammed sequence of steps hoping nothing goes wrong.
I really disagree that almost anyone who has built tool assistance into someone's workflow will view their workflow as simple and view the employee with contempt - actually getting to know the BS they have to put up with on a daily basis actually deepens my appreciation for what they're doing. Sure I can cherry-pick examples that make folks look bad - employees today, even millennials, are for some reason allergic to learning keyboard shortcuts and will use right-click to paste. That's somewhat on them, but more on the company for not providing training time to teach better data entry habits.
If one of these highly skilled agricultural workers was given a robot that used scent to determine which fruit trees likely had ripe fruit that needed to be harvested - I'm sure they'd appreciate it for what it is, a tool that lets them concentrate on more important things. And, if we as a society are moving into a era where automation becomes so widespread that jobs are much fewer and far between 1) we'll waste a lot more time building freemium games or whatever BS labour sink society comes up with next and 2) we should realize that taxes need to go way up on the people still employed and UBI or other social services are needed to keep us functional as a society. Having too much productivity is such an easy problem to solve - you just don't do what America is doing and let it all coagulate into the pockets of a few hundred billionaires... and then everyone gets to live a good life.
With that, things like getting a plumber will skyrocket in price.
I mean why should I fix your toilet if I can be at home playing video games right now? I'm going to have a roof over my head and food in my mouth either way.
Current UBI experiments don't have the experience of going 100% of the population, so there's no way to know what will happen.
It would also depend on the specifics of how UBI would be implemented. Where I live I could stop working and live off welfare, providing a minimal standard of living. Yet almost nobody chooses this way of life as people strive for higher living standards, and for many work plays a part in giving meaning to life.
If more and more jobs get automated this might become a problem though, assuming no new non-automatable jobs are created.
If construction costs are cheaper than ever (are they really?) why can't I buy a box to live in without spending $300,000?
Why is the permits process nationally broken? And despite being skilled, that doesn't mean they're informed: why do contractors not know what they're supposed to do? Why do the people handing out permits not know what they're doing?
See also: trying to renovate or buy a home in Arizona.
It really isn't a box for $300k. You can buy a mobile home for what $60k?
Permit process isn't broken.....this is a view often held by incompetent contractors. I have performed various jobs like building retaining walls, bathrooms, electrical, decks and never had an issue as long as my work was up to code.
Permits can protect a homeowner, when the contractor bails without finish the work.
But back to why a home can be $300k, it is a lot of work and materials. From an article, "According to NAHB, the average material cost to build a house is $296,652, with the average square footage of a house being 2,594. That means your cost per square foot is $114.36 ($296,652 / 2,594)."
You are underestimating how much it costs to get lumber, copper(can use pex), brick, concrete etc.
You sound like someone saying, I can have a wordpress site for $1000, but why does it cost $500k to build an ecommerce platform.
Permits cost tens of thousands of dollars, and months of processing in my area. They’re happy to sign off at the end after they’ve extracted the maximum they can from you.
OTOH, insisting on permits potentially saved my ass at least in one case - I had a contractor remodel my bathroom say that "they do everything up to code". Their work did not pass inspections on the first try (had to fix things up both for plumbing and electrical).
The intent in many jurisdictions is to tarpit change as much as possible and it shows.
It isn't the permits themselves which are usually the problem.
I was paying for a house to be built by a contractor who, it turns out, screwed up the permitting process and didn't do a hydrological survey of the erosion pattern on the land. The town yanked his permit to build.
I was grumpy, but when I drove by I could see clearly what the problem was: he hadn't planned for drainage at all, and the new construction and alteration of the terrain had already caused rainwater to start pooling at the base of my future neighbor's foundation. If he'd been allowed to continue unhindered, he would basically have guaranteed my neighbor's property would have been destroyed in 5 years.
Manufactured homes are like $20,000. The plot of land that they go on however, is subject to political rules. Turns out that people don't like cheap housing to be available, because it attracts the underclass.
Its important to remember that housing is mostly a solved problem. The politics of housing are not solved however: too much "progress" (aka: higher prices) and people complain about gentrification.
Bringing down the cost of housing is akin to inviting gangs and thieves into your neighborhood. Everyone wants to support the poor and needy, but just not in their backyard.
Its very difficult to convince a typical city / town / county to support the development of a new trailer park. That's just how politics are today. Politicians are screwed if prices go up. Politicians are screwed if prices go down. Politicians are screwed if prices remain the same.
