The claim that “more than 33,000 lives would be saved annually in an entirely Waymo’d world” appears stronger than the evidence presented. The cited statistics concern injury-causing crashes, not fatalities, so it’s unclear how they support an estimate of lives saved. In addition, Waymo operates in a limited range of locations and driving conditions compared with the full spectrum human drivers face. Extrapolating its current performance to a nationwide estimate therefore depends on assumptions that are not explained here.
You're right -- these are all great points. The purpose of my back-of-the-napkin math (while imperfect) was to attempt to quantify the very real impact that autonomous driving will have on road safety. It's the part of this conversation that is frequently overlooked.
You're focusing on uncertainty in the exact figure while overlooking the broader evidence that autonomous driving substantially reduces serious crashes. Even if 33,000 is overstated/inaccurate, the underlying conclusion is still be well supported.
The longshoremen comparison skips over something though: a negotiated exit only helps if you actually have time to use it.
Not all displaced drivers are in the same boat. A 27-year-old driving Uber between gigs has options. A 54-year-old who took out loans to buy a taxi medallion, raised a family on that income, and never had a reason to think the floor would drop — that’s a different conversation entirely. By the time any retraining fund shows up, that person might already be underwater.
The rug pull is the whole problem. You can catch yourself when something shifts — but it’s a lot harder when you’re older, locked into debt, or just weren’t watching because you’re too busy trying to keep your head above water. These jobs were sold as a real path to the middle class. People built their lives around that promise.
The Teamsters might be fighting the wrong battle tactically. But writing off the resistance as pure Luddism misses why some people are scared — it’s not that they don’t understand technology wins. It’s that for older workers who can’t realistically pivot, slowing it down is the strategy. Not because they think they’ll stop it. Because they’re just trying to outlast it.
They can still keep their Taxis, their service will simple become a differentiated premium service, what an absurd comparison to make. So for the sake of sealing a few people in their forever jobs we're preventing a revolution in mobility?
Here in Austin, Waymo's drive slowly through neighborhoods all night without passengers. When homeowners call Waymo to complain, they just move to another neighborhood. Until we have independent stats, it's hard to know how many of those miles are in a risk-free culdesac at 2am.
I think Mr. Chiolan may be missing the point regarding the opposition to autonomous vehicles and to AI more generally. To most people, it is not just the fact that millions of drivers are at risk of being replaced by an AV; it's the fact that the rollout of this technology will be paid for by the lowest earning members of our society while the profits will redound the most well heeled. It's such a typical pattern in a capitalist economy. His solution is to require the AV companies to contribute to a fund to ameliorate the effects on the displaced workers. Such a "solution" seems totally inadequate when one expands his perspective slightly. It seems that all workers are "at risk" in the modern American economy; first, they will take the drivers' jobs, then they will take the programmers', lawyers', accountants', and doctors' jobs. Soon no one will have a job. And the richest of the rich will continue to get richer, with no thought for those displaced by the new technologies that are making them richer. We need a wider, more encompassing solution to these problems. Progressive taxation of income and WEALTH, universal basic income and an otherwise expanded social safety net, universal health coverage, and a society and an economy that generally support social and economic security for the greatest number, not just for those at the top. The rich are eating us alive, and it must stop.
A problem is that we never should have needed the gig economy in the first place. (I worked multiple of these jobs including rideshare driving for Uber & Lyft. It's hard work that doesn't pay well.)
Now, we have to face the reality that these are people whose livelihood depends on "bullshit jobs" (Graeber) to not-quite-get-by. As many have pointed out, we have a massive shortage in good jobs. Jobs that make people feel a sense of purpose. This is unacceptable.
Just about every day now, I hear references to "financial nihilism" and fears of a "permanent underclass". This is the world we live in. This is the state of the economy for real Americans.
