Hacker News new | more | comments | ask | show | jobs | submit login
Ride That Nearly Killed Me Changed How I Think About Ride-Hailing Apps (bloomberg.com)
50 points by monsieurpng 35 days ago | hide | past | web | favorite | 53 comments

What a strange article. It had plenty of emotional language that led me to believe there was some build up to a story about the driver being horribly unqualified to drive, but in the end the driver already had a taxi license.

What was the point? That we should use taxi's without apps instead? Why? Because sometimes wrecks happen and this time it happened in a ride-share?

There were some anecdotes about a driver driving for 12 hours and another driver claiming they hadn't been behind the wheel in years, but how reliable is that data I wonder?

I was hoping to learn something other than the authors bias.

> ... the driver already had a taxi license. What was the point?

The article came to a pretty clear and definite conclusion, did you read to the end? The point made is that she realized the company bore no responsibility and no damage in her accident, and these companies need to bear some of the responsibility and some of the liability for the system they’re creating. She realized the driver isn’t wholly responsible and the driver suffered some pretty big consequences himself.

“Me: left vertebral artery. Driver: livelihood + S$3,700 in fines and expenses. Grab: S$20, the refund they’d given me after the accident.”

“I’m also more aware of the profound compromises we’ve made in giving tech companies, even those with well-intentioned CEOs, so much power over our daily lives. Tan showed me kindness in the months that followed my accident. But we deserve more than expressions of remorse from companies such as Grab. Detailed safety records would be a good place to start.”

Would it be different with a taxi company?

Would a taxi customer be able to obtain anything from a taxi company beside the money paid out by the insurance? Does it matter who pays for the insurance, company vs driver? Aren't a lot of taxi driver insured themselves?

I don't see technology to call the ride (an apps vs phone), or work arrangement (cab might be an employee or a contractor to a large company vs an "independent worker" makes a difference. Would a taxi company bear more responsibilities? How?

Perhaps it would not be different. But maybe this whole time it should have been different, and the centralization provided by ride share apps like Grab now make it feasible for it to be different.

So I don’t buy this argument, and I think it’s a dangerous one. The hyperbolic, oblivious version of what you’re saying is: “well, when there were many taxi companies, it was a Wild West with no tracking and little accountability, why should that change now that the tech lets you call a cab from a phone?”

The technology we use to hail cabs provides new business opportunities, and those are actively being pursued. But also the centralization of these operations also provides new regulatory opportunities — perhaps we should be taking advantage of those too!

The author talks about compromises we made in accepting technology. I'm pretty sure she doesn't argue it should have been different.

Now, let's address your question, should it be different? Sure, but not taxi vs people driving. All drivers should be targeted, not just professional drivers. This is not like airliners vs general aviation, because those two very rarely share paths and runway, while we all share the same roads. If I die as a taxi passenger because the other driver made a mistake, I'll still be dead.

Also, deaths on the road in the western world has been drastically declining because of campaign and laws against drunk driving, enforcement of speed limits, better engineering of roads and car safety. What are you proposing here for technology to increase road safety when you are saying maybe it should be different?

People driving taxis and Ubers certainly do have reason to be vetted at a higher standard than regular drivers, since for the most part they're spending much more time driving than anyone else; this leads to a much higher risk-per-driver if one is unsafe.

This is comparable to how there are special licensing requirements, above and beyond the typical driver's license, in order to drive a truck or a bus (and, IIRC, even higher requirements to drive a schoolbus.)

I didn’t get the sense that the article was about taxi vs ride sharing, I don’t think her point depends on ride sharing being worse than taxis. They should both bear some responsibility, in my interpretation of the article.

> Would a taxi company bear more responsibilities? How?

I don’t know about Singpore, but I think the answer in many cities is yes to some degree, and the reason is that taxi companies are more established and have more regulatory oversight today. In some places the taxi company owns the car, and licenses the car and the driver, where ride sharing apps benefit from hiring people who use their own car. That is, after all, the main reason why ride share apps are currently under-cutting taxis.

I haven't been to singapore for a copule of years, but I didn't notice Uber undercutting taxis, it was just easier to get them (especially out of the city centre), and easier to expense, with a single bill at the end of the month.