-------------
Its not very difficult to physically build any kind of "underclass" home. Trailer parks, low-income housing / mixed-income housing (be it a high-rise or low-cost apartment/condo complex), etc. etc. But its very, very difficult to convince local board members to zone an area for that kind of home.
Machine learning is not a high return investment on its own. It is only leveraged as a technical piece of an already cash flow healthy larger pie in most organizations.
The division of labor and supply of various workers makes most labor intensive tasks self-explanatory in cost.
$300,000 can buy you a property that takes years to double in value in comparison. They're completely different categories of capital utilization.
Low-hanging fruit I've seen in this area has included automating or augmenting labor-intensive manual processes (e.g., classifying insurance claims), replacing human judgement with time-series forecasts within an ordering system, and optimizing email campaigns with RL.
They're not cheaper than ever. Materials are expensive and in short supply. Labor is expensive because demand is high and few people want to do manual construction work.
But you're not literally buying just a structure. You have to buy a property for the house. In many locations, the valuation of the land a house sits on is higher than the value of the structure itself. My house could burn to the ground and my property would lose less than half its value, and I'm not even in a very high cost of living area.
You can buy a house for $300,000 if you're not picky about where it is. One of my family members had a new construction house built in the 2500 sq. ft. range for under $300,000 recently. The trick was that they live in a small town where property can be purchased for around $10K.
> See also: trying to renovate or buy a home in Arizona.
A quick search on Zillow shows over 1,500 listings for freestanding single family homes (not apartments, condos, townhomes, or other properties) for $300,000 or less. Many of them are even recently renovated, have 2-car garages, and frankly look quite nice.
Usually when someone is talking about starter houses being exorbitantly expensive, they're actually referring to the property being expensive. Loosen your requirements for location and it becomes much easier to find affordable housing. In most states you don't even need to move that far out of city centers. There are hundreds of properties within Phoenix and Tuscon that are $300K or less.
It goes both ways
If these numbers are representative, it's no wonder it's hard to hire a contractor. The pay sucks and it's hard work that requires a lot of training. Overall I don't understand how the economics of this industry work such that doing anything is expensive as hell, yet the workers are not paid very well.
Ask me what HVAC companies make in the hottest city in the nation.
The costs vs sale price are relevant to the builder in making the decision on whether to build and whether to sell, as they wouldn’t do it if they’re going to lose money.
Also, housing isn't expensive because of labor costs. In many places, the land itself is worth more than the improvements made to it.
The chief cost of housing is land, and that follows the laws of economic rents and assets, in two principle ways:
- For goods subject to rents, that is, where supply is nonresponsive to price, any surplus value accrues to the seller rather than the buyer. The marginal price tends toward the marginal value rather than the marginal cost.
- Moreover, as an asset, land is subject to the general behavioural tendency that holders of assets will tend to act such that the value of assets inflates. This is through constraining supply, sale, credentialling, regulation, certification, and other elements. This shows up in zoning, construction, certifiction, lending, insurance, and other obligations or limitations on new construction.
- Unskilled labour is subject to wages, which tend to fall to the level of marginal cost, if not below that. Incidentally, in a simple society of labourers vs. rentiers, surplus value accrues to the rentiers as a case of the law of rent vs. the iron law of wages.
- Skilled labour wages deviate from those of unskilled as skill itself is a rent-generating factor, and hence skilled labourers can command somewhat higher wages. (Adam Smith's discussion of the five factors involved in the wages of labour are a fascinating read.)
- Materials follow natural resource pricing, which ... is very poorly understood by modern economic theory, but tends to follow the marginal cost of provision by the marginal supplier. That is, the supplier with the highest viable cost structure sets the market price for the commodity. Perversely, when the market is oversupplied, prices often fall (expected) and production rises as the only way producers have of meeting their own fixed costs is to produce and sell more at the lower price. (This occurs because marginal costs may fall below fixed costs.) The situation is not long-term sustainable, but may persist for a substantial period of time. For purely extractive resources (mining, gas, oil), the replacement cost of the good is not factored in at all, and market prices may be hundreds to millions of times below any rational fully-accounted cost basis.
Net productivity of labour has increased markedly. The land-cost of housing has increased far beyond that. Materials have shifted in several ways, with numerous older materials having been replaced by nominally-cheaper modern equivalents. Restrictions on construction methods, materials, designs, etc., impose additional costs and constraints in design and building as well.
