I'm often frustrated by the extremely cautious attitude some people have toward technology with the potential to have negative consequences, when it seems much more likely to create positive ones.
For example, self-driving cars and the possibility of an accident a human would have prevented.
I think sometimes we need to be a little bit better at accepting risk when it comes with positive things.
I have the attitude you are talking about, and the source of it is not defiance of technology. It's coming from the experience that humans are morons that will not weight the consequences of their actions and destroy things on their way. And once the pandora box is opened, the energy required to close it is tenfold the one that was required to open it, meanwhile the destruction continue.
We knew asbestos was dangerous since freaking 1920s. We had to wait till the las decades to eventually deal with it.
Do I think AI, nano techs and GMO can be fantastic opportunities ? Sure I do. Do I think the people that wants to make money with it can be trusted to ensure they will not destroy humanity ? I certainly don't.
The more powerful a technology is, the more precautions you need to take.
GMO, AI, and nano techs are incredibly powerful technologies, and misuse have the potential of cascading consequences of such a scale that it's only natural, as a human being, to be extremely concerned with it.
And yes, it should be so regulated it's not even funny.
Read up on black swans. There are risks that can destroy everything that has come before it. These are worth defending against and building resilience against. Messing with genomes has unknown unknown risks that could turn into black swans. Tread lightly where there be dragons.
Its true black swans are unpredictable but we can know to some degree which systems have risks that can eliminate the entire system. Biological systems are one of those because we've seen plagues and viruses come close. Financial systems as because of interdependence. What the article describes is an external modification to the gene line, which if you follow out a few generations, has permanent consequences on the entire gene pool.
This is overblown. Nearly all "the entire gene pool" lacks this particular mitochondrial defect. The population who have it, have a great deal of trouble reproducing. This procedure allows them to reproduce without passing on the defect. Unless some other trait has been previously limited to this population and will now be passed on to viable descendants at a higher rate, the effect is very small. Unless that particular trait is so advantageous that those descendants will outcompete all other humans, it certainly won't have "consequences on the entire gene pool."
If you have some time, The Gene [0] is a good read. One of the cases the author describes is when a lot of people decided to be very optimistic at the same time and things didn't end well.
Not questioning your frustration though, but it's good to have extremely cautious people around.
This. Society, wrongly in my opinion, seems to have converged on a 'do no harm' philosophy. Essentially we seek to minimize the risk of type 1 errors (approving a technology that causes harm) but in the process increase the number of type 2 errors (delaying or rejecting a technology that would have been safe). I think the reason is because those type 1 errors are visible and can be seen, so cause backlash, whereas the type 2 errors cannot be seen so easily. I do think we'd be better off balancing these two risks rather than just minimizing type 1s (of course this is hard as many type 2 risks are of the unknown unknown variety).
It seems to me that your reasoning comes from not caring about an increased turnover of potential effects on human lives (one such effect being death).
It's quite Mendelian an attitude, and frankly I fear people who share your view - advancement at the cost of "breaking a few eggs" - a line not unlike what you'd hear from an evil villain in a movie, or, sadly, 2017 hackernews.
It's not self concern, it's called ethics. It's not about "being linked to the action," it's simply that one situation involves someone going out of their way to imperil people, and that is simply unethical behavior. It's not about me, it's about anyone who does that. It's an ethical principle, not a personal liability. To put it another way, accidents are morally superior to murder.
Whereas the other situation, where one has cancer - and you're sadly incapable of helping them - involves no moral compromise. Nobody gave the person cancer. Your intention is to help the person, but you can't. It's tragic, but comparing the situations, only one of them has foul play.
If you think that's self concern, then you're basically resorting to "all good behavior is ultimately selfish anyway" which, like moral relativism, is a slippery slope which removes any kind of standard.
So far in this thread we've seen:
- The end justifies the means
- People are going to die anyway, may as well cause a shift (read: increase) in that number
> involves someone going out of their way to imperil people
No it doesn't, where did you get that from?
> accidents are morally superior to murder
Predictable accidents vs murder? What's the difference but personal liability?
> Nobody gave the person cancer
You say it's not about "being linked to the action", yet you think this relevant. Who cares if no-one gave them cancer, does that make it morally superior to die of natural cancer?
> Your intention is to help the person, but you can't
Only because of the way you've framed it. What if you could have helped, if another course of action had been taken?