Likewise in the UK, all uber drivers are licensed private hire drivers, same as Sal's taxis. It's just that they tend to be far better. The only illegality I've seen is potentially breaking the regulations I've heard exist to prevent a taxi based in stoke and thus licensed under Staffs driving me from Crewe (Cheshire) to Knutsford (Cheshire). Far better than the unlicense minicabs that used to be so common in cities on a friday night.

Isn't it about taxi vs ride sharing? The author wrote:

"I’m also more aware of the profound compromises we’ve made in giving tech companies, even those with well-intentioned CEOs, so much power over our daily lives"

How do you interpret that?

Compromises over what if this isn't taxi vs ride share? What compromises have we made here?

“we’ve made in giving tech companies, even those with well-intentioned CEOs, so much power over our daily lives. ”

This line also stuck out to me when reading the article.

It seems as if the author thinks we had a choice to make Uber/Grab/etc so wildly successful and popular. I would argue instead that the current state of the ride-sharing industry is merely a natural progression due to technological advancement (smartphones, Internet, etc.) and strong economies - it’s not 1940 and we aren’t all building military planes and making bullets in factories, after all. People (human capital) are free to pursue rather wild and crazy ideas like auto-lacing shoes and blockchain WiFi-enables dog feeders, as well as quite useful things like transportation and self-driving cars.

People say my grandfather was a dreamer. They told me he would say things like “one day, we’ll have TVs that you hang on your wall, like a picture!”. Well, we’ve done that, and even some of the stuff from cartoons seems to be coming true.

Whoops, off-track a bit! So, back to my argument. Taxi cabs could have been the ones to develop apps and put them in the hands of millions of people. Not at the scale of Uber/et al. But they of course, did not, because the smartest people in tech aren’t going to work for Yellow Cab Company of (NY/Chicago/major city), and they wouldn’t have been able to change the habits and buying patterns of millions of people around the world. Ride-sharing works because you have people who need income providing you with their driving skill and their car, anywhere in the world, just to use your app. And it’s a losing $ situation for the drivers, but many of them are too desperate to see that.

We didn’t choose to give the big Tech players power over our lives. It just happened. Technology changed the way the world operates, and thus, it changed the way we experience life and interact with each other. If it weren’t Uber, it would be another startup with a silly name. Maybe Goober. Güber.

I would argue instead that the current state of the ride-sharing industry is merely a natural progression due to technological advancement (smartphones, Internet, etc.) and strong economies

Not to mention complacency by entrenched industries -- before the Uber days, San Francisco had a half dozen competing cab companies, you'd call for one cab (talking to a human dispatcher), he'd tell you a cab was on the way, then you'd wait and hope your cab showed up, meanwhile you might see several cars from other companies drive by... and as you wait, you consider whether to ditch the cab you called and flag down one of the others.

If they'd pooled their resources, come up with a central dispatch system for all companies and some sort of app that tells you when your cab is going to arrive (even an SMS saying "Driver XXX is .5 miles away and should be there in XX minutes" would let you know that there really is a driver on the way), I wonder if Uber would have taken over there.

Yeah, I found the article to be very confusing too. The author seems to be upset that the driver and the passenger lost a lot of money/ body parts, but Grab did not lose anything. Sure, but the author fails to explain why Gab should lose anything?

It's like complaining that my Prius had an accident with somebody's Tesla, but Toyota and Tesla CEOs did not lose anything. Why should they?

Maybe the company should have insurance if something bad happens? But if the law does nor require this then I am sure this companies would try to shift the blame and have the driver pay.

It’s more like complaining that you were in an accident that was the fault of a Toyota employee, but Toyota didn’t lose anything.

Should a company be liable for losses caused by their workers while they’re on the job? Seems to me like the answer is yes, at least most of the time.

Sal's cars, based in Northwich in the UK, charge drivers a 'radio fee', and send cabs their way. They don't employ the drivers, they don't own the cars.

I struggle to see how Uber is any different to Sal's cars.

In the UK, Uber requires its drivers to have taxi insurance. In addition, it has 'backstop insurance', just in case an uber driver had lost their insurance / license.

I'd far rather be in an Uber in the UK than a typical minicab in the UK. Or a typical yellow taxi in new york.

