Yes, it's a simplified toy model that may not perfectly describe the exact actions of a crow. However, within the model assumptions (which aren't unreasonable for real-world assumptions), the actions make sense. One component of the model is the time length tau that the critter "thinks ahead" to maximize entropy. Considering flying away to find food elsewhere requires taking more distant times into consideration; within a few-second "think ahead" horizon, there's nothing else for the crow to do than deal with accessible objects in its immediate environment. Another not unreasonable model assumption is that the crow's mental "world" at the time consists of only the stick, log, food, and itself; the crow isn't deeply pondering every other object it might encounter in the universe. Between the limited "think -ahead" time and the limited "objects" in the crow's mental universe, the stick-grabbing behavior does indeed follow the entropy-maximizing model.
A key component to the paper's model is that entropy maximization is not just through "local" maximization of the gradient, but total entropy maximization over an interval:
Inspired by recent developments [1–6] to naturally generalize such biases so that they uniformly maximize entropy production between the present and a future time horizon, rather than just greedily maximizing instanta- neous entropy production, we can also contemplate gener- alized entropic forces over paths through configuration space rather than just over the configuration space itself.
So, indeed, entropy maximization --- not just instantaneous entropy gradient maximization, which might "miss" solutions requiring passing through low-entropy-gradient regions to reach an even higher entropy final state --- is important. Of course, the result of maximizing entropy over a time interval is the maximization of the average gradient over the interval.
My "critter not going into a corner" example was based on the first toy model in the paper, a particle that drifts towards the center of a box when "driven" by entropy maximization. In some of the more "advanced" examples, there are more complex factors coming into play that may maximize entropy by "ending up in a corner," depending on how the problem is set up. However, if you read the paper (instead of just glancing at videos), the mathematical formalism that drives the model is all about maximizing entropy.
The point in the paper that addresses the "burn shit to be smart!" concept is that the "intelligence" is operating on a simplified, macroscopic model of the world, which doesn't pay attention to the microscopic entropy of chemical bonds (increased by setting stuff on fire). In this simplified "critter-scale" world, shorter-term entropy gain *is* the driving compulsion. The toy model "crow reaching food with a stick" example wasn't driven by the crow thinking "gee, if I don't eat now, I'll be dead next year, so I'd better do something about that." Instead, the problem was "solved" by the crow maximizing entropy a few seconds ahead --- e.g. it moves to reach the stick, because there are a lot more system states available if the stick can be manipulated instead of just lying in the same place on the ground. The "intelligent behavior" only needs to maximize entropy on the time-scale associated with completing the immediate task --- a few seconds --- rather than "long term" considerations about nutritional needs.
The idea is that the urge to resist entropy yields a competitive advantage and leads to intelligence.
Actually, the opposite: "intelligence" functions by seeking to maximize entropy. Note, however, we are talking about an approximate "macroscopic scale" entropy of large-scale objects in the system rather than the "microscopic" entropy of chemical reactions, so "intelligence" isn't about intentionally setting as many things as you can on fire ("microscopic" entropy maximization). So, the analogue statement to "all the gas molecules in a room won't pile up on one side" is "an intelligent critter won't want to get backed into a corner" --- in both cases, the "system" works to maximize "entropy," or the number of available future states (a lot less possibilities for where molecules can be if they're all confined to half the room; a lot less places where a critter can go if stuck in a corner).
The paper has nothing to do with the "design" and/or evolution of intelligence. It proposes a general mechanism by which "intelligent" brains may be able to "figure out" how to perform a wide variety of tasks (by maximizing future available states based on a simplified internal world model). Plain old evolutionary selective pressures would favor critters with brains good at carrying out this type of cognition.
For instance, on the planet Earth, man had always assumed that he was more intelligent than dolphins because he had achieved so much—the wheel, New York, wars and so on—whilst all the dolphins had ever done was muck about in the water having a good time. But conversely, the dolphins had always believed that they were far more intelligent than man—for precisely the same reasons.
Yes, what you're missing is the entire point of the paper. Here's my attempt at a quick summary: Suppose you are a hungry crow. You see a tasty morsel of food in a hollow log (that you can't directly reach), and a long stick on the ground. The paper poses an answer to the question: what general mechanism would let you "figure out" how to get the food?
