Well, quality varies, but I suspect like most things most of it is mediocre. That's true of nearly all genre fiction. Pick a random mystery or sci-fi novel off the rack at a bookstore and chances are it's not very good -- although a serious reader of the genre may get things out of it that you as a casual reader don't.
One of them is a rarest of birds: a successful self-publisher who supports herself by her writing. Another is an equally rare bird: a prolific, traditionally published mid-list author with over a dozen books currently in print. And no matter how dumb you think romance stories are, these are very, very smart women who understand what they're doing extremely well -- enough to have better success than literally 99.9% of people who try their hands at it. It's not something anyone could do; yes, it's a lot of the same thing over and over, but you've got to do that and yet at the same time stand out. That ain't easy.
Anyhow I was discussing an urban fantasy manuscript my traditional author friend had earlier asked me to comment on. She'd pitched it to a paranormal romance imprint, and was bemoaning the rewrites they wanted. Now I was aware that each imprint has very specific guidelines as to what they're looking for. Buying a romance is a bit like buying jarred spaghetti sauce -- the manufacturers maximizes their sales by customizing each sauce offering to the preferences of a clusters of customers. So if you like your sauce chunky and spicy but with no vegetables, there's a jar of exactly that sitting on the shelf. That explains the vast space dedicated to spaghetti sauce in the supermarkets. Romance imprints have similar target clusters. A given imprint will be very specific about the level of sexual experience and or voracity of the heroine; exactly how roguish the love interest should be, how exotic the setting should be (some imprints don't want exotic, others do), etc.
What surprised me was how specific their editorial requirements were for sex. The manuscript had to be between 80 and 85 thousand words long (no surprises there), have at least one sex scene in the first six thousand words and thereafter another explicit sex scene every ten to fifteen thousand words. Such guidelines guarantee customers of the imprint a consistent experience; if you like one novel from the imprint it's pretty much guaranteed you'll like any of them, because for practical purposes they're almost the same damned book. Artistically it's stultifying, but I can easily imagine a formula generating a rough draft for the editors for that very reason. The editor could then send the manuscript to a human writer to tweak the dialog, setting, sensory descriptions etc.
Well, once they figure out how to generate simple, formulaic stories of one kind, they can potentially generate endless stories, not necessarily simplistic or of that particular type.
I can think of many applications for this beside putting the more artistically marginal authors out of business. You could generate the basic story outlines for a season full of American TV shows, which are for the most part filler anyway given the vast number of episodes per season. That'd free human writers up from the basic narrative spadework to add artistic touches to the script.
Or you could have random, open-ended storylines to videogames, perhaps customized to user preferences.
If you want a convincing AI, it's got to be able generate stories about itself, because that's what humans do: we take the pile of random experiences we have and transform them into a story which we call our "self".
Maybe more appropriate than you think. I've done some thinking about how fascists present themselves in stories and art, and romance plays a bit role in that. Not the bodice-ripper kind, obviously, but still a kind of fantasy of power, self-admiration and personal significance through association.
You didn't read the summary. The person suing wasn't a user; he was hit by a user. In fact his car was rear-ended, so he bears no responsibility for the accident.
Now speaking as someone who has actually designed products to people, I advocate a different strategy than blaming people who get hurt by products: common sense. If someone uses your balloon as a condom and it doesn't work, that's not your fault. But if you design a balloon that's supposed to rise into the air and burst into flame, if your users light someone else's house on fire, that's your fault because anyone with the brains of a sea cucumber would anticipate that outcome.
Well, that's true. But some of the rich people I've known are among the best people I know. That's because they have the luxury of being the kind of person they'd like to be, rather than focusing on putting bread on the table.
The bottom line is that they aren't better or worse as a group than middle class people or poor people, but their material wealth amplifies any extreme tendencies they may have, for good or ill.
In any event, the experiment wouldn't necessarily disprove free will. People have already done experiments that show that the conscious, rational part of the brain becomes engaged after it initiates an action.
Taken altogether these experiments show (a) people rationalize their actions after the fact as the result of conscious decision-making. (b) remembered sequences of events are unreliable and thus subjective attributions of actions to conscious decision-making are unreliable.