Do I, as a customer, need to carefully scrutinize the employment relationship used for every service I use?

Let’s say I call a plumbing company to fix a sink and the plumber floods my house. The company tells me, “That plumber is an independent contractor not employed by us. You’ll have to get restitution directly from him. Good luck, bye!”

I certainly wouldn’t accept this. I didn’t sign up to have some unknown third party do the work and be responsible for the outcome, I signed up to have the company I’m paying do it.

Imagine if your car is defective and the manufacturer tells you to get restitution from the subcontractors that made the part. Or if you bought a sandwich full of arsenic and they told you to talk to some random farmer who happened to grow the affected vegetable.

These companies want to have it both ways. They want to completely own the relationship with the customer... until something goes wrong. Then it’s all, “we’re just a broker for independent contractors, they’re responsible for everything, don’t talk to us.”

There’s nothing wrong with just being a broker, but you have to be consistent.

> The author seems to be upset that the driver and the passenger lost a lot of money/ body parts, but Grab did not lose anything. Sure, but the author fails to explain why Gab should lose anything?

One argument for Grab to be responsible would be that Grab is probably in the best position to estimate the frequency and severity of such accidents and buy insurance to cover them.

Riders probably don't have the information necessary to specifically buy insurance for this, and so if riders have to use their own insurance they will have to make it part of some general purpose insurance which won't be as efficient and will spread the costs among everybody instead of just people using ride services.

This past week I rode a lyft to a tech meetup in Atlanta. The drive was about an hour away, which meant that the fastest route was through the interstate (I75).

The car was a 4 year old Toyota Camry. It looked very nice from the outside and was very clean inside. The driver was a very friendly man who drove very carefully. Rather surprising here in Atlanta.

However, the one thing that made the ride highly dangerous:

The car had a blown strut (shock absorber) on the front right side. It was not very noticeable at low speeds. There were little vibrations coming through the floorpan but it could have easily been the tires needing a rotation or slightly out of balance.

When we got into I75 that's when I was scared for my life. Doing 70+ MPH on a car with a blown front strut is very dangerous. It vibrated and shook as if it was going to break in two any moment. I told the driver to slow the fuck down because we would have an accident. He did. Then I explained why. The driver mentioned being in an accident a month ago. He had fixed the cosmetics, but not the mechanicam because "a friend told him it was ok".

I made it to the tech meetup in one piece. Yes, I could have requested another car, but decided to not stop on an unknown area of Atlanta.

Lyft must make sure the cars are in good mechanical condition. It must also provide a way to report this sort of incident directly.

How do I know about what was wrong with the car? Im a master mechanic who used to specialize and work for Toyota.

This incident made me realize Im better off driving downtown. Which is a bit insane given how awful people drive there.

I've had similar experiences with Uber and Lyft in NYC. I was once in a Toyota Camry that had suspension so bad the car bottomed out on every single bump on the highway. I'm not a master mechanic so I don't know if it was unsafe, but it sure felt unsafe.

I also had a crazy experience with a Lyft in Seattle where I got a driver who must have been a brand new arrival to the country, couldn't speak english, and didn't know how to drive. He literally didn't know traffic laws, tried to make a left turn into oncoming traffic, and almost killed us. Because of the language barrier I couldn't explain to him what he did wrong. I called Lyft safety after that ride and told them the driver needs to be removed from the platform.

At least with taxi companies the licensing requirements force them to do regular safety inspections of their vehicles and require drivers to have some training.

How would this be any difference with a legacy taxi company?

I understand where you are coming from, but the question is flawed. All taxi / ride companies must meet some kind of safety and maintenance standards. No matter who it is.

Looking back and saying "Taxi companies were also as bad" is a terrible way to dismiss the need for regulation.

These contractors should have their cars inspected every n months or n amount of miles by a certified shop.

Lyft should also provide a way to allow users to report safety / mechanical problems and temporarily remove that car from service (not the driver) until it is fixed.

In the UK taxis have a more rigorous safety test each year (more rigorous than the normal safety test all UK cars over 3 years have to have), but that wouldn't help in a situation where it was in an "accident a month ago"

Not sure about lyft, but Uber certainly does allow you to report such issues -- certainly more easilly than the majority of taxi or private hire companies.

