"Digital" cooperatives using (optionally) open source platforms is an idea that should be much more explored imho.
The concept applies to so many verticals involving small companies or individual operators (from taxis, to hotels, to restaurants, to real estate agents, local bookstores, cinemas, to game developers etc. etc.)
The inability of large numbers of small entities to coordinate and cooperate in the digital era means they create the conditions for third parties to completely take over the intermediation with clients and use that leverage to extract unreasonably large rents.
In fact the extreme "taxes" small businesses and individuals have to pay to shrewd digital platform owners are multiples of the cost of development of the platforms themselves.
In other words, if a collection of independent entities cannot get they act together to codevelop digital infastructure (that will benefit all of them) somebody else will do it for them and will be eating their lunch.
This reclaiming of their market / ownership of digital tools might be challenging in conditions of rapid technical innovation. The autocracy of a top-down corporate structure is more effective than a democratic collective. Yet despite incessant marketing noise currently we are in a rather stagnating phase so there is a window of opportunity to reclaim some agency.
If there's a 3rd party aggregator (for taxis, restaurants or whatever), the final price that the customer pays would include both the aggregator's profits + the individual service provider's profits.
If the individual service providers all came together and made an app themselves, the aggregator-profits can be eliminated, resulting in potentially reduced prices for the customers.
I think (probably wrongly) one of the reasons they don't usually come together is because by making such an app they would effectively be helping their competitor(s), which they may not necessarily want to (because they don't see the bigger picture?).
In my city (in Asia) a subset of taxi drivers realized this and even made an app, but it didn't really work well due to poor marketing and because visitors to the city didn't want to set-up a new app just for 2 days etc.
On the other hand, if all the service-providers do co-operatively make an app for their services, what prevents a cartel from developing? The government?
I guess there should be a high baseline level of trust to make it work.
> Maybe it would work better if the local government runs the app (contracted out obvs) and bans all the others.
I like the idea of that but the issue is software that people have to use with no alternative end up having no real motivation to be user friendly or stable or anything like that. They own the market and nobody can ever leave it
If every city has an app, you also lower the trust of the drivers because you don't have a way to know if the passengers are behaving well (good reviews) or not.
So much "this!". The hotel/booking is the most blatant example, because the bar for "good enough" wouldn't really be that far up.
The only moat they have is worldwide discoverability (a regional cooperative just won't reach people who visit once, cooperatives would have to start global, or at least be really, really good at federating) and that the threshold for critical mass in that market is winner takes all: at least when you are not looking for that one spot for a two-weeks getaway (you might even enjoy trawling multiple platforms) but going through serial booking for a road trip (doesn't even matter wether it's bulk ahead or on the go),
Depends on the vertical. I would not be real comfortable booking lodging on some random portal, especially if we're talking more of an AirBnB thing vs hotels.
Yeah, that's one of the moats some have but booking.com has not: you don't just pop up a hotel from nothing, and the hypothetical failure mode of complete fakes appearing out of thin air is not only solved by payment on site but also by the local hotel owner community having a financial interest in policing their turf.
Or not, could be a fun plot for some speculative fiction: on a cooperative post-booking.com platform, hotel owners in a remote valley discover that they can earn quite nicely by offering "rescue nights" to guest who fell for a series of fake listings they keep up, untraceably because the fakes lack a payment channel to follow. Add some Bates vibes to the one hotelier who refuses to take part in the scheme for multi-layered tension.
A cooperative booking organization would have strong incentive to keep their members in the straight & narrow. It is a cooperation problem and institutional enforcement is a proven recipe for ensuring that.
If there were/are lots of small portals for each region/group of hotels, how would you know which portal is the (a?) genuine portal?
People on the other end of genuine transactions want/need some kind of payment from the people booking (to prevent at a minimum people booking lots of rooms with no intention of staying), but that's an attractive target for scammers. Especially as people often book things a significant length of time in advance.
> The autocracy of a top-down corporate structure is more effective than a democratic collective.
If that was the case then Soviet Union's central planning would be more efficient than Liberal democracy's market capitalism. The efficiency of corporations stems from unfair advantages states gives them like intellectual property rights or from socialising their losses and privatising their gains from public budget.