The most that these demonstrate is that some of the time we're mistaken when we think our actions are the result of conscious deliberation, which should come as a surprise to no observant person. It doesn't show that we never act as a result of conscious deliberation, which would be very hard to demonstrate in my opinion. Nor is it entirely clear to me that actions taken without conscious deliberation are necessarily not the result of "free will"; they only happen outside the bounds of one particular model for how people act (e.g. that we make decision exclusively through processes that are both logical and conscious).
Actually you have the perspective exactly backward. Yes it's true that on geological time scales the kind of climate change we're talking about is minor. However none of us live on geological time scales; and as you point out human civilization hasn't even existed on geological time scales.
It's the increase in suffering and impoverishment of people already born that's the problem, not whatever is inhabiting the Earth millions of years from now.
My wife hates house spiders; it's the one animal she kills without any shadow of compunction. I once pointed out to her that since animals are heterotrophs, those house spiders had to be living on something else. This she vehemently denied. "They eat each other," she insisted.
A slow demise for traditional TV won't precipitate any sudden crisis. As new network contracted content dries up, more studios will just pitch their stuff direct to streaming services. There will be a period of coexistence between broadcast TV, cable TV, and streaming, the way cable and broadcast coexisted. Then the weakest competitor dies off, and the cable networks probably become streaming services.
I can't imagine anyone watching synchronous TV in twenty years.
I agree we should focus more on economic class, but black people do still face a variety of discrimination. Insofar as this holds them back economically then class oriented programs will benefit them disproportionately, but this doesn't mean a program which targets blacks is necessarily bad.
When minorities begin to raise their status, they set their sights on the kinds of high status, high paying jobs you see in the media; doctors, lawyers, and so on. This means they remain under-represented longer in professions like engineering that aren't very exciting to the kind of people who are TV showrunners or movie producers. That's not to say there haven't been outstanding minority engineers all along, but as a white anglo you're much more likely to have an engineer in your family who is a role model, maybe helped you with a school or scouting project.
So trying to bring more minorities into professions like engineering (or, I suppose, accounting) where they're under-represented has value in building a world where everyone is more likely to find some kind of employment that really suits him. Sure, if you had to choose only one factor to address, you'd choose socioeconomic class. But you don't have to choose just one.
Except we're not talking about Baghdad getting uniformly 2C warmer, every day of the year. We're talking about 60 additional hot days in the summer with the rest of the year being roughly similar to today. Those sixty days would, almost own their own, raise the year-round average by more than 4C (not 2C). For that to happen the temperature increase on these days be closer to +20C, rather than +2C.
A lot of denialist reckoning does this kind of simplistic reckoning -- e.g., assuming 2C global warming means exactly 2C warmer, uniformly distributed in space and season across the globe. That of course would only amount to a trivial change. But what you're actually going to get is vast increases in extreme weather (hot AND cold) which averaged out across the globe.
Think of it this way: imagine we hold the global increase in temperature to 2C. The amount of additional kinetic energy per unit volume represented by that additional 2C, integrated over the immense volume of the atmosphere, works out to be a staggering amount of total energy, which will change the patterns of weather. Since that immense fluid is rotating, the additional energy cannot mix and diffuse out uniformly and neatly; instead it will drive massive eddies of hot and cold migrating out of their old geographic limits. You'll get all kinds of extreme weather both hot AND cold, it just all averages out to +2C temperature-wise.
To put that perspective, current estimates are that we're capturing an additional 8 x 10^21 joules of extra solar energy per year, every single year.
Americans born after 1970 or so have little memory of what smog was like before the introduction of emissions controls dramatically improved air quality. I remember as a kid coming back from vacation and getting the first glimpse of the city a giant brown smudge on the horizon. When you got back you could actually feel how much harder it was to breathe; it was like a mild asthma attack.
If you have no memory of pre-emissions control smog, try this image search. Yes, it really looked like that. Not every day mind you, but regularly in the summers.
Humans have colonized every habitat on the planet that supports any kind of population of animals, and to a large degree we did this before we had science and engineering. If anything survives, it'll be us and the cockroaches. Therefore I set the chance that humans as an entire species will disappear at zero.
Having known and met many prominent environmentalist thinkers, by in large the threat of the extinction of the entire human species doesn't occupy a lot of their thought. To the degree that they're concerned with human extinction, it's of localized populations.