Reiterating what I said about jobs with a sense of purpose... we will lose an unknown number of jobs through AI automation in the near-term but the hope is that, as with most new boogey man technologies, that the long-tail will create more jobs. Doesn't seem entirely realistic to prophesize the reality. Some will say, "It's different this time." (and we know how that goes, historically, with regards to the market... just sayin)
Reid Hoffman (in his book 'Superagency') noted we could do Americans a big favor by making this newer iteration of the Industrial Revolution (the AI Revolution) easier on everyday people. Many have said that the only way to do this is a form of Basic Income. (Separately, Scott has suggested we give individuals a living wage (at least $25/hour)) With regards to Basic Income, Scott has suggested a lower amount than some and then adding services (such as childcare) that brings individuals up to ~$80,000/yr.
Much more to this conversation but I'm glad Scott and the Prof G team take deep dives and force us to look at hard realities.
Driving safety is wrong way to look at the problem but jobs. We should look at what other gig jobs can be created if Waymo is offering rides. I am sure there are other jobs available if they are not all driving. Perhaps doing delivery on autonomous vehicles or other blue collar jobs. We would help the situation by providing employment ideas, or bring in the businesses that may need such work force.
Another safety factor would be the premise that as the number of Waymos and Self Driving vehicles increase, their overall safety factor would increase. It would be interesting to see the number of Waymo accidents attributed to human drivers.
Mamdani is an idiot ideologue that panders to a certain demographic but forgets the reason why cities like London or New York became the success they are. And it's got nothing to do with worker protections.
Or all the fletchers or all the charcoal producers. Imagine where New York would be. Anti progress demagogues want to freeze people in the same jobs forever like little communists instead of helping train and distribute obsolescence into nicer and better things.
Worker protection is somehow a bad thing? 146 people died a horrible death in that fire. London's Matchgirl's strike helped stop something far far worse. The world is a better place when there are limits on how badly employers can act. Ideologues who place profit over people like little neoliberals, want to cut employee salaries out of their balance sheets entirely. Pandering to the 1% won't make anything nicer, or better.
Thanks, Prof G for this clear-headed analysis. All great points, but this one stands out:
"The more honest and useful fight is the one the longshoremen eventually settled for: not blocking the technology altogether but demanding that the companies deploying it bear some of the cost of the disruption they’re causing."
Technology has always been and always will be, a two-edged sword. The concept itself is disruptive. Speeding up a human activity by definition puts people out of work. Machines that do the work of humans will displace them. The simple economics dictate that. Gutenberg put a bunch of scribes out to pasture. And now we must factor in, as you did, the broader human impact of the tech, and the safety versus human operated machines reveals the same issue that's always been there: what is the real benefit of this machine versus putting a human out of work?
I believe the outcome of this debate is inevitable. You can't put the genie back in the bottle. Future humans will not be driving anything, and they almost certainly won't own a car. They'll summon one when they need it.
So yes, make the big tech companies pay for the retraining needed to help displaced workers find gainful employment elsewhere.
But while we're at it, why not apply this same logic to other more pressing issues such as our climate and the death by plastic that we are allowing companies to keep producing unchecked. Make plastic too expensive to use, to force people to invent better alternatives, clamp down on the oil and coal industries to make them help clean up.
My personal view of all this is defined by this acronym: GLAG. Greed, lust, ambition and gluttony are the heartbeat of the world. Flub-dub, flub-dub and we'll keep beating on, until enough of us stand up and say, Not on my watch. Has to stop.
The advent of business office apps and suites, especially Word and Excel, displaced millions of typists, file clerks, and bookkeepers. This unemployment "crisis" occurred over decades, with many industries slower to adopt than others. New tech job displacement should be evaluated by its speed, the sectors affected, and the quality of alternative jobs created in the broader economy.
Technology did not win as far as nuclear power plants. If you look at the risk factor, nuclear power is incredibly safe, but people do not want it.
Isn't it to be expected that technology brings cheaper services for us individual consumers? Does Waymo cost less?