Again in the UK you can always simply complain to the licensing authority -- easy to do with uber (you have the registration and driver's name), but fairly easy to do with http://www.wiltshire.gov.uk/taxicomplaintform.htm

Despite all this, entrenched taxi drivers still complain about uber in the UK ("it's too popular and causes congestion and pollution, unlike black cabs with their express lanes for rich people to bypass the traffic")

Legacy taxi companies are required to have insurance and their owners can be held criminally responsible for not maintaining their vehicles.

Depends on the jurisdiction, in the UK Uber are exactly the same as a legacy taxi company and have to have the exact same insurance coverage. They're just more likely to do so.

The CEO himself sent texts to the article's author, and visited her several times. How many people get a visit or a text from Uber's CEO when there's an accident, when a disabled person is refused service because of their assistance dog, etc? Not many, because they're not journalists. If a company lays on everything they've got when they know the subject of an accident is a prominent journalist, but that person ends up thinking they should be doing more regardless... that makes it pretty clear these companies can do better.

While I feel sympathy for the writer, there are few items of note before applying this beyond Singapore:

- Believe all drivers for hire in the US are required to have mandatory liability insurance; that would address at least part of the financial consequences of the accident.

- Don't the market leading US ride-share apps track driver performance and post ratings?

- Driver ride ratings are likely a poor predictor of severe events; you probably need something more along the lines of hard breaking incidents per mile (used to predict issues for trucks and service vehicles). IOT can get you this.

I think more importantly, the driver already had a legally issued cab license... if he wasn't driving for a ride-share company, he would have been driving a regular cab. The risk management failure was at the state level not the company.

I was looking for some type of uber vs traditional cab comparison and couldn't find any justification for the headline. The quote on your last point is:

> I asked him how he’d passed Grab’s driver-screening process. He said there was no issue, because he already had a taxi driver’s license.

Well, that's pretty much the end of the story then. At least, the story that I thought it was based on the headline. Driving is dangerous. The fact that an accident occurred in a car hired via app vs traditional street side hailing or telephone call doesn't seem very relevant.

Re hard braking: Smooth driving may be a nice-to-have, rather than an clear indicator. Or, to put it another way, if there is a useful signal in apps that score driving smoothness and lack of distraction from phone calls and texts, extracting that signal from the noise of whether the driver frequents dangerous roads, drives at night, drives fatigued, intoxicated, is using a second mobile device, in bad weather, etc. isn't easy. Or, to put it yet another way, the real signal isn't in the score provided to the driver, it may be something the driver would consider unfair and outside their control in many cases.

These apps influence drivers to drive more smoothly, which is useful for ride hailing since it improves the perceived quality of the ride, but that's much less directly connected to safety.

is using a second mobile device

Initially I was like “versus using a single mobile device? What?” But I see what you’re saying here... the apps should be able to prevent or detect use by drivers while in motion.

On the other hand, this prompted some thought about how distracting in-car entertainment systems can be. The late 2000s early touch-screen and some of the non-touch screen systems are particularly horrible, like the early BMW iDrive and its predecessor. Changing the air conditioning settings in those cars is a UX nightmare.

How are hard braking incidents used in the other industries? Too many and the driver is canned?

For the use case I saw, it was one of several variables fed into a statistical modeling engine that predicted various forms of trouble. Mostly safety and employee retention.

If I recall correctly, the absolute value was less predictive than watching for changes and relative performance vs. peers. It's a good signal that things aren't going smoothly behind the wheel; sometimes is may just be the time and place, in other situations you could have a driver issue. The usual intervention was to trigger someone back at dispatch to get the driver on the horn and ask them if everything was ok... "whoa cowboy, you all right?"

Long haul trucking is an interesting place; everyone hates it, pretty much, since you're away from the family for weeks at a time. Turnover rates approach 100%. Bad driving metrics often can be linked to other forms of stress which prompt an unsafe attitude behind the wheel.... I think they wove sentiment analysis into the product as well, looking at what the driver said. Recommended intervention is to get the driver on the horn and ask them if they needed a little help; apparently that tended to cure a good share of the problems...