> then you're basically resorting to "all good behavior is ultimately selfish anyway"
I disagree. You'll have to argue why this is the case.
> involves someone going out of their way to imperil people
No it doesn't, where did you get that from?
____________
That's the entire premise of the discussion! What do you mean where did I get that from?
>Predictable accidents vs murder? What's the difference but personal liability?
One is tragic, the other unethical and tragic. This point is exactly where you wind up with "all good behavior is ultimately selfish anyway" - particularly when you write: "Isn't that just self concern? You can't be linked personally as the cause of negative consequence, so you want to remain divorced from it." Here, you represent doing the right thing as avoiding being linked personally, as if it's just a self-interested maneuver. You're literally, plain as day saying that ethics is merely self interest, hence "all good behavior is ultimately selfish anyway."
The example you wrote above about a girl being run over and nobody helping is completely a false analogy (and a car analogy, no less :)). The assumption from the start, which I made very clear, is that you can't help without hurting others. Saving the girl who was run over does not imperil anyone and is completely moot!
The points in your response have been tactical rather than open (e.g. forcing me to go back and re-state the basic premise, which you suddenly no longer acknowledge, forcing me to go back and prove that you actually made the claim about ethics being self-interested, even though you did so plainly) and I don't appreciate that. These tactics further my conviction that your position is unjustified and ultimately dangerous.
> Except lots of eggs get broken today as a matter of course. If a technology saves lives, losing them to get it quicker might be a matter of balance.
The premise is that losing lives can be offset against saving them. "going out of their way to imperil people" suggests needless harm.
> You're literally, plain as day saying that ethics is merely self interest
No I don't, I give a specific example where I think this is the case; I don't apply this to all situations, or ethics in general.
I also don't "represent doing the right thing as avoiding being linked personally" - I do represent this as the only difference between two examples of doing good, (and hence don't think they are any different), but I don't question the value of doing good, but the relative values of one over another; I see value in saving lives, I don't see value in saving fewer lives because there might be personal liability.
> Saving the girl who was run over does not imperil anyone
I find it to be the slipperiest of all slopes. It's an insidious threat to humanity. It reminds me of moral relativism, but instead of cultural factors being used to justify absolutely anything, here it's some bizarre yet convenient calculation of worth, which by total coincidence (sarcasm), allows all the would-be human experimenters to feel good about themselves.
We haven't learned a thing from the scientific community's hype cycle. Scientists and technologists need to humble ourselves - the polar opposite of presuming to operate on humanity based on some assessment of human worth.
We don't even know what a good diet looks like, never mind grandiose agendas based on assumptions of human worth!
(I'd also like to correct what I wrote earlier - I meant Mengelian, not Mendelian.)
We'll have to answer the questions of "How much is a life worth monetarily", "How much is this specific life worth in relation to others (or monetarily)", "Does this specific situation justify the loss of lives; if so, how many lives and, specifically, whose?", and "Who gets the power to answer these questions?"
We will have to act on the answers to those questions.
I believe those questions open an answer space that would give power to greedy, selfish, callous or indifferent people and ideologies to kill or let others die.
I believe if we accept a utilitarianism as justification for death, we'll be inviting proponents of human actuarial science, realpolitik, genocide, sterilization and murder of certain demographics 'for the betterment of society' to kill for profit, power or any of the other reasons people have killed others in past.
I don't believe that people determine value well, so I don't buy into the utilitarian argument for much. I also haven't seen good evidence that we can accurately quantify the outcomes of situations used in the utilitarian calculus before they've happened. While I reject utilitarianism as a whole, I find applying the utilitarian principle to human lives to be particularly detached.
> If you don't believe life can be quantified, why don't we spend unlimited dollars saving it?
I realize the questions I listed above have been asked and answered, directly and indirectly, at different times and by different people. Those questions and answers have justified a disregard for life and atrocities against humanity.
I don't believe we should base policy on the answers to those questions.
Roads claim lives, buildings claim lives, products claim lives. We balance costs in all those areas too.
I believe failing to answer those questions openly is a greater harm; You have no sense of improving quality of life, or increase human worth if you pretend it's universally priceless. If you refuse to set the price, others will.
> we'll be inviting proponents of human actuarial science, realpolitik, genocide, sterilization
why do you believe this? this are extreme examples. Conflating fascism with utilitarianism seems borderline slanderous.