Elaborate on this please, after all even in "autocratic" corporations we have internal power struggles etc. So are they democratic too? Suddenly everything is like democracy?
USSR, with the exception of the elections, wasn't much different from what we recognize as democracies.
i.e. a local bureaucrat notices something, he contacts his bureaucrat supervisor, let's call him the city manager, who then decide to command the bureaucrat to investigate further or ignore the issue.
Once it is established that the issue actually exists, the local bureaucrat supervisor contacts his bureaucrat supervisor, let's call him the area manager who once again analyzes the report, or more probably gives it to his subordinates to analyze it, because he has a lot of more important work to do, he's a manager!, and then decides what to do next.
Once it is established that the issue actually exists, the area manager contacts his supervisor, let's call him the regional manager, who once again analyzes the report, or more probably give it to his subordinates to analyze it, because he has a lot of more important work to do, he's a manager of managers for god's sake!, and then decides what to do next.
Rinse and repeat a dozen times in a system deeply nested (full employment was a primary goal in USSR, meaning that the machine was slow and inefficient but provided job security to almost everybody) and it is clear that what Stalin said or commanded was the result of an information gathering process that wasn't so swift and to deploy the solutions that the central committee deliberated they had to face some slowdown and a lot of friction.
To understand how fragile the system was, imagine that the fall of the Berlin wall was caused by a trivial communication mistake made by Guenter Schabowski during a momentary inability to contact the bureaucrat who knew the correct answer to the journalists questions.
Democracy simply means that we vote and have many different parties to chose from (not so many in the US, for example), a parliament made by one, two or more "houses" where elected representatives discuss matters, etc. etc., but what we call "the machinery of government" it's not very different from system to system.
Take China for example, yes, they have only one party, but the CCP ha almost 100 million members, with many different political currents, that constantly discuss different point of views and try to find a compromise, making it one of the largest, if not the largest, political institution in the World and in human history.
We do not call it a democracy, but it is still humans discussing politics that have real life impact on over a billion people that form their community.
Not sure this a good counterexample. In fact we have plenty of proof that loose collectives dont perform well under rapid change conditions.
There is no reason why the ubers and bookings etc should not have been actually platforms developed and owned by the respective sectors using more or less the same technologies. In fact there are some examples.
The lag in responding to a risk of opportunity is intrinsic to a consensus system. The long term fitness is an entirely different matter.
Problem is that web3 tries to tie everything to a cryptocurrency token, not willing to accept the fact that as soon as something happens off-chain and in the real world, most/all advantages of using cryptocurrencies and blockchains go away.
It really is a shame that all blockchains (AFAIK) are tied to cryptocurrency as reward for participating in their consensus mechanism when it is not necessary the only way to do things.
In fact, I believe blockchain will never be decentralized as long as there is a means of accumulation of power over the blockchain, in many cases the accumulation of PoW computing power, or in PoS the accumulation of the cryptocurrency itself.
I also believe (without any basis other than the spirit of the correspondence from Satoshi) that Satoshi released Bitcoin in 2008 as a proof of concept, not as a project meant to be adopted by every institution. Bitcoin was created to address the Byzantine Generals problem, and AFAIK it was the first to truly do so. It was not created to be the blueprint to every other blockchain ever.
The whole concept of decentralized trust-less environments requires some way to prevent malicious attackers from taking over the network. Proof of work, while extremely wasteful, does that job.
The main problem is that trust-lessness and decentralization isn't actually necessary, especially for most things that exist in the real world and are at the mercy of the legal systems they operate within (this applies to cryptocurrencies as well - the government doesn't need to break crypto, it can just make interacting/transacting with it illegal).
For a lot of real-world projects, the legal system is an adequate protection. Centralization doesn't require a profit motive - the central party could be set up as a non-profit and then have all the advantages of a centralized system while avoiding a lot of the drawbacks we see in for-profit operations.
> Proof of work, while extremely wasteful, does the job.
True, yet you could even have Proof of Work without cryptocurrency. Cryptocurrency is used to incentivize PoW which attracts more people to the blockchain and strengthens the network and so on.