The question largely reveals a large ignorance of the factors involved.
Watching those people on youtube who make tutorial videos that go super fast and make it look really easy... You can record yourself writing a python script and make it look easy. Those people are pros, and have skills and experience to back it up. It isn't easy.
Can you think of any other reasons that people might believe something like automated driving is right around the corner?
Or is "contempt for the skill and intelligence of ordinary people" the only reason one might believe this?
Computers can do really well in controlled, specced situations. Not so well in a complicated environment where humans live and operate.
Computer people tend to work in office settings. The reason the old joke about threatening to replace people with shell scripts works is because a lot of office jobs can be replaced with shell scripts. This is because, rather than physically working in the real world, most office jobs entail manipulating symbols, which is easy for computers but hard for humans. (That's why we invented computers).
A lot of work in the real world entails some decision-making component that can be expressed in terms of symbol manipulation. This overlaps with the part that many people find difficult, because people are naturally good at physically manipulating the real world. So when we talk about automating driving or picking tomatoes or bricklaying, we are only accounting for the difficulty of automating the symbol-manipulation abstraction of that task.
(This is basically the conceit of "Manna", a science fiction story where a ton of jobs are "automated" by a computer that just gives a continuous stream of verbal instructions to a poorly paid human being.)
If you're measuring the skills of manual workers against the standard of how hard it is to build a robot that replaces them, they are surprisingly hard. The same is true for the skills of trained animals. Maybe we should start worrying about replacing human beings with robots after we've managed to figure out how to replace dogs with robots. Once we have enough bomb-sniffing robots, drug-sniffing robots, livestock guardian robots, police pursuit robots, search and rescue robots, and so forth, to the point where we don't need dogs, then maybe we can start to worry about the human jobs. (Or maybe not--dogs have a really good sense of smell that might be harder to replicate than some human capacities.)
Isn't that basically amazon warehouses today? (though text based)
Yeah, we pretty much have... it just hasn't been widely adopted yet. https://www.youtube.com/watch?v=yjztvddhZmI
And honestly, from a business perspective, most people really don't care about the skills ordinary people have.
Look at the wine industry for example. Top shelf wine is hand-picked, for exactly the reasons you outline, and some people are willing to pay top dollar for that end product.
But the vast majority of grapes are harvested by very simple machines, because scale and profit, and most people just want a $10 bottle of decent-ish wine.
We absolutely can automate cooking a fucking hamburger, and as soon as it's cheaper than the cheapest labour, McDonald's, Burger King, etc, WILL automate cooking a fucking hamburger.
Will it be as good as a "hand made" burger? Maybe.. maybe not?
Will the people paying a few bucks for a burger care? Nope.
This isn't an issue with "computer people"... this is an issue with economics.
Curiously, McDonald's is probably as close as you can get to automating burgers and fries without crossing that final step of removing human labor entirely. It's not really cooking anymore, it's process chemistry[0], with some final assembly required.
--
[0] - Which I actually consider to be a good thing.
Once upon a time I'm sure some dude had to constantly monitor the temperature of the cellar that holds the wine casks and occasionally call into town for a block of ice to keep things cool. Now we have refrigeration - and that mind-numbingly tedious portion of running a winery can now be conveniently avoided (except for the occasional maintenance) so you can concentrate on more important things.
Eventually it might be that McDonalds automates it's POS and menu to such a degree that for every ten franchises there's just one dude in a car that drives around responding to error codes and, ideally, the people who worked there will find something more creative and fulfilling to do with their lives. I don't hold those people with contempt - I feel like their skills (and everybody has skills) are being underutilized.
A fun part of cooking by hand is dealing with the quirks of fresh produce. These once-living plants and meats are inconsistent enough to require decisions on the fly, mid cook. One cut is 50g thicker at one end than the other, one parsnip is 10% more dense at the top, this cod retained more salt than yesterday and requires less seasoning, etc.
Of course this kind of cooking is miles from what happens in fast food, but there’s a reason the burger station is the most senior position. Cooking patties properly requires skill to get them crispy, fully cooked, but still juicy. Hot damn I’m hungry now.
There must be, running an open-loop process is rather tricky. If it's like any other process in food manufacturing, there are likely both continuous automated checks of trivially measurable properties and regular (every hour or less) manual lab tests[0].
> These once-living plants and meats are inconsistent enough to require decisions on the fly, mid cook.