These are the kinds of things they worry about:
(a) Loss of things that can only be replaced on an evolutionary timescale, like the passenger pigeon. (b) Degradation of productive biological systems, like the Aral Sea. (c) Exposure of vulnerable human populations to environmental costs that have been externalized by business, e.g. Bangladeshis to sea level rise or Appalachians to coal tailing dam failures. (d) Political destabilization due to the creation of environmental refugees (e.g., Syria/ISIS).
As one well-known environmentalist put it to me, "We need to stop living off our ecological capital and start living within our ecological means." In other words rather than liquidating the last two cod in the sea, you need to live in equilibrium with the ocean's ability to generate cod, and manage the resource so that you don't damage that (e.g. limiting the damage to the base of the food chain caused by dragging nets across the sea floor).
If cod disappears, if topsoil disappears, if we develop pesticide-resistant crop pests, if breadbasket regions fail to produce, we won't see the extinction of the entire human race. We'll see localized extinctions. Fishing towns disappear; farm communities age as youth abandon them and move to cities looking for work. In all of these scenarios, some people end up doing well. But even in a global disaster that manages to disrupt our entire civilization (unlikely, but much more likely than total human extinction), there will be places where a few of us will be able to prosper.
Which is why I always say never to assume that people use a system the way you would. But I do think the proportion of people who do this is a lot smaller than it used to be.
If the UI doesn't suit the task, it's a bad UI. There'd be nothing unusual in that, that's for sure. I'm just saying that you can't criticize a UI for not being like a paper artifact; you can only criticize them for not supporting the task adequately.
Of course understanding how people use paper artifacts is an important first step in automating something that heretofore has only been done manually, something I've done frequently in my career but I suspect is becoming much less common these days.
Paper maps are highly versatile, general purpose tools. You can do all kinds of things with them, and generally speaking the more data they cram in (in a clever way of course) the better. They're like a swiss army knife; you want them to serve in any possible occasion.
Digital map displays are embedded in a user interface; they're a backdrop that provides the user with useful contextual information as he attempts to perform some specific task. The better you understand how the user performs that task, the more you can pare down irrelevant context that might detract from that task. Now there have been many times I wished the Google Maps UI was a little more versatile, but in general it does really well at the kinds of search and navigation tasks people mainly use it for, e.g. finding all the doughnut stores in Quincy, MA.
So basically you can't automatically apply criteria you'd use in a paper map to a digital display, although it's certainly helpful to under stand those criteria.
Well, it's not the method, it's action that I'm ambivalent about. What limits are there on the president deciding anyone, anywhere in the world has to die?
And while I applaud the degradation of Al Qaeda's and ISIS's operational capabilities, I also wonder whether we're setting ourselves up to win the battle and lose the war against terror. The very precision of our weapons creates the impression that when innocents are killed by them it was deliberate.
You don't get easy choices in situations like this.
Well, the cure-all for gun violence seems to be (in the large part) to ignore it. Your chance of suffering from it has gone down across the developed world regardless of gun ownership rates, which have gone up in the US and down in other places.
If you believe in evidence based approaches then there doesn't seem to be any grand, overarching scheme needed to achieve overall reduction in gun violence. Specific measures for specific situations (e.g. school shootings) are of course a different measure; what's needed in aggregate isn't necessarily the same as what's needed in some specific situation. And Grass Range Montana doesn't necessarily need the same measures as the Bronx.
Take Esperanto. Clearly this falls under the "useful system" rubric; it was intended by its creator to be a useful system to exchange ideas.
Now consider a hypothetical. Suppose Paramount produced a movie about an eccentric scholar who invents a universal language. They go so far as to hire a linguist to construct the rudiments of such a language. In that case the language is constructed to give the most convincing artistic impression of a constructed language possible. It's not designed to be useful, it just is, like fantasy weapons and armor which are probably not very practical but in some cases could hurt someone if you finished them properly (e.g. putting an edge on Sting from LotR). Is the movie Esperanto essentially different from a realistically designed fantasy sword?
Slashdot Mirror
Well, quality varies, but I suspect like most things most of it is mediocre. That's true of nearly all genre fiction. Pick a random mystery or sci-fi novel off the rack at a bookstore and chances are it's not very good -- although a serious reader of the genre may get things out of it that you as a casual reader don't.
That literally was how my comment started out. "Duh!"