However, one could have Proof of Work without cryptocurrency, for example, one could reward the PoW with participation in the blockchain, or in this case, have drivers PoW to be able to take rides.
> Centralization doesn't require a profit motive
Yes, I also agree that centralization doesn't require a profit motive, and a centralized solution is deal for 99% of the cases.
Shout out to Namma Yatri [1] app in Bengaluru, India. It is developed by a payments startup called Juspay. It is endorsed by the Auto Rickshaw Drivers Union. It was launched in November of 2022 and has since done about 5m rides and drivers have earned 9m USD [2]. The entire app, backend and protocol is completely open source [3]. Seems like it uses a fair bit of Haskell and Nix!
> FOSS is a great path for workers to seize the means of production.
It's... not. Open Source is not a business model, nor a collective organizing principle, bargaining process, etc. It's just source code a couple of people have commit access to. I'm all in favor of everybody sharing source code too, but let's not pretend it's magic. Building and sustaining any kind of organization takes work, and has nothing to do with whether your source code is open or closed.
>It's... not. Open Source is not a business model, nor a collective organizing principle, bargaining process, etc. It's just source code a couple of people have commit access to.
in fact that's one of the driving characteristic differences between open source and FOSS; one is accompanied with a set of guiding moral and ethical principals regarding fundamental user freedom.
FOSS absolutely does have an effect on how your movement is organized and if you depend on it it can be an accelerator.
But FOSS doesn't create some kind of socialist utopia as some people seem to believe. It does something much more mundane yet infinitely more valuable in our fractured modern world: it unites bitter enemies and gets them to collaborate.
Look at the contributors to the Linux kernel, which even includes Microsoft now. It's hard to imagine any other force which could unite all those people and businesses under a common cause. But thanks to the GPL, even as their wars with each other continue they are all contributing resources to the public good.
It's so popular these days to flame and dehumanize the other side of any debate but all that does is destroy. FOSS builds and extends humanity's common heritage. It won't win the war for you but it will redirect your energies into creation and ensure that you leave something good behind.
"Bitter enemies" work together all the time without open source. Competing companies partner on initiatives, join the same boards, help each other pass new standards. Companies do not have emotions. They have profits and losses, marketing, products, coalitions, industry associations, etc. They are fine with working with competitors if it has no downside. Often they have to work with competitors, if only because their parent conglomerates are so large that they each depend on some product or service the other provides, or their customers use some product of the customer's and supporting it keeps their customers around. It's just part of doing business.
There is no magic to FOSS. It's just a weird quirk in the global capitalist market. It has no real bearing on business, unless your business is catering to the tiny minority of people who care only about FOSS.
Thank you for writing this out, I never worded it like this to myself but it makes so much sense.
I love the FOSS movement's use of the legal system, mixed with ease of adoption, to force big players to contribute.
FOSS, while not being a socialist utopia, is one of the few truly cooperative systems in our society, and it also goes to show you cannot trust people(corporations) to contribute out of the good of their heart, you have to force their hand through the very tools(ip law) that these people and corporations use to wield their own power.
This may stray past your original point, but whenever I hear someone complain about something being "authoritarian", I find it hard to take seriously because the profit motive almost always goes against a general common good. ie covid lockdowns, lives saved vs dollars earned, where do your priorities lie?
I'd like you to imagine a world where GNU and Linux never took off, where we'd probably all be paying for Windows Server licenses in order to run websites. One where you need to pay to download a compiler, and all you get is the binary.
That's hard to imagine, given the freedoms we enjoy now. I can download a world-class OS for free and distribute changes to it without any hinderance other than my own skill barriers. Same for DVCSes, video players, browsers, modelling software, desktop environments, compilers, and so on... That was not the case in the early 90s.
FOSS wasn't a complete nor a permanent revolution, but revolutions rarely are. Point is, it changed the world, it changed business, it changed developers' day to day experience, and it's disingenuous to pretend otherwise. We still rely on corporations to make software, but they are now socially pressured and incentivized to release the source of a large portion of what they make for free. Maybe we haven't seized the means, but we are right there with the big guys at the means, sharing power in a (relatively) new and exciting way.