> don't believe that people determine value well
They aren't going to improve if there are barred from doing so. What fills the gap if analytical methods don't? Even science started from fanciful roots, could we have had modern medicine without medieval doctors cutting up cadavers?
> While I reject utilitarianism as a whole
do you mean 'don't'?
> justified a disregard for life and atrocities against humanity
You seem to have a sabotaged "brave new world" definition of utilitarianism.
> why do you believe this? this are extreme examples. Conflating fascism with utilitarianism seems borderline slanderous.
Those people already try to make a utilitarian argument for their actions and beliefs.
I didn't make up utilitarian genocide[0] to slander utilitarians.
> Two given examples of this form are the genocide of indigenous peoples in Brazil and the genocide of indigenous peoples in Paraguay.[4]
> This form of genocide was highly prominent during the European colonial expansions into the Americas, Oceania, and Africa. The colonial expansion into the Americas was markedly different in its approaches to the accumulation of wealth. The French colonization of the Americas through exploitation and the fur trade had a minor impact on the indigenous peoples. The Spanish colonization of the Americas however was devastating to the indigenous population, as was the British colonization of the Americas.[5] Dadrian has also given as further examples of utilitarian genocide the murders of Moors and Jews during the Spanish Inquisition and the killing of Cherokee Indians during the colonial expansion of the United States.[6]
From STERILIZED in the Name of Public Health [1]:
> If utilitarian pursuit of the common good required mandatory vaccination to inoculate against communicable diseases, it also necessitated “immunizing” the hereditarily defective in order to prevent the spread of bad genes. Once seen as integral to health prophylaxis and as a cost-saving recourse, sterilization programs intensified at a clipped pace across the country in the 1930s.15 By 1932, twenty-seven states had laws on the books and procedures nationwide reached over 3900.16 Not only did operations increase markedly during this decade, but some states, such as Georgia and South Carolina, passed legislation for the first time.17
> In California, at least into the 1950s, compulsory sterilization was consistently described as a public health strategy that could breed out undesirable defects from the populace and fortify the state as a whole. Convinced of its efficacy, sterilization proponents pushed for implementation of the law beyond the walls of state institutions. For example, in his Los Angeles Times Sunday magazine column “Social Eugenics” (which ran from 1936 to 1941), Fred Hogue claimed that “in this country we have wiped out the mosquito carriers of yellow fever and are in a fair way to extinguish the malaria carriers: but the human breeders of the hereditary physical and mental unfit are only in exceptional cases placed under restraint.”18 To rectify this situation, Hogue recommended broader intervention and argued that eugenic practices, above all sterilization, were essential to “the protection of the public health” and “the health security of the citizens of every State.”19 second edition of their popular textbook Applied Eugenics, Popenoe and colleague Roswell H. Johnson underscored that “if persons whose offspring will be dysgenic are so lacking in intelligence, in foresight, or in self-control that they do not control themselves, the state must control them. Sterilization is the answer.”
This is why I mention actuaries, realpolitik, genocide and sterilization: people in those camps often make utilitarian arguments for their causes.
> You seem to have a sabotaged "brave new world" definition of utilitarianism.
I've heard too many utilitarian arguments for sterilizing or eliminating the poor or other minorities pitched as efforts to maximize happiness.
Sometimes, they're pitched as efforts to maximize the happiness of those being killed or sterilized.
Often, it's "children brought up in poverty suffer, poor children are statistically likely to remain poor into adulthood and their kids will likely be poor; therefore, less children in poverty is better. Here is some napkin math to justify the sterilization of the poor" or "X minority group is a net-detriment to society, therefore less of X minority group would increase society's happiness and prosperity. Here's some napkin math on expulsion and death camps".
The internet is filled with people who make such arguments.
> do you mean 'don't'?
Utilitarianism is useful in some of the ways the market is useful. The market is great at evaluating goods, assets and labor, but it has a terrible track record when it comes to life and human rights.
> They aren't going to improve if there are barred from doing so. What fills the gap if analytical methods don't? Even science started from fanciful roots, could we have had modern medicine without medieval doctors cutting up cadavers?
The philosophies of medicine and healthcare address medical ethics while trying to answer the questions I listed in my OP with regards to medicine.
> No, but when asked about utilitarianism, you go to utilitarian genocide.
I was talking about utilitarianism with regards to quantifying human lives, not strictly utilitarianism in general.
I believe cases from the past where utilitarianism, with regards to quantifying human lives, was used to justify objectively bad things like genocide or sterilization deserve to be considered.