One of them is a rarest of birds: a successful self-publisher who supports herself by her writing. Another is an equally rare bird: a prolific, traditionally published mid-list author with over a dozen books currently in print. And no matter how dumb you think romance stories are, these are very, very smart women who understand what they're doing extremely well -- enough to have better success than literally 99.9% of people who try their hands at it. It's not something anyone could do; yes, it's a lot of the same thing over and over, but you've got to do that and yet at the same time stand out. That ain't easy.
Anyhow I was discussing an urban fantasy manuscript my traditional author friend had earlier asked me to comment on. She'd pitched it to a paranormal romance imprint, and was bemoaning the rewrites they wanted. Now I was aware that each imprint has very specific guidelines as to what they're looking for. Buying a romance is a bit like buying jarred spaghetti sauce -- the manufacturers maximizes their sales by customizing each sauce offering to the preferences of a clusters of customers. So if you like your sauce chunky and spicy but with no vegetables, there's a jar of exactly that sitting on the shelf. That explains the vast space dedicated to spaghetti sauce in the supermarkets. Romance imprints have similar target clusters. A given imprint will be very specific about the level of sexual experience and or voracity of the heroine; exactly how roguish the love interest should be, how exotic the setting should be (some imprints don't want exotic, others do), etc.
What surprised me was how specific their editorial requirements were for sex. The manuscript had to be between 80 and 85 thousand words long (no surprises there), have at least one sex scene in the first six thousand words and thereafter another explicit sex scene every ten to fifteen thousand words. Such guidelines guarantee customers of the imprint a consistent experience; if you like one novel from the imprint it's pretty much guaranteed you'll like any of them, because for practical purposes they're almost the same damned book. Artistically it's stultifying, but I can easily imagine a formula generating a rough draft for the editors for that very reason. The editor could then send the manuscript to a human writer to tweak the dialog, setting, sensory descriptions etc.
Well, once they figure out how to generate simple, formulaic stories of one kind, they can potentially generate endless stories, not necessarily simplistic or of that particular type.
I can think of many applications for this beside putting the more artistically marginal authors out of business. You could generate the basic story outlines for a season full of American TV shows, which are for the most part filler anyway given the vast number of episodes per season. That'd free human writers up from the basic narrative spadework to add artistic touches to the script.
Or you could have random, open-ended storylines to videogames, perhaps customized to user preferences.
If you want a convincing AI, it's got to be able generate stories about itself, because that's what humans do: we take the pile of random experiences we have and transform them into a story which we call our "self".
Maybe more appropriate than you think. I've done some thinking about how fascists present themselves in stories and art, and romance plays a bit role in that. Not the bodice-ripper kind, obviously, but still a kind of fantasy of power, self-admiration and personal significance through association.
To slashdot's new masters: your readers aren't so ignorant that we think that Solar Impulse 2 means we'll be seeing solar powered 747s. Sheesh.
You didn't read the summary. The person suing wasn't a user; he was hit by a user. In fact his car was rear-ended, so he bears no responsibility for the accident.
Now speaking as someone who has actually designed products to people, I advocate a different strategy than blaming people who get hurt by products: common sense. If someone uses your balloon as a condom and it doesn't work, that's not your fault. But if you design a balloon that's supposed to rise into the air and burst into flame, if your users light someone else's house on fire, that's your fault because anyone with the brains of a sea cucumber would anticipate that outcome.
Well, that's true. But some of the rich people I've known are among the best people I know. That's because they have the luxury of being the kind of person they'd like to be, rather than focusing on putting bread on the table.
The bottom line is that they aren't better or worse as a group than middle class people or poor people, but their material wealth amplifies any extreme tendencies they may have, for good or ill.
In any event, the experiment wouldn't necessarily disprove free will. People have already done experiments that show that the conscious, rational part of the brain becomes engaged after it initiates an action.
Taken altogether these experiments show
(a) people rationalize their actions after the fact as the result of conscious decision-making.
(b) remembered sequences of events are unreliable and thus subjective attributions of actions to conscious decision-making are unreliable.