MINIX and BSD were already free. More free tools would crop up, just like free Windows tools abounded since their inception, like the hobbyist software that was traded for free before that. Nerds who just like making software would always have kept giving it away for free, because they just like the fact that they can and it's fun to do.
It didn't change the world that much. Nearly every company that makes internal software refuses to release it for free. They'll happily use whatever you give them for free, of course. But even if it weren't around for free, they'd just pay for something else. Most people would probably use Android as the example of the "revolution" paying off, but before Android everyone just paid to license mobile OSes, and they will again when Android eventually dies. Business does not care about the morality of a license, nor sharing code. You pay for what you need when you need to, don't pay for what you don't need to, and pass the cost on to the customer. Simple as that.
There is no seizing here. FOSS is available to all, no one actually loses the means of production.
It’s “distributism” rather than “socialism”.
Anyway — FOSS can be crowdfunded by selling utility tokens to future customers. If you are worried about the SEC, then just use a broker-dealer like Rialto Markets to help you do Reg CF.
There's good margins in any software that people want to use. There's money in Nextcloud hosting, Matomo services, or Bitcoin. All of which are open source projects that attempt(or attempted to) put power back in hands of users. There are many more examples.
VC will gladly fund a business model with solid margins. They often prefer those that might get a monopoly, grow exponential etc. sure. But it's certainly not the only mode.
The thing about software is it's remarkably cheap in the grand scheme of things. To choose two random modern examples, Signal and Lemmy both got their start with a little bit of funding from nonprofits. Signal then basically funded its bills with a Whatsapp consulting contract and now it pays for a CEO that goes around doing prime TV appearances where she shoots down politicians who are trying to rob you of your freedom. That's a hell of a story arc.
The customers want to donate money to local taxi operators and drivers?
The taxi drivers who barely understand what open source is?
This can work, but definitely not through crowdfunding. This better works through say business associations, say a national taxi association (composed of all small/regional taxi operators). That association can allocate a small portion of the funds to build this open taxi framework.
The truth is that I'd be happier donating to the project of a non-exploitative marketplace platform than tip a driver out of starvation while virtually all of the nominal rate goes to making a unicorn golden.
This comes across as overly snarky. At best, you delayed productive conversation by the time needed for two asynchronous comments. At worst, you shame someone out of getting an answer to their question.
Unfortunately, seems like it hasn't been maintained for quite a while (https://github.com/ro31337/libretaxi/graphs/contributors), doesn't support credit/debit cards (cash only) and also doesn't seem to offer any safety features (any driver can sign up at any time without any sort of verification)
Scheduled rides are a neat strategy, but they're tough for consumers to use effectively. For one thing, if you're attempting to schedule a last mile ride after you get off the subway, your arrival at the pickup point could be subject to minor delays. But on the network side, a few minutes on each ride impacts efficiency in a major way.
Personally, my biggest (non ethical) issue with Uber/Lyft et al. is that they're unreliable. There are common rides you can request (e.g. leaving the urban core after work when drivers want to be heading in for bar service) that drivers simply don't want to do. I don't think there's a faster way to create negative user associations than walking an hour home in a heat wave because the pricing model failed.
The platforms care little about individual customers (as opposed to drivers, which are fewer and have higher onboarding costs) so they are reluctant to set up the right incentives to force drivers to be reliable.
A simple solution (in most gig economy markets - not just ridesharing) is that any accepted task costs you a deposit which gets returned on completion. This instantly makes it unprofitable to offer sub-par service or cancel, and provides stability to both customers and the system in general as any accepted task is more likely to be delivered (and tasks which are unprofitable to accept will not get accepted, so the surge pricing mechanism can do its job and raise prices until acceptance rates go up).
The problem is that this will only happen if there is either a legally-enforceable service-level-agreement (such as compensating passengers for unreliability or missed arrival times) or strong competition, which is not currently the case. Most gig economy markets are dominated by a handful of equally-mediocre players who have no incentive to out-compete each other.
They could have a successful business that way, but I'm not sure it's one that can handle 9,000+ drivers like the article calls out with any sort of equitable scheduling. For context, that's more drivers than the entire city of San Francisco typically has at peak. Maybe I'm just vastly underestimating the size of the NYC airport market though.