Again, these are my beliefs, I'm not out to slander others.
> And is the time-cube guy considered representative of science?
Given that people with power and authority have used the utilitarian principle to justify bad things, along with the prevalence of such arguments today, makes this an unfair comparison.
Timecube guy has no authority or serious adherents to his ideas.
> I believe cases from the past where utilitarianism, with regards to quantifying human lives, was used to justify objectively bad things like genocide or sterilization deserve to be considered.
There's bad utilitarianism, much like there's bad science; bad science isn't science at all.
That people claimed their intentions were pure doesn't mean we have to accept that they were, and that we might be skeptical towards claims of utilitarianism as justifications for inhumanities doesn't suggest much about utilitarianism itself to me.
> I'm not out to slander others.
What I mean by this, is if you characterise utilitarianism as a justification for the above, people who support utilitarianism appear to be supporters of a terrible ideology, whereas those are cases are all deeply flawed, and not a fair representative of utilitarianism.
> people with power and authority have used the utilitarian principle to justify bad things
The comment was a response to "people on the internet" supporting terrible utilitarianism. they also have little authority.
Terrible things where done by doctors in the past, this says little about modern medical practice.
Systems and incentives within them help explain this. Society at large doesn't incentivize the type 1 errors you described. Ben Horowitz created a 2x2 to describe this.[1][2]
I tend to agree with you, but you should think about what that harm might be. Many genetic defects that result in miscarriage also result in severe problems for the child if they're carried to term. If we're going to fix genetic issues to allow a healthy birth, it's only fair (to the baby and parents) that we also provide that baby with a healthy life.
The thing is we never know whether approving X is a type 1 error or not approving it is a type 2 error. So right now we cautiously make, say, 99.9% of type 2 errors and only 0.1% of type 1 errors. We need to relax the requirements a bit, so that the split is, say, 95%/5%, where the 5% type 1 errors are minor (as in, won't cause a huge catastrophe).
It comes down to reality vs idealism for me. The reality is that the market has perverse incentives to use these technologies in ways that benefit a few people, but can harm many more along the way.
Automation is a prime example. In an ideal world, automation would benefit everyone: perhaps eliminating scarcity and freeing up time for all humans to pursue other goals than working to survive. In reality, automation serves the needs of the few and will put millions in a situation where survival might not be possible at all.
"I'm often frustrated by the extremely cautious attitude some people have toward technology with the potential to have negative consequences, when it seems much more likely to create positive ones."
This trait, like all other traits, has been selected for over eons.
It appears to be very well preserved.
We might step back and consider the good reasons that this trait would be so well preserved - regardless of how uninspiring or inconvenient it appears to be.
It's good to allow people to take risks that only affect them, but even then we have to make sure they are properly informed of the risk. When so much of tech is spread on hype, people aren't aware how much of risk they're signing on for.
The example that frustrates me most is self driving cars. Tesla's marketing department got ahead of the rest of the org, and you end up with people doing this https://www.youtube.com/watch?v=qnZHRupjl5E. Or you see excited people comparing number of crashes per self-driving mile to number of crashes per human driven mile and thinking that's an apples-to-apples number to compare, and thinking that therefore their risk has decreased.
Don't get me wrong, I'm all for self driving cars and personal autonomy, but something has to balance all the PR and marketing bullshit around every product. How many misleading half truths and flat out lies have you seen in in advertisements lately? Or in the news? Science journalism isn't exactly renowned for accuracy.
Why? Shouldn't the bar be "better" instead of "perfect"? If we can drop car deaths by a large percentage by allowing self driving cars that kill a lot, but not as much as humans, shouldn't we?
Funny you should say that. I thought self-driving cars were a bad example for GP to use because they have almost no down side, especially in the long run. Humans suck at driving, and at dealing with high speed phenomena in general. Computers are great at that. They'll do even better when they don't have any idiot humans in their way on the road. Even if they were more dangerous than human-driven cars, they wouldn't pose anything like the existential threat that nukes, bio-engineering, AI, and nano-tech do.
> they wouldn't pose anything like the existential threat
They will further the decline of the city. If you are hoping to see America densify and embrace vibrant walkable communities, there will no longer be that visceral pressure of commute-hate to encourage it.
Additionally, there will be 3+ million Americans that will be out of a job because of it. You witnessed just how backwards a huge abandoned workforce took us on November 8th, don't underestimate the impact another wave of displacement could have.