The most that these demonstrate is that some of the time we're mistaken when we think our actions are the result of conscious deliberation, which should come as a surprise to no observant person. It doesn't show that we never act as a result of conscious deliberation, which would be very hard to demonstrate in my opinion. Nor is it entirely clear to me that actions taken without conscious deliberation are necessarily not the result of "free will"; they only happen outside the bounds of one particular model for how people act (e.g. that we make decision exclusively through processes that are both logical and conscious).
Actually you have the perspective exactly backward. Yes it's true that on geological time scales the kind of climate change we're talking about is minor. However none of us live on geological time scales; and as you point out human civilization hasn't even existed on geological time scales.
It's the increase in suffering and impoverishment of people already born that's the problem, not whatever is inhabiting the Earth millions of years from now.
My wife hates house spiders; it's the one animal she kills without any shadow of compunction. I once pointed out to her that since animals are heterotrophs, those house spiders had to be living on something else. This she vehemently denied. "They eat each other," she insisted.
A slow demise for traditional TV won't precipitate any sudden crisis. As new network contracted content dries up, more studios will just pitch their stuff direct to streaming services. There will be a period of coexistence between broadcast TV, cable TV, and streaming, the way cable and broadcast coexisted. Then the weakest competitor dies off, and the cable networks probably become streaming services.
I can't imagine anyone watching synchronous TV in twenty years.
Try "has" for anything but breaking news or sports.
People like you never see any other alternatives than "piss off".
I agree we should focus more on economic class, but black people do still face a variety of discrimination. Insofar as this holds them back economically then class oriented programs will benefit them disproportionately, but this doesn't mean a program which targets blacks is necessarily bad.
When minorities begin to raise their status, they set their sights on the kinds of high status, high paying jobs you see in the media; doctors, lawyers, and so on. This means they remain under-represented longer in professions like engineering that aren't very exciting to the kind of people who are TV showrunners or movie producers. That's not to say there haven't been outstanding minority engineers all along, but as a white anglo you're much more likely to have an engineer in your family who is a role model, maybe helped you with a school or scouting project.
So trying to bring more minorities into professions like engineering (or, I suppose, accounting) where they're under-represented has value in building a world where everyone is more likely to find some kind of employment that really suits him. Sure, if you had to choose only one factor to address, you'd choose socioeconomic class. But you don't have to choose just one.
Well, I've known some rich people; they aren't necessarily assholes.
The problem is that a rich asshole is still an asshole, and unlike an ordinary asshole he's got an enormous societal and political footprint.
Except we're not talking about Baghdad getting uniformly 2C warmer, every day of the year. We're talking about 60 additional hot days in the summer with the rest of the year being roughly similar to today. Those sixty days would, almost own their own, raise the year-round average by more than 4C (not 2C). For that to happen the temperature increase on these days be closer to +20C, rather than +2C.
A lot of denialist reckoning does this kind of simplistic reckoning -- e.g., assuming 2C global warming means exactly 2C warmer, uniformly distributed in space and season across the globe. That of course would only amount to a trivial change. But what you're actually going to get is vast increases in extreme weather (hot AND cold) which averaged out across the globe.
Think of it this way: imagine we hold the global increase in temperature to 2C. The amount of additional kinetic energy per unit volume represented by that additional 2C, integrated over the immense volume of the atmosphere, works out to be a staggering amount of total energy, which will change the patterns of weather. Since that immense fluid is rotating, the additional energy cannot mix and diffuse out uniformly and neatly; instead it will drive massive eddies of hot and cold migrating out of their old geographic limits. You'll get all kinds of extreme weather both hot AND cold, it just all averages out to +2C temperature-wise.
To put that perspective, current estimates are that we're capturing an additional 8 x 10^21 joules of extra solar energy per year, every single year.
Americans born after 1970 or so have little memory of what smog was like before the introduction of emissions controls dramatically improved air quality. I remember as a kid coming back from vacation and getting the first glimpse of the city a giant brown smudge on the horizon. When you got back you could actually feel how much harder it was to breathe; it was like a mild asthma attack.
If you have no memory of pre-emissions control smog, try this image search. Yes, it really looked like that. Not every day mind you, but regularly in the summers.
The only place in the US where you can still experience that kind of epic smog is in parts of LA, and even that has improved over the last several decades.
Humans have colonized every habitat on the planet that supports any kind of population of animals, and to a large degree we did this before we had science and engineering. If anything survives, it'll be us and the cockroaches. Therefore I set the chance that humans as an entire species will disappear at zero.