Ignoring the details of this particular initiative, from a high-level view this is wonderful.
Similarly to how Tesla's future failure or success is irrelevant as they've already succeeded in the most important task of proving that EVs can work and thus driving the industry to produce them en masse, Uber and Lyft have driven a revolution.
They've shown that the old model for running a taxi service was horrendously outdated and customer-unfriendly (and also driver-unfriendly, in some cities) and now others can innovate and compete along the same lines, differentiating via areas that Uber and Lyft either can't or won't. I know it's essentially a meme on here to diminish Dropbox or Twitter to being basic services that one could easily replicate in an hour or a weekend, but it's similar for Uber and Lyft: the underlying concept and technology is relatively easily replicable - meaning their moat is pretty insecure. As time goes on, one can envisage them dying a slow death by a thousand cuts.
From a consumer's perspective, some sort of federation or interoperability would be needed between these competing apps. If it is necessary to research and download a different taxi app for every city --including going through the usual painful signup process each time-- it will drive people back to the common denominator of the big incumbents.
Uber and Lyft seem like a step back: less safety for drivers and passengers as cars are not company owned and inspected and there are no checks on drivers that a taxi company would do. Also the gig economy is overall probably a negative for society as working people do not have the safety of employment along with benefits. What risk or investment does e.g. Uber have that justifies taking a 20% cut?
From a consumers perspective there might be more choices and cheaper options, but that doesn't outweigh the issues it brings with it I think.
The worst taxi I've been in had a completely broken suspension that bottomed out on every bump and threw the car all over the road. My wife got car sick, and we were both afraid we'd crash due to how out of control it was.
The cigarette smell and broken AC were obligatory, of course. Don't forget the check engine light and squeaky brakes.
The taxi owners did not care about car quality whatsoever.
I've never been in an Uber or Lyft that was as bad.
> Also the gig economy is overall probably a negative for society as working people do not have the safety of employment along with benefits.
A lot of drivers paid to rent both cars and taxi medallions and started the day off in the red. Many were contractors and didn't get benefits.
Not sure if it's still there, but in Queens, NY you'd see lots of these taxi medallion and rental places in Long Island City.
In Mexico it's even a matter a safety, there's rampant pirate taxis in Mexico City that if you get on one they might try to rob or kidnap you. Most women avoid taxis at all costs throughout Mexico, Uber and Didi are just so much safer
> shown that the old model for running a taxi service was horrendously outdated and customer-unfriendly
Prior to Uber, you had to pick a taxi company, call them, wait between twenty minutes and the heat death of the universe for the taxi to show up, and then maybe have the driver claim that his POS machine is broken and please pay cash.
After Uber showed up, suddenly the largest taxi service came out with an app where you could order a cab, watch it approach in real time, and pay through the app.
Indeed - and these may be some of the areas that U/L can't or won't improve further, but new competition (like that in the article) can and might.
--
But U&L showed how awfully customer-unfriendly and antiquated the existing industry was. Introducing an app which can call a taxi in seconds to your GPS location, offering easy destination selection, seamless electronic payment, and a feedback system, were the innovations that the incumbent taxi and mini-cab industry was unable or unwilling to introduce, until competition forced them to. These innovations aren't linked to the employment model - that is something that U&L chose.
(Just like Tesla with the Model S, anyone could have done this first, but no-one did, until someone else did.)
> Also the gig economy is overall probably a negative for society as working people do not have the safety of employment along with benefits.
This is not exactly new with Uber/Lyft. I remember a ride with the local taxi company (before they came on the market) where the driver was actually boasting how some previous customer was complaining about him to the dispatch, but he's a contractor so he's safe from any repercussions.
In that situation, at least the 1-star review at Uber/Lyft has a measurable effect.
> Similarly to how Tesla's future failure or success is irrelevant as they've already succeeded in the most important task of proving that EVs can work
They’re still pushing forefronts. NACS and infrastructure roll-out. Iron-based battery chemistries. Cars versus only SUVs.