Damn, that sounds like a good response.

I travel a lot - typically two weeks each month. I view Uber as very risky compared to cabs. I once had an Uber driver whose phone got flaky during the trip, so he said that he would drop me off (he turned into an industrial park area that was completely deserted to leave me.) He said that if Uber didn't see him complete the trip then he wouldn't get paid and would be working for free, so he wouldn't take me any further than where we happened to be. Another time I had a 1 mile commute in 115 degree heat. Well, no Uber driver would come to pick me up for a one mile commute. I tried and tried for over an hour, while other people were catching Ubers with no problem. Another time I had to go to the next town over, and nobody wanted to pick me up for that trip, so I again found myself stranded by using Uber. Then there is the fact that so many Uber drivers are obviouly just barely surviving financially, and all that entails.

I've been taking taxis for decades - and never had any problems in the United States.

Cabbies aren't much better. I had a cab driver once that apparently didn't realize the gas and brake pedals had a range of motion. He stomped on the gas pedal to the floor to go, and stomped on the brake to the floor to slow down. And so we went down the street every 100 feet, accelerating like mad then screeching the tires to stop, over and over again.

An Uber driver did this with me recently. It was late at night (bar pickups). The driver was an older, professional looking gentleman but told me he was tired and “had not driven in a few months”.

That's code for "I've never driven before today and I don't have a license, but my family member let me login to his Uber account and drive his car..."

I don’t really understand what this article is trying to accomplish. Most of it is purely about the emotional impact of a car accident and realistically that’s going to be largely the same even if you were the one driving the car, it follows the same basic grief cycle as any tragedy.

> Singapore’s government requires drivers for services such as Grab to be licensed, a process that includes a background check, medical screening, classroom instruction, and a written test. The city-state used to require 60 hours of training for taxi drivers to earn a vocational license. That’s now 25, which is still longer than the 10-hour course required for private-hire car drivers. In those 10 hours—two of which can be done outside the classroom as “self-study”—applicants are supposed to learn all they need to know about service quality and road rules.

> Grab notes that Singapore law forbids the company access to driving or criminal records, and that authorities wouldn’t have issued a vocational license if they judged Chia to have a poor record.

Those two quotes pretty much sum up the rest and, at least to me, paint a different picture than the one she drew. Even if you somehow single out a ride hailing service as being different than an actual taxi or driving yourself / carpooling with a friend / etc. (which hardly seems to be justified in this case) it seems like Grab followed the law and if anything there is blame to be placed upon the government of Singapore by restricting their access to additional information.

> But we deserve more than expressions of remorse from companies such as Grab. Detailed safety records would be a good place to start.

I’m not sure how healthcare works in Singapore, but there was no mention of medical bills, so I’m curious what exactly Grab could or should have done above and beyond having their CEO visit her multiple times?

> Grab says that at the time of the accident, Chia had a valid license and had completed more than 500 rides on the platform with a good passenger rating, and his record was spotless.

So the driver had a spotless record while at Grab and the company has no access to his prior driving history. Exactly what other detailed safety records should they start with?

Even though this driver had a license, many drivers in other countries where Uber operates do not. Part of why classic Uber is not in Germany is because you need an extra license to be allowed to drive passengers around for money, this includes deep knowledge of the area you are working in and a bunch of other tests and checks.

I think it makes sense that not just anyone is allowed to do this, even though i see a lot of cab drivers driving pretty recklessly, but in general they at least seem very experienced and the cars are usually very modern and safe.

There's one major problem with apps of any sort of scale that involve any scenario where dangerous accidents are possible. The problem is that as these apps scale, these accidents will happen, and with a degree of regularity that also scales alongside the growth of the app/company - even when there is absolutely 0 negligence involved. And the media then takes these anecdotal bits and implies they are representative of the service itself, further implying negligence.

In this case the article mentions that the driver was an actively licensed taxi/vocational driver that had been driving taxis on and off for 40+ years. And had completed more than 500 flawless rides on Grab in his month of work with them, with a good overall review rating. The only thing the author mentions as something trying to be framed as a red flag is that in his decades of driving, the driver states he had "several small and big road accidents."