Many cognitive models might approach this by assuming the crow has a big table of "knowledge" that it can logically manipulate to deduce an answer: "stick can reach food from entrance to log," "I can get stick if I go over there," "I can move stick to entrance of log," => "I can reach food." This paper, however, proposes a much more general and simple model: the crow lives by the rule "I'll do whatever will maximize the number of different world states my world can be in 5 seconds from now." By this principle, the crow can reach a lot more states if it can move the stick (instead of the fewer states where the stick just sits in the same place on the ground), so it heads over towards the stick. Now it can reach a lot more states if it pokes the food out of the hole with the stick, so it does. And now, it can eat the tasty food.
The paper shows a few different examples where the single "maximize available future states" principle allows toy models to "solve" various problems and exhibit behavior associated with "cognition." This provides a very general mechanism for cognition driving a wide variety of behaviors, that doesn't require the thinking critter to have a giant "knowledge bank" from which to calculate complicated chains of logic before acting.
You can tell this is a physicist's paper. It lacks spherical cows, but only because the toy models were set up in 2D. So, instead, we get a crow, chimpanzee, or elephant approximated by circular disks.
Those were economists, not scientists. When your job is to be an apologist for the wealthy, accuracy and statistical rigor has little to do with the standards by which your work is judged.
Exactly. "Statistics for people who don't understand math" courses are counterproductive and actually dangerous to the sciences. They churn out people who treat statistical analysis as a magical black box --- they've memorized which incantation sequences to type into some calculator or statistical analysis software package without the least understanding of what they're doing. The result is folks churning out research analysis applying all sorts of sophisticated-sounding statistical methods in inapplicable and utterly wrong ways. If calculus and linear algebra are above your comprehension level, then your understanding of statistics will do nothing but put you in the "knows just enough to be dangerous" category where you're too ignorant to tell that you're incompetent.
Anyway, a simple example of what "wasted" data would mean: suppose you are measuring some small signal with a background of similar or larger magnitude. You can measure signal+background or just background. So, you tell your lab tech to spend the next couple weeks measuring the signal and background, and at the end of the time they come back and say "I measured the signal+background for 100 hours, interspersed with 10 hours of background data." D'oh! If the signal is smaller than the background, then the statistical sensitivity of 100h Sig+Bkg and 10h Bkg is barely better than 10 hours of each; you should have measured ~55 hours of each. Your signal measurements are statistically crippled by uncertainty in the background, and you've wasted a lot of lab time for suboptimal results. You needed people with expertise on all parts of the experiment --- what the expected signal and noise components are, how to allocate time for maximally useful statistics --- to be in on the planning.
No, "double blind" doesn't mean the people performing an experiment don't know what hypothesis they are testing for or how the data is being analyzed. In fact, precisely determining and understanding the protocol for data analysis *before* doing the work is a critical component of proper "blind" experiments. "Blinding" means that the experimental protocol is designed so the researchers will not be able to tell which way particular results will skew the outcome while doing the experiment, not that they are ignorant of what they're doing.
Blinded protocol example: researcher 'A' randomly fills numbered bottles with either a solution being studied or an identical-appearing control substance, recording which bottle contains which solution. Researcher 'B' takes the bottles, not knowing which contains which contents, sprays them on a series of petri dishes, and measures how stuff grows on each one. After 'B' collects the data, 'A' reveals what was in each bottle, so analysis (according to a rigorously predetermined procedure) can indicate what effects the solution under study had compared to the control. Blinding has nothing to do with 'A' and/or 'B' being ignorant about what their hypotheses or analysis methods are.
Math may come second, but it does need to come. If Einstein had just been the guy who went around saying "dude, everything is like totally relative! Cosmic space-time bendy-warp, all-one time-cube, dude!," and expecting someone else to fill in the mathematical formalism, I doubt he'd be all that famous now. Einstein was able to write down his insights as tensor calculus equations --- that's why he's remembered as a famous scientist, not an incoherent ranting quack.
Collecting data without having a darn good grasp of how the data analysis works is a great way to waste a huge amount of time and money collecting mostly useless data. It may not be the same person doing both, but the data-collector definitely needs to be intimately "in the loop" about how their experimental work impacts uncertainties in the final analysis.
Intuition and the part of math that involves being good at grinding through lengthy, dense calculations without making sign errors don't have to be the same person. However, a strong and intuitive sense of what math is capable of (which requires advanced mathematical education) do need to go together for scientific productivity. Otherwise, it's just like the techno-incompetent manager asking engineers to implement his "brilliant" physically impossible designs.
The first scene, in the first movie, is a slo-mo shot of Jar-Jar Binks getting his head sliced off with a lightsaber.