"net loss of just $318,000 in 2022"
"earned $5.9 million in revenue from 162,294 successful trips — and $5.2 million of that went directly to driver wages"
Let's do the math: they earned $700K in net revenue and lost $318K - for a margin of -45.43%... sounds grim...
But... if they can double the net revenue to $1.4M while only growing costs 25% to 1.3M then voila they're breakeven/ramen-profitable...
I don't quite understand the logic here.. the 5.2MM wasn't fixed costs, that was costs of goods sold because they needed to provide the service to earn in income.
So as far as I can tell they lost 318K, but they don't need to grow gross revenue by very much to close the profitability gap.
Exactly: if they double total revenue then NET revenue (after driver payouts) only grows to $1.4M
Sounds awesome but currently the net revenue was 700k which means they spent ~$1mm on other stuff besides driver payouts. Some of that will grow with the business (e.g. customer service, server hosting costs), some is fixed (e.g. you only need one CEO).
If they can grow the revenue without also growing expenses then they'll hit breakeven by ~doubling revenue, maybe more, maybe less.
Interestingly, i just checked their prices to JFK and they're about the same as Uber. So if they grow beyond breakeven, then they could undercut Uber's current prices...
They developed a new app and essentially built the business in 2022, I am going to expect that the vast majority of the $1MM is fixed and one-time costs. (Obviously they are amortizing the app cost.. but nonetheless.)
But regardless of that if you are launching a new service business and you've got $5MM is sales and you need to hit $8MM in your second full year of operation to become profitable, I'll call that almost profitable!
It says they took in $1.6M of crowd funding. Not sure what the repayment terms are but presumably less predatory than VC. That gives them a few years of runway. And this is a non-profit, so "ramen-profitable" is probably the end goal. Every penny past break even just goes into the cash reserves.
This is an interesting startup filling a need, but I don't think it is really competing with Uber and Lyft, because they are only doing prescheduled rides.
The killer feature for Uber/Lyft is being able to get a ride at any time on demand in a matter of minutes. It doesn't seem like this will be able to get there.
I have wondered if a pure market making app would work... let drivers and riders place bids... a driver can lower their bid to get hired faster and a rider can increase their bid to get picked up faster. The app would just take a fixed cut.
Someone who doesn't mind waiting at a friends house for a while longer could pay less, but the person in a rush could get a ride fast for a bit more. Similarly, some drivers might be ok only doing a few expensive rides.
I know Uber does something like this with surge pricing, but it is not transparent at all, and you can't change your bid.
> I have wondered if a pure market making app would work... let drivers and riders place bids... a driver can lower their bid to get hired faster and a rider can increase their bid to get picked up faster. The app would just take a fixed cut.
There’s a service in Ukraine called Uklon[0] that does pretty much exactly that. If you bid very low, it’d both take longer to find a ride, and the car/driver would usually not be great. Think a 25+ year old daewoo with no ac that smells like smoke on the lowest end vs. newish Camrys or other regular fleet cars on the mid-higher side.
> I have wondered if a pure market making app would work... let drivers and riders place bids... a driver can lower their bid to get hired faster and a rider can increase their bid to get picked up faster. The app would just take a fixed cut.
This is a good idea. But it's important to note that price is not the only relevant parameter here (as opposed to e.g. buying company shares, where two shares are exactly the same)
Other parameters:
1. Waiting time — how long is the driver and the passagner willing to wait?
2. Driver rating — the there a minimum rating a driver must have (and if passengers are rated, too, then vice versa)?
So there's the combination of, at least, these three parameters, which gets complicated. For example: as a passenger, would you rather ride with a 4.5 star rated driver with a 5 minute waiting time for $30; or a 4.0 star rated driver with a 2 minute waiting time for $26?
Could just offer it to the user. They might have said only drivers of 4.5 or greater. So the first one would be shown, but have the cheaper $26 one to the left and the $35 4.9 star one to the right. Let the user decide which they want.
I wonder whether they've considered the possibility of open-sourcing their apps and backend. Perhaps this is just Pollyannaism, but I tend to believe that workers' rights and open-source are the perfect match in a market like ridesharing. "Just goodenough" open source solutions have difficulty breaking through in consumer software (e.g. the mythical "year of the linux desktop"), but in a market like ridesharing, worker coordination could easily provide the push to critical mass.