Do you seriously think that compares to nuclear holocaust? Ecological collapse caused by reckless tinkering? Hard AI takeoff? Any sort of out-of-control replicating technology? We're worried about the literal extinction of humanity. Urban decay and unemployment don't even register on that scale.
Desperate humans are a great fuel for any type of change. Those desperate people will not care about the world burning around them, or prevention of it burning.
Walmart, insurance, and shipping companies are one of the top 5 or all in the top 5 as employer's for a frightening number of States. All of these are ripe for AI disruption.
It happens all the time. Remember when trans fats were good? Remember when we were 100% sure dietary cholesterol was bad?
There tends to be a ~20 year lag for the discovery such catastrophic failures. By the time we catch on, the people who built their careers on them have already enjoyed decades under the assumption of success and/or are long gone.
It seems the mad scientist trope is based on a very real morbid scientific temptation that many people seem to have, that handwaves away ethics with specious arguments, and fails to realize that permanent, unrecoverable mistakes that can cost people their lives are infinitely worse than abstaining from things you have doubts about.
I don't find extreme over-caution to be too far-fetched at all. I often am frustrated that people ignore possible new technology and its possible positive consequences (i.e. in simple cases where there really isn't any negative in sight) but on the whole people are generally correct to be fearful of potentially negative unknowns.
Often this leads to extremely important safety measures. For example, you mention "self-driving cars and the possibility of an accident a human would have prevented" - but that is a very poor metric, as driving a vehicle is one of the most dangerous things you can possibly do with your time, and one of the most likely activities anyone does day to day to kill or injure yourself, another driver or passengers, a cyclist or pedestrian. It seems the end-game to this extreme over-caution will be a redesign of cities that, in fact, puts a final end to this: the bar is set for self-driving cars much higher than "necessary" (to reap benefits), but the end-game will be good.
For other examples of extreme over-caution: how many nuclear reactors have melted down in the United States? How many times does a Mars Rover just explode or stop receiving or answering communications, since a simple bash script was good enough? (And in that case, no lives are even at stake.) How many times do pacemakers have a blue screen of death? How often does a drug get taken off of the market since it turns out that if you take it, suddenly you'll keel over and die, oops.
I am very often frustrated by people ignoring benefits, but I simply don't object to the level of caution people place on these things.
I mean we are talking about engineering super-humans. Humans are the only biological organism that is currently not being genetically engineered for better characteristics. Do you want people to just splice in DNA willy-nilly, seeing if we can get smarter or faster or stronger babies, assuming they survive? Where does this end?
I am excited by the potential of technologies but by no means are people's skepticism less than justified! We have a fragile equilibrium with the environment and the technological environment we have created and caution, where there are very obvious negative potentialities, is not at all misplaced in my opinion!
This stuff is genuinely scary. One bad click and you engineer a psychopathic superhuman Hitler who does not share our morals but is able to manipulate, as a sociopath, all of our systems. This isn't "hypothetical" -- examples throughout history are real.
Genetic engineering of humans is very scary, in my opinion. The reason I do not have quite the same view of GMO is that a plant might very well be nothing more than the sum of its characteristics: its rate of growth in certain climates, its nutrient needs, how much protein, carbohydrates, fats it provides.
But a human is not just a set of characteristics, for the benefit of the rest of humans. Optimizing for these is very scary. Perhaps more than people are raising the alarm about.
You know how large supermarket fruit tastes bland and awful? (But it ships really well and is huge and looks great!).
Do you really want our humans to look like that in a few years? (with respect to some wrong metrics that were chosen to be optimized for.) These things are alarming. Really, they are. Really, really.
I mean we are talking about engineering super-humans. Humans are the only biological organism that is currently not being genetically engineered for better characteristics. Do you want people to just splice in DNA willy-nilly, seeing if we can get smarter or faster or stronger babies, assuming they survive? Where does this end?
That isn't what was done in this case. This is cellular surgery, not genetic engineering.
Thanks. (Though to be clear, I was directly responding to the parent point about over-caution. :) Not making a statement about this particular news item.)
> One bad click and you engineer a psychopathic superhuman Hitler who does not share our morals but is able to manipulate, as a sociopath, all of our systems
For example, self-driving cars and the possibility of an accident a human would have prevented.
I think sometimes we need to be a little bit better at accepting risk when it comes with positive things.