As I understand their process (from what little I read and watched about it in the past), the keystone of it is minimizing input variance at every stage. They stick to specific varieties of potatoes, they keep bun production centralized, they place specific requirements on slaughterhouses for the meat they then blend into standardized pulp, etc. This makes it possible to constrain variance downstream - e.g. the meat pulp has roughly uniform properties, so it can be shaped and processed in tightly controlled fashion.
At the end come pre-processed food pieces that are then shipped to restaurants for assembly. McDonald's venues don't really cook this food, they put it through the last stage of processing pipeline - heating it in programmable ovens/fryers. Restaurant workers aren't making any cooking decisions. That's how McDonald's ensures its product consistency - there's no place left for a human to screw things up.
--
[0] - Industrial food processing plants have labs on-site, which receive samples from the production line on an ongoing basis. This is used to check the batches for contamination / pathogens, verify measurements from automated sensors, and to provide ground-truth data for models used in model predictive controllers, if such are used in the process.
It's time to link to the AMFare video again.[1]
This was done 50 years ago, but wasn't cost-effective.
Country Garden in Beijing has several automated restaurants. They have 6 restaurants on this system now, one of which seats 600 people. So at scale, this is starting to work. Maybe. Looking at the videos, it looks hard to clean. Commercial food processing equipment has a look of "can be blasted with boiling soapy water daily", and not seeing that here.[2]
"Pretty much". heh, there are 6 distinct levels of car automation for a reason.
> it just hasn't been widely adopted yet. https://www.youtube.com/watch?v=yjztvddhZmI
Now try this in Wisconsin in January with 8 inches of snow on the ground. I consider Wisconsin wide adoption, do you?
We don’t have to perfect automated driving; we just have to get it better than the average human… and that’s a pretty low bar.
And we don’t even have to automate driving everywhere. Who cares if we can’t automate it for 5%? Can we automate it for 95%? Hell, if we can automate it for 30-40% with a lower accident rate… that’s already a win.
1. Get a standing desk, so you can alternate between sitting and standing.
2. Learn to use a mouse with both left and right hand. And/or ball wheel mouse.
3. Try Speech to text software
4. Try an Ergo keyboard. Logitech makes an ergonomic keyboard (Ergo k860)
5. Get up and Stretch every hour
6. Go the the bathroom as much as you're allowed to
7. Go for a Walk during your lunch break
8. Go for a walk after work
9. Do some yoga. It's totally free.
https://www.youtube.com/channel/UCWN2FPlvg9r-LnUyepH9IaQThe things you mention do not particularly happen to everyone and even less so since safety standards have improved.
Because sitting at a desk all day for me definitely isn’t it.
Also, if you speak with people in ML/AI, they will not tell you it's around the corner. They know that physical stuff and robotics in general is hard.
Also, builders and the like often have a great amount of contempt towards any desk job - is that OK? Why?
Jobs that require interaction with the physical world and people will almost always be a hard to automate thing. There are simply too many variables.
The real problem is when we make the results of a job worse to aid in automation. The prime example here is tomatoes being breed with thick skin, picked before ripened, and then squirted with some chemicals to make them red so people will buy them. All while sacrificing taste.
One can absolutely admire the skill of a bricklayer, while celebrating the eventual fact that bricklayers will not be forced to perform that labor if they don't want to. Those are not mutually exclusive things.
"We haven't automated driving"
One can, today, fly to Phoenix Arizona, download the Waymo app, and summon a self-driving vehicle. Although it would be disingenuous to claim that self-driving is a completely solved problem for ALL use cases, it absolutely is "good enough" for very many of them, and will be scaled out faster than you expect.
"but many more have not and most likely never will be."
Unless there's some yet-to-be-discovered supernatural process at-work inside the human brain, this is just plain false. The human brain is an incredibly complicated machine, and nothing more. It's only a matter of time until we understand, or in the worst case, simulate it with sheer computational power.
Exactly, and those actions are able to be simple and repetitive because the machines don't have to try to identify objects in the environment, they just reach X meters, open their grappler Y centimeters, close it until it reaches Z Pascals of pressure, move, repeat.
No real "AI" required, just programmed to handle expected conditions. But get just a little bit outside the pre-programmed constraints and... https://www.engadget.com/2018-12-06-amazon-workers-hospitali...
As a result, the humans are now forced to accommodate the failings of the machines. https://www.thesun.co.uk/tech/8346932/amazon-factory-workers...