I'm all for the safety argument, but let's apply that evenly and favor funding mass transit. Isn't the lobbying of the auto industry monumental and against mass transit funding? Show me the numbers on that one. The public funding of roads is at the top of budget items for ever unit of local government and states. We never talk about how much safer mass transit is and question why we continue to fund the more deadly mode.
I'm referring to the decline in use starting in the 1960s. There was adequate evidence that nuclear was a much better choice. It was shut down not by evaluating risk but by citizen action.
was it cheaper than gas/oil/coal at the time? I asked chatgpt, it says the cost is actually not better than traditional fuels + it needs a bigger upfront cost which make it not competitive in the market.
"The pattern is undefeated: Technology always wins..."
Is this not a rather deeply nihilistic stance--not to mention glib, and conveniently resigned?
Technology always "wins" WHAT, exactly? "Wins" for whom?
Looking at the current state of our small world and humanity's future in it (i.e., living on borrowed time and resources, while nevertheless incessantly accelerating and declaring "technology will (somehow!) save us!" despite technology being the proximate cause of the planet's degradation)...seems to me that this POV is precisely the kind of existential shrug that accounts for our very worst qualities as a species.
Consider at least the *possibility* that we actually have choices. Even we're too numbed--or voluntarily distracted--to see them. ¯\_(ツ)_/¯
I do believe that we have choices! In this case, I think the choice in front of us is how we respond to autonomous driving: let it run unfettered, think critically about how to beneficially deploy it, or attempt to will it out of existence. What I'm arguing against is that third option, which seems to be gaining traction.
Except that all these industries have fought tooth and nail against any kind of regulation and spent billions lobbying politcians to push it through. Are we to believe they are doing that out of altruism?
The claim that “more than 33,000 lives would be saved annually in an entirely Waymo’d world” appears stronger than the evidence presented. The cited statistics concern injury-causing crashes, not fatalities, so it’s unclear how they support an estimate of lives saved. In addition, Waymo operates in a limited range of locations and driving conditions compared with the full spectrum human drivers face. Extrapolating its current performance to a nationwide estimate therefore depends on assumptions that are not explained here.
You're right -- these are all great points. The purpose of my back-of-the-napkin math (while imperfect) was to attempt to quantify the very real impact that autonomous driving will have on road safety. It's the part of this conversation that is frequently overlooked.
You're focusing on uncertainty in the exact figure while overlooking the broader evidence that autonomous driving substantially reduces serious crashes. Even if 33,000 is overstated/inaccurate, the underlying conclusion is still be well supported.
The longshoremen comparison skips over something though: a negotiated exit only helps if you actually have time to use it.
Not all displaced drivers are in the same boat. A 27-year-old driving Uber between gigs has options. A 54-year-old who took out loans to buy a taxi medallion, raised a family on that income, and never had a reason to think the floor would drop — that’s a different conversation entirely. By the time any retraining fund shows up, that person might already be underwater.
The rug pull is the whole problem. You can catch yourself when something shifts — but it’s a lot harder when you’re older, locked into debt, or just weren’t watching because you’re too busy trying to keep your head above water. These jobs were sold as a real path to the middle class. People built their lives around that promise.
The Teamsters might be fighting the wrong battle tactically. But writing off the resistance as pure Luddism misses why some people are scared — it’s not that they don’t understand technology wins. It’s that for older workers who can’t realistically pivot, slowing it down is the strategy. Not because they think they’ll stop it. Because they’re just trying to outlast it.
They can still keep their Taxis, their service will simple become a differentiated premium service, what an absurd comparison to make. So for the sake of sealing a few people in their forever jobs we're preventing a revolution in mobility?
re: "Number of Incidents per Million Miles"
Here in Austin, Waymo's drive slowly through neighborhoods all night without passengers. When homeowners call Waymo to complain, they just move to another neighborhood. Until we have independent stats, it's hard to know how many of those miles are in a risk-free culdesac at 2am.