Unfortunately, the second scene has the camera view swooping through the door marked "sekrit cloning lab" into a room filled with tens of thousands of mechanical pods. Lids on the pods slide open in unison, as the camera zooms in to the blank soulless gaze of a Jar-Jar clone. Scrolling title text rolls from the bottom of the screen, receding to a vanishing point:
Looks more like trollbait than any one political ideaology.
Have you seen the political ideologies this country generates? All too often, the reality transcends attempts at parody --- the lobbyist pundits brought in by the media as "serious" viewpoints are more sick and incoherent than the foulest internet troll.
I'm not denying there's no harm, but it's nothing near to a whole day's economic output lost. A careful analysis of actual societal costs and benefits is necessary. A few people might suffer or die from medical issues (in life-or-death emergency situations, however, an ambulance will still come, and probably faster than if the streets were clogged with snowbound vehicles); on the other hand, many lives might be saved from traffic accidents as people skidded through unusually dangerous road conditions to get to/from work. True, some people might get stuck at the airport --- assuming flights would be running anyway, instead of also all being canceled for the weather.
Is it a "slippery slope" for government to declare a special "holiday" when cost-benefit analyses indicate significant overall savings for society? If you're a total libertarian government-should-do-nothing type, then you won't be happy with this. On the other hand, this is something the "free market" will not fix on its own: employers, especially, will be too happy to offload risks onto their employees, and demand people come in to work (with the employee taking the added risk of personal injury or death for the employer's benefit) --- the same type of safety issues that unregulated markets tend to fail spectacularly at. The economic criteria and cost-benefit analyses for making such decisions should be made public and transparent (and subject to voter modification should people disagree with the government technocrats' decision), but I don't think it's out-of-place (or a slippery slope to fascism) for the government to regulate "general public welfare" safety concerns like road travel during dangerous weather conditions.
Very few companies of any sort are running at 100% capacity all the time, and limited in their output by the number of hours in the day --- basically, whenever someone hits this point, it's profitable for them (or a competitor) to expand operations, until economic output is limited by how many units (of whatever one is producing; goods or services) one can sell at positive margins instead of how much can be produced at maximum capacity in a day. At worst, you have to spend a tiny bit more on overtime pay (a small fraction of total production costs) to catch up. Basically nothing public consumer-facing will cause overall economic loss, because even if customers don't return to the same place they'll still spend their money somewhere. And most businesses selling to other businesses have fixed order sizes, that they can make up with higher production rates (or a few extra hours) on other days. And again, most money not spent in one place still gets spent in another --- net economic damage only occurs when something is actually destroyed (e.g. food that spoils during the "missing" day) rather than shuffled around to another time/place.
And I'm sure a local food service establishment could have supplied food too. Oh wait, that's what happened. Aside from the "derp, cops love donuts!" funny factor of the particular choice of vendor, what happened seems like a perfectly fine and practical use of available resources. If the cops had called in a military convoy to do their coffee runs, many of the same folks complaining about what did happen would be whining instead about the massive waste of taxpayer money to send a squadron of hummers with MREs instead of just using local food vendors.
I don't know how intrusive the emergency drills on your campus are, but the ones I've seen really don't stop much productive work. It's not "everyone cower under their desk in the fetal position and tremble in silence until the all-clear is announced," but "stay in the room doing what you were doing anyway, and hold off on any trips into the hall/outdoors that you can wait on." No one gets tackled by the campus police for needing to get to the bathroom, or puts their experiments on hold, or stops their lecture, or quits working on homework.
In retrospect, it's interesting that the bomber didn't kill more people when they actually had the chance. During their escape, they held up a convenience store and stole a car --- without shooting the robbery victims. An interesting artifact of human psychology, even at its most twisted: the terrorists willing to blow up random strangers weren't willing to look a shopkeeper or driver in the eye and shoot them; in panicked flight and personal contact with potential victims, they showed far more restraint and respect for human life than their premeditated impersonal cold-blooded murders just hours before.
So, making sure officers on an intensive stakeout can get a quick calorie boost is a bad thing (herp derp, cops and donuts!)? Would it have been better to tell the police force "everyone go grab lunch in a far-away safe location, and be back in two hours"? Or, should the police force maintain their own mobile food preparation facilities, so they're ready to bring their own snack trucks to the site of once-in-a-few-years major operations? Keeping a close-by source of quick-energy-rush stay-awake food available (by the simple and easy method