This is a fine effort, but FOSS is a minor part of it. The big thing is the money -- how much of the fare do the drivers keep; how many staff are required to run it; is the allocation of rides fair; can customers get what they want; how are drivers hired and fired; how are the executives chosen; etc.
There are lots of co-ops in many different industries. The Berkeley Cheeseboard is a co-op that's existed for decades. A lot of agricultural commodities are marketed via co-ops. It's all in how the business is run.
$5.9M in revenue over 162294 trips works out to about $36/trip. Average of 444 trips/day.
$5.2M in driver wages over $60k/driver/year (conservatively, $30/hour * 8 hours/day * 250 work days/year) works out to about 86 drivers; less if they pay each driver more or each driver works longer hours; more if drivers aren't full time. Each driver gets, like, 5 fares a day. (I would also believe $120k/driver/year and 10 fares a day.)
Prescheduled rides seems like a nice niche; better than sitting around in a TNC lot for hours waiting for an airport fare. Would be curious to see how and if they get to the max 2.5x return for the WeFunder though.
>Scheduling trips is a tricky calculation, and it all comes down to routing — you want drivers to spend as much of their shift as they can actually driving clients around, but you have to leave enough time in-between trips so the driver isn’t late for their next client. So far, the only way to do this successfully has been by using a team of human dispatchers, though the co-op is working on an automated solution using artificial intelligence.
I worry about this. Travelling salesman is a solved problem. Throwing "AI" into the mix sounds like someone was sold a bridge.
Travelling salesman is a largely solved problem, but I'd say this problem is larger.
As soon as humans and the real world are involved, seemingly contradictory criteria need to be met.
Algorithms like A* search may not be able to capture these criteria robustly enough.
Example: a driver may be assigned the optimal route, but because they are human will not want to sit and wait for 15 minutes between several of their rides
A* is great for path finding but it's not a global panacea for all subproblems related to ridesharing
This is true, however these known criteria can be added in, thats the whole point of the heuristic in A*.
Likely one of Uber/Lyfts greatest strengths in their algos is discovering the E[X] of those heuristics. This could probably be captured from the human dispatchers as well.
This is a different problem, because the TSP is about reaching every point in the shortest distance, but in this case you can't just hit every point in any order you want. Each rider is going to want a ride at different times, and each rider is going to make their request at different times, and multiple drivers are going to be available at different times.
You are optimizing for least amount of time riders have to wait, not just for shortest distance driven.
>You are optimizing for least amount of time riders have to wait, not just for shortest distance driven.
Point being though that it's fundamentally a deterministic combinatorial problem. And they have the added benefit in this case of only worrying about prescheduled trips that need to be calculated the day before. It's not trivial, but it's an algorithmic problem that has no need for inductive output.
I don't think having "drivers spend as much of their shift as they can driving clients" is the best objective function. It seems like you should want to get customers to their destinations as quickly as possible, not have cars driving around slowly to maximize client drive time.
"You have to leave enough time in-between trips so the driver isn’t late for their next client." also seems bizarre, as you should just be routing the minimal time cost set of vehicles to those awaiting rides. This seems to imply most rides are scheduled pickups?
It's a perfectly fine objective if you trust your drivers to meet a minimum standard of a good job. It won't work in an adversarial system like Uber where drivers are just numbers but if the drivers are also invested in the success of the platform it's fine.
Given that drivers are driving the shorted distance/time route at the speed limit, treating customers with respect, arriving on time to scheduled pickups, etc. The next thing to focus on is scheduling/filling orders such that drive time is maximised.
Possibly, but today EVERY nontrivial piece of software is "AI". It's the new "synergy" for dipshits spouting nonsense.
FYI: in MBA speak, "AI" is now anything that can't be hacked together on top of airtable, outsourced, or done by someone with a GED and six months of bootcamping. If you need someone who isn't perceived as interchangeable cog labor to help build the thing then it's now called AI. Was just in a meeting where a major component for an optimizing compiler was called "the AI".