I think Mr. Chiolan may be missing the point regarding the opposition to autonomous vehicles and to AI more generally. To most people, it is not just the fact that millions of drivers are at risk of being replaced by an AV; it's the fact that the rollout of this technology will be paid for by the lowest earning members of our society while the profits will redound the most well heeled. It's such a typical pattern in a capitalist economy. His solution is to require the AV companies to contribute to a fund to ameliorate the effects on the displaced workers. Such a "solution" seems totally inadequate when one expands his perspective slightly. It seems that all workers are "at risk" in the modern American economy; first, they will take the drivers' jobs, then they will take the programmers', lawyers', accountants', and doctors' jobs. Soon no one will have a job. And the richest of the rich will continue to get richer, with no thought for those displaced by the new technologies that are making them richer. We need a wider, more encompassing solution to these problems. Progressive taxation of income and WEALTH, universal basic income and an otherwise expanded social safety net, universal health coverage, and a society and an economy that generally support social and economic security for the greatest number, not just for those at the top. The rich are eating us alive, and it must stop.
A problem is that we never should have needed the gig economy in the first place. (I worked multiple of these jobs including rideshare driving for Uber & Lyft. It's hard work that doesn't pay well.)
Now, we have to face the reality that these are people whose livelihood depends on "bullshit jobs" (Graeber) to not-quite-get-by. As many have pointed out, we have a massive shortage in good jobs. Jobs that make people feel a sense of purpose. This is unacceptable.
Just about every day now, I hear references to "financial nihilism" and fears of a "permanent underclass". This is the world we live in. This is the state of the economy for real Americans.
Reiterating what I said about jobs with a sense of purpose... we will lose an unknown number of jobs through AI automation in the near-term but the hope is that, as with most new boogey man technologies, that the long-tail will create more jobs. Doesn't seem entirely realistic to prophesize the reality. Some will say, "It's different this time." (and we know how that goes, historically, with regards to the market... just sayin)
Reid Hoffman (in his book 'Superagency') noted we could do Americans a big favor by making this newer iteration of the Industrial Revolution (the AI Revolution) easier on everyday people. Many have said that the only way to do this is a form of Basic Income. (Separately, Scott has suggested we give individuals a living wage (at least $25/hour)) With regards to Basic Income, Scott has suggested a lower amount than some and then adding services (such as childcare) that brings individuals up to ~$80,000/yr.
Much more to this conversation but I'm glad Scott and the Prof G team take deep dives and force us to look at hard realities.
Just to note that the data on how safe Waymo is comes from… waymo
“If it saves just one life …”
Damn the full torpedoes costs — full speed “ahead.”
These sentimentalities always founder on cooked books foundations which are actually hull-rending reefs.
And already too fragile automatons can’t ever sate their appetite disorders for more automation … & fat-melting syringes.
Driving safety is wrong way to look at the problem but jobs. We should look at what other gig jobs can be created if Waymo is offering rides. I am sure there are other jobs available if they are not all driving. Perhaps doing delivery on autonomous vehicles or other blue collar jobs. We would help the situation by providing employment ideas, or bring in the businesses that may need such work force.
The problem is that it’s the wrong technology.
The other problem is the cheering section I’m seeing here for capital, capital, capital over everything.
Another safety factor would be the premise that as the number of Waymos and Self Driving vehicles increase, their overall safety factor would increase. It would be interesting to see the number of Waymo accidents attributed to human drivers.
Mamdani is an idiot ideologue that panders to a certain demographic but forgets the reason why cities like London or New York became the success they are. And it's got nothing to do with worker protections.
If they'd only let the Triangle Shirtwaist Factory open more firetraps, imagine how successful New York could be!
Or all the fletchers or all the charcoal producers. Imagine where New York would be. Anti progress demagogues want to freeze people in the same jobs forever like little communists instead of helping train and distribute obsolescence into nicer and better things.
Worker protection is somehow a bad thing? 146 people died a horrible death in that fire. London's Matchgirl's strike helped stop something far far worse. The world is a better place when there are limits on how badly employers can act. Ideologues who place profit over people like little neoliberals, want to cut employee salaries out of their balance sheets entirely. Pandering to the 1% won't make anything nicer, or better.