Agreed. My mother was involved in my local city theatre, and always told us kids that "there would be no show without someone to lift the curtains." These people had their cameo in every finale. That has always stuck with me.
Ever wondered what would happen if there wasn't someone flipping burgers to satisfy your 2AM craving? You, no matter your riches or power or fame, would have to do it yourself.
The most innovative thing you could do is shut the fuck up and start listening. Instead of trying to replace the bricklayer with robotics, how about asking them how robotics could help them? They are the SMEs, they do a job considered easy to automate daily (and that job isn't easy to automate). They can tell you how they are wasting time on inefficiencies.
The job of technology is to allow the human to stop worrying about the rote. It is not to replace the human, or to solve imaginary problems. That's when problems arise: Facebook, racist/sexist AI, etc.
They miss that the contempt is not directed at those doing the work but instead directed at the skills-- this is evidenced by how they tend to trivialize the skills that are trying to be automated.
This is especially true when looking at the spectrum of quality of human work for a given field. Take, for example, masonry. There is simple bricklaying for the front of the house. Then there is intricate stonemasonry that requires a lot of careful planning, know-how, and knowledge of historical techniques.
There is machining a piece of metal in a simple way. Then there is intricate hand-finishing of that metal at the other extreme.
I suspect that over time we will see the less artisan, less "skillful" aspects of manual work be replaced but that fine craftsmanship won't really disappear. The more you learn about what goes into making certain things by hand the more you appreciate how foolish it is to claim that it can all be automated easily.
I could do the job of every logic gate in my computer, but I wouldn't want to try and boot a copy of Windows with a pad of paper and a pencil.
As an example, not something you'd want to make on a hand operated milling machine: https://www.youtube.com/watch?v=RnIvhlKT7SY
But to look at something that's easy to do for people and think "That's easy, a machine could do it!" It's always going to depend on why it's easy. Handling situations that differ from one job to the next is a pretty strong point for humans, especially when those differences are handled by grabbing and moving things.
But then, five years ago we would've put "playing Go" on there with grabbing and moving things, so you never know.
Automated cement-pouring in place of brick laying at https://www.nbc12.com/2021/06/24/virginias-first-3-d-printed...
The robots have arrived. Unskilled labour needs to be concerned.
EDIT: Yes, they didn't last long: https://hackaday.com/2018/10/11/i-ate-a-robot-hamburger-befo...
I remember how people were amazed when robots like Asimo could not be used during the Fukushima disaster. Anyone who works in these fields, even peripherally, has a very different perspective of reality vs. fantasy and knew the score as it pertained to robots like Asimo, their actual utility and capabilities.
The most automated things have been designed by engineers with robotic assembly in mind.
Which is why the points about non-Newtonian fluids like mortor changing viscosity seems right on the money. Anyone trying automation with normal mortor gives me zero confidence. I like the robot truck idea, seems like a winner compared to others.
Exo suits / better humans in the meantime, purpose built machines later.
So, the building tasks are set to be a compromise of materials and human skill.
If we want machines to build them, then we need different systems and materials.
I think I laid my first brick before the age of 9, but first just a couple here and there. Bricks are heavy! I think was 12 when I was given a straight stretch of the wall close to foundation to, and after review of my work I had to take everything apart, it was saggy, uneven and ugly. Eventually I learned to do straight walls, but the corners eluded me until at least a year doing it with very qualified masons coaching. I don't think my joints were ever pretty enough with that lovely groove that some people do, and avoiding mortar onto the brick is devilishly hard.
Where I live there is little brick construction but I love seeing what is there, it is always so pretty.
Based on my experience of the changing by the minute mortar, heavy bricks, continuous looping feedback, I don't believe it is something that will get automated.
Besides that It's not my experience at all that "computer people" think that these things will be automated in a short time frame. It's the opposite. The "computer people" are highly skeptic because they know much better then anyone what a computer is capable of. It's almost always the managers and marketing people parading automation around.
Things will become automated. That is the natural human progression. We are creative, and their are 7+ billion minds with new giants to stand on the shoulder of every day.
AI, soft/compliant robotics, and globalization are strong unrelenting factors/motivators.
You probably wouldn't have to go too far back to find people sceptical about the possibility of automating that.
The article suggests efforts go back over 100 years, so it seems bizarre to suggest this has something to do with "computer people".