Having known and met many prominent environmentalist thinkers, by in large the threat of the extinction of the entire human species doesn't occupy a lot of their thought. To the degree that they're concerned with human extinction, it's of localized populations.
These are the kinds of things they worry about:
(a) Loss of things that can only be replaced on an evolutionary timescale, like the passenger pigeon.
(b) Degradation of productive biological systems, like the Aral Sea.
(c) Exposure of vulnerable human populations to environmental costs that have been externalized by business, e.g. Bangladeshis to sea level rise or Appalachians to coal tailing dam failures.
(d) Political destabilization due to the creation of environmental refugees (e.g., Syria/ISIS).
As one well-known environmentalist put it to me, "We need to stop living off our ecological capital and start living within our ecological means." In other words rather than liquidating the last two cod in the sea, you need to live in equilibrium with the ocean's ability to generate cod, and manage the resource so that you don't damage that (e.g. limiting the damage to the base of the food chain caused by dragging nets across the sea floor).
If cod disappears, if topsoil disappears, if we develop pesticide-resistant crop pests, if breadbasket regions fail to produce, we won't see the extinction of the entire human race. We'll see localized extinctions. Fishing towns disappear; farm communities age as youth abandon them and move to cities looking for work. In all of these scenarios, some people end up doing well. But even in a global disaster that manages to disrupt our entire civilization (unlikely, but much more likely than total human extinction), there will be places where a few of us will be able to prosper.
Which is why I always say never to assume that people use a system the way you would. But I do think the proportion of people who do this is a lot smaller than it used to be.
If the UI doesn't suit the task, it's a bad UI. There'd be nothing unusual in that, that's for sure. I'm just saying that you can't criticize a UI for not being like a paper artifact; you can only criticize them for not supporting the task adequately.
Of course understanding how people use paper artifacts is an important first step in automating something that heretofore has only been done manually, something I've done frequently in my career but I suspect is becoming much less common these days.
It's UI design, and that's task-oriented.
Paper maps are highly versatile, general purpose tools. You can do all kinds of things with them, and generally speaking the more data they cram in (in a clever way of course) the better. They're like a swiss army knife; you want them to serve in any possible occasion.
Digital map displays are embedded in a user interface; they're a backdrop that provides the user with useful contextual information as he attempts to perform some specific task. The better you understand how the user performs that task, the more you can pare down irrelevant context that might detract from that task. Now there have been many times I wished the Google Maps UI was a little more versatile, but in general it does really well at the kinds of search and navigation tasks people mainly use it for, e.g. finding all the doughnut stores in Quincy, MA.
So basically you can't automatically apply criteria you'd use in a paper map to a digital display, although it's certainly helpful to under stand those criteria.
Well, it's not the method, it's action that I'm ambivalent about. What limits are there on the president deciding anyone, anywhere in the world has to die?
And while I applaud the degradation of Al Qaeda's and ISIS's operational capabilities, I also wonder whether we're setting ourselves up to win the battle and lose the war against terror. The very precision of our weapons creates the impression that when innocents are killed by them it was deliberate.
You don't get easy choices in situations like this.
Well, the cure-all for gun violence seems to be (in the large part) to ignore it. Your chance of suffering from it has gone down across the developed world regardless of gun ownership rates, which have gone up in the US and down in other places.
If you believe in evidence based approaches then there doesn't seem to be any grand, overarching scheme needed to achieve overall reduction in gun violence. Specific measures for specific situations (e.g. school shootings) are of course a different measure; what's needed in aggregate isn't necessarily the same as what's needed in some specific situation. And Grass Range Montana doesn't necessarily need the same measures as the Bronx.
Take Esperanto. Clearly this falls under the "useful system" rubric; it was intended by its creator to be a useful system to exchange ideas.
Now consider a hypothetical. Suppose Paramount produced a movie about an eccentric scholar who invents a universal language. They go so far as to hire a linguist to construct the rudiments of such a language. In that case the language is constructed to give the most convincing artistic impression of a constructed language possible. It's not designed to be useful, it just is, like fantasy weapons and armor which are probably not very practical but in some cases could hurt someone if you finished them properly (e.g. putting an edge on Sting from LotR). Is the movie Esperanto essentially different from a realistically designed fantasy sword?