Thanks, Prof G for this clear-headed analysis. All great points, but this one stands out:
"The more honest and useful fight is the one the longshoremen eventually settled for: not blocking the technology altogether but demanding that the companies deploying it bear some of the cost of the disruption they’re causing."
Technology has always been and always will be, a two-edged sword. The concept itself is disruptive. Speeding up a human activity by definition puts people out of work. Machines that do the work of humans will displace them. The simple economics dictate that. Gutenberg put a bunch of scribes out to pasture. And now we must factor in, as you did, the broader human impact of the tech, and the safety versus human operated machines reveals the same issue that's always been there: what is the real benefit of this machine versus putting a human out of work?
I believe the outcome of this debate is inevitable. You can't put the genie back in the bottle. Future humans will not be driving anything, and they almost certainly won't own a car. They'll summon one when they need it.
So yes, make the big tech companies pay for the retraining needed to help displaced workers find gainful employment elsewhere.
But while we're at it, why not apply this same logic to other more pressing issues such as our climate and the death by plastic that we are allowing companies to keep producing unchecked. Make plastic too expensive to use, to force people to invent better alternatives, clamp down on the oil and coal industries to make them help clean up.
My personal view of all this is defined by this acronym: GLAG. Greed, lust, ambition and gluttony are the heartbeat of the world. Flub-dub, flub-dub and we'll keep beating on, until enough of us stand up and say, Not on my watch. Has to stop.
The advent of business office apps and suites, especially Word and Excel, displaced millions of typists, file clerks, and bookkeepers. This unemployment "crisis" occurred over decades, with many industries slower to adopt than others. New tech job displacement should be evaluated by its speed, the sectors affected, and the quality of alternative jobs created in the broader economy.
Technology did not win as far as nuclear power plants. If you look at the risk factor, nuclear power is incredibly safe, but people do not want it.
Isn't it to be expected that technology brings cheaper services for us individual consumers? Does Waymo cost less?
I'm all for the safety argument, but let's apply that evenly and favor funding mass transit. Isn't the lobbying of the auto industry monumental and against mass transit funding? Show me the numbers on that one. The public funding of roads is at the top of budget items for ever unit of local government and states. We never talk about how much safer mass transit is and question why we continue to fund the more deadly mode.
Nuclear is not winning because there is an even better tech called solar which is more economical
I'm referring to the decline in use starting in the 1960s. There was adequate evidence that nuclear was a much better choice. It was shut down not by evaluating risk but by citizen action.
was it cheaper than gas/oil/coal at the time? I asked chatgpt, it says the cost is actually not better than traditional fuels + it needs a bigger upfront cost which make it not competitive in the market.
Great piece! Have to think about where I stand but appreciate the perspective
Thank you!
"The pattern is undefeated: Technology always wins..."
Is this not a rather deeply nihilistic stance--not to mention glib, and conveniently resigned?
Technology always "wins" WHAT, exactly? "Wins" for whom?
Looking at the current state of our small world and humanity's future in it (i.e., living on borrowed time and resources, while nevertheless incessantly accelerating and declaring "technology will (somehow!) save us!" despite technology being the proximate cause of the planet's degradation)...seems to me that this POV is precisely the kind of existential shrug that accounts for our very worst qualities as a species.
Consider at least the *possibility* that we actually have choices. Even we're too numbed--or voluntarily distracted--to see them. ¯\_(ツ)_/¯
I do believe that we have choices! In this case, I think the choice in front of us is how we respond to autonomous driving: let it run unfettered, think critically about how to beneficially deploy it, or attempt to will it out of existence. What I'm arguing against is that third option, which seems to be gaining traction.
Except that all these industries have fought tooth and nail against any kind of regulation and spent billions lobbying politcians to push it through. Are we to believe they are doing that out of altruism?