Surprisingly you can solve massive TSPs using heuristics. The solutions are not guaranteed to be optimal, but they can get very close to optimal without a crazy amount of computing power.
The Concorde TSP solver can apparently solve instances with 85.6k cities to optimality. Pretty amazing!
Sorry but how is ride share problem a TSP problem? Rides are assigned one at a time and I do not think they would take into account future rides while doing the assignment optimization.
Not my industry, but imagining that they accept rides within a 48 hour windows from now. You’d quickly get something somewhat similar to a TSP. But given that you have lots of additional constraints (people tend to have a clear idea when they want to get from A to B) compared to the classical TSP, it might actually be easier to solve. As those constraints limit the paths you must evaluate.
Being NP-hard doesn't mean it's unsolvable at all. It means it's not solvable quickly. Google maps does this millions of times per day... the number of points on the trip are few enough that you can brute force it perfectly or use a heuristic to solve it faster but less optimally. Very solvable just not optimizable algorithmicly.
A cab service with unknown, frequently updated nodes where edge weights - driving distance from last drop to next pickup- can vary, would qualify as pretty bad version of TSP.
You can come up with a good enough solution but it’s not “solved”.
This “good enough” solution starts to break down whenever there is a huge concentration of drivers in a location. If this weren’t the case, ride shares wouldn’t have had to add a cancellation fee and hidden destinations from drivers.
what if drivers painted their cars a distinct color, like yellow maybe, and they were always out there... and you could just spot them on the street and wave your arm, and they'd come get you? you wouldn't need an app for that, and wait around for them to show up.
Tried that. Didn't work well - you never knew if you could find one, they often refused trips to places they didn't like, they would cheat you on the fare, and it was impossible to get one in a less popular area without calling, and you never knew if they were actually going to come.
You must live somewhere sufficiently dense. In a midsized city life before Uber was awful for getting a late night taxi. Have to call a central line, No shows, multi-hour waits, etc.
parent gave five examples of issues they found while using a taxi in the manner you described
you could have just believed them instead of being antagonistic about it, this is especially unnecessary given the topic has a history that supports the experiences the parent poster listed
One difference I noticed is that drivers are basically employees - they sign up for a shift and take all rides assigned to them.
I tried to understand why they are almost profitable and Uber/Lyft not and as best as I can tell it's simply because Uber/Lyft are trying to grow and expand, and this co-op is not.
They are not employees because they can stop working at any point in the day and they decide to pick up rides or decline them instead of being assigned tasks they have to complete. Bad rating just means they get less priority not a bad performance review that gets them fired. They choose their tools (car), work location and take breaks whenever they want between rides.
And yes, Uber and Lyft are intentionally a "growth hack" companies.
This is why we need Unenclosable Carriers, and decentralized/distributed "apps" that cannot be de-platformed, and ...
There is a reason that Holochain was invented, and this entire conversation re-validates it.
The primary marker of Mercantilism (what we erroneously call capitalism, here in the west) is the erection of barriers to competition by entrenched players, via their purchased politicians.
When their control of the money supply, artificial barriers to competition, etc. is made much more expensive and difficult, their hand will be revealed. They will have to expend increasingly expensive (and obvious) force against ever more independent and smaller adversaries (their fellow citizens), to maintain their ill-gotten market positions.
The concept applies to so many verticals involving small companies or individual operators (from taxis, to hotels, to restaurants, to real estate agents, local bookstores, cinemas, to game developers etc. etc.)
The inability of large numbers of small entities to coordinate and cooperate in the digital era means they create the conditions for third parties to completely take over the intermediation with clients and use that leverage to extract unreasonably large rents.
In fact the extreme "taxes" small businesses and individuals have to pay to shrewd digital platform owners are multiples of the cost of development of the platforms themselves.
In other words, if a collection of independent entities cannot get they act together to codevelop digital infastructure (that will benefit all of them) somebody else will do it for them and will be eating their lunch.
This reclaiming of their market / ownership of digital tools might be challenging in conditions of rapid technical innovation. The autocracy of a top-down corporate structure is more effective than a democratic collective. Yet despite incessant marketing noise currently we are in a rather stagnating phase so there is a window of opportunity to reclaim some agency.