There is a commercial burger cooking machine north of me in SF (https://www.creator.rest/), and there are similar machines for pizzas and other fast food. That isn't an unsolved problem.
There's folks on YouTube already who can automate the assembly of a hamburger using Lego. It's just slower than having a person do it in a semi automated manner.
Per the article, we have automated some of bricklaying (for roads), and I've definitely had ads show up for tools that reduce the skills needed to do bricklaying.
We've also automated parts of driving - staying the same speed, parking, etc. The automation all changes which skills are important, and steadily makes high skilled jobs into less skilled jobs.
Devil's advocate asks, what about the skill of the devs that can automate that kind of stuff? The types of things that might be learned in building articulating robotically controlled apendages to pick tomatoes without crushing them or the entire plants, the skill of a robot to cook the perfect burger to the perfect temp, etc. It would be impressive.
Sending people to the moon and building atomic weapons brought us some very cool and useful unintended things. So who knows what could come from some other boondoggle type of projects?
No. They just don’t know what makes it hard. I don’t know where the robot train drivers are. We’re doing cars and planes but it seems trains have fewer things to worry about. I don’t feel contempt for train drivers, the limitation is mine. I don’t understand. I still think I deserve to live.
I love this guy’s videos. He’s a fantastic insight into a world I barely know: https://youtu.be/qP1AmDRhoas
In order to make automation economically viable there would need to be a niche for "cheap" brick powered by automation. It's unclear to me that this niche exists and can't be similarly achieved by scaling up the blocks or switching to more labor efficient building materials.
Kind of an unfair generalization, I think. I see far more comments in defense of the skill of masonry and masons, insisting that automation may be too difficult, and almost none that "display contempt" as I see it. But I could be missing something. Would you be willing to expand on what you see here?
I wouldn’t know whether their claims hold water, but http://www.automatorobotics.com/autonomous-single-tomato-har...:
A robot that can reliably detect ripe tomatoes and harvest them with consistent quality at a rate of 10-12 tomatoes per minute.
[…]
reducing the harvest costs by 50%
(The videos give me the impression that “10-12 tomatoes per minute” is best-case)
Heirloom tomatoes look like a topology project gone horribly wrong, and the colors aren't found in any Pantone catalog. They are also the most delicious.
There's definitely a place for both. We have to get ketchup ingredients from somewhere, and gourmands will always want the ugly heirlooms.
Throughout history, our main trick when solving a problem with technology was to constrain the problem to something that's much easier to solve, by changing the environment.
For example, all-terrain mechanized transport is hard, and to this day, nobody except hobbyists and militaries even bothers to try tackling it. Our solution was to flatten the terrain into roads, so that we could just use wheels. We've been iterating on that idea ever since.
That one didn't make money either, and it also broke a lot, and had all sorts of other problems. Believe it or not, the core technology has improved a lot less than you might think. They may have put a prettier skin on it but it's still the wrong approach.
Where do you get that? While you focus on the "skill" and "intelligence", automation is about productivity and freedom (from labor)
> learn some humility and try to understand the skills
What do you think automation is, if not to try to understand the skills, then turn it into the program?
(1) increases the value produced per unit of labor input in the field at the low end of quantity, but potentially resulting in reaching the point of rapidly diminishing returns faster.
(2) and, increases the domain in which creating/maintaining automation has value, increasing the value of automation work and pushing out the point where automation work reaches diminishing returns.
Which is why programming (viewed broadly)—itself part of automating any field—keeps being resilient against adverse impacts from automation. Because as long as we are increasing the scope of automation generally, we’re pushing out the diminishing returns point for programming, and automation of programming isn't pulling that point down fast enough to offset that effect. And automation—both of programming and of everything else—keeps pushing up the starting value of programming work. (It also makes it more abstract and about providing input to higher-level automation tools.)
Technology is the domain of means to ends. It's literally how things are done, and includes such factors as skill, training, technique, organisation, and logistics.
As such, your comment is self-negating.
I'm confused - based on your Github profile I assume you'd fall under the category of "computer people". Are you saying you feel this way? Because I certainly don't.
Frank Bardacke in his book, _Trampling out the Vintage_ describes the great skill that agriculture workers bring to the job that they do, from knowing whether something is ripe enough to pick to the dexterity required to pick it without destroying the fragile fruit or vegetable. Some jobs have been automated, but many more have not and most likely never will be.
Please, learn some humility and try to understand the skills that every human possesses and that they bring to their work.