Let's look into that. This [1] article from the NYTimes indicates that taxis get into crashes about 4.6 times per million miles driven, which was substantially better than the rate for non-taxi drivers. Another search indicated that taxi drivers hit around 70k miles per year. [2] So we now have the numbers we need. 1 million / 70k / 4.6 = 1 crash per 3.1 years. That means a perfectly average taxi driver gets into a crash every 3.1 years of driving on average. A taxi driver that's been driving for 40 years would expect to get into an average of about 13 crashes. Given that the article mentions our driver drove taxis "on and off", his accounting of "several" means he was probably just about perfectly average.

The point of this is that anecdotal evidence is irrelevant. I suppose like the old saying goes, Dog bites man is not news. Man bites dog is news. Yet when all you see from media is reporting these incidents of man bites dog, it leads to a distorted worldview where there seems to be a rash of men biting dogs.

[1] - https://www.nytimes.com/2006/04/28/nyregion/that-wild-taxi-r...

[2] - https://www.quora.com/How-many-miles-does-a-NYC-taxi-do-in-i...

One thing I notice is she says the ride app GPS took the driver through local roads instead of the highway.

I had a similar experience with Uber recently.

There is a road that goes pretty much straight from the airport to my house, though it's a busy, slow road. Instead the uber gps had us going all over the place, through many intersections and left turns. It might have saved 1 or 2 minutes, but surely cost way more in gas (and CO2) and was much more likely to result in an accident. I felt bad for the driver, having to drive such a confusing and stressful route.

I'm not sure if it made my route more expensive. The cost is set before we leave, so any route has the same cost. But the drivers are paid more the farther they go, so it costs uber more to pay him, when he could have gone a way that's much shorter and doesn't take any longer.

How closely do Uber drivers have to follow the suggested route? Do they get penalized if they often don't follow it?

We're generally not very good at handling market externalities like this. Injuries. Housing. Pollution. Corruption. We tend to think the market will sort everything out. The question is, how many people have to die or suffer while it corrects itself?

The title is misleading; the author was using a ride sharing app, not a cab.

In Singapore, Grab drivers need a commercial license. The driver in this case had a taxi license. I'd argue that "cab" is not a misnomer here.

There is a deliberate attempt to conflate the two, since they are essentially the same service, and many people want them regulated the same.

There is a correspondingly deliberate attempt to ensure that the two remain separate in peoples minds, to ensure that Grab, and the other "* sharing" apps, don't get the same regulations applied to them as the industries that they are trying to disrupt.

But make no mistake, the services that Grab/Uber/Lyft/other "ride sharing apps" provides is for 99.99% of consumers a direct, possibly cheaper, replacement for a cab service.

> the services that Grab/Uber/Lyft/other “ride sharing apps” provides is for 99.99% of consumers a direct, possibly cheaper, replacement for a cab service.

I think that, reading between the lines, this article is talking about that point. I know cab services are crappy most places, but ride sharing apps are cheaper precisely because they’re trying to build a market without any regulation and they’re undercutting cabs in part because cabs cost more. The business innovation of ride sharing apps is in externalizing the costs of ride sharing, from the maintenance costs of cars onto the drivers, to the lack of unions & benefits, to the “disruptive” startup mentality where Uber justifies breaking laws because it’s good for business, to lower regulatory oversight and fees.

Taxi drivers in the US, where Uber/Lyft started, were already considered contractors without unions & benefits, and they paid to rent the cars with medallion from cab companies (they started each day over $100 in the red).

In Singapore, at least while I was living there, Grab taxi was used to hail actual taxis. Think traditional taxi company, but happens to have a mobile app that (mostly) works well. I guess now that they bought up Uber’s operations it’s a mixture of traditional and Uber style freelancers.

Yeah I'm not sure why the author chose that title, it makes you think its going to be a "pro-ride sharing/anti-traditional taxi" piece. I wonder if this is a click-bait strategy.

Could be an allusion to the fact that the driver was a licensed taxi driver.

I don’t know about Bloomberg specifically, but typically authors don’t write the headlines in media.

If you wanted ride sharing to be treated the same as traditional taxis, you could erase the difference in people's minds by referring to them as the same thing.

I see they changed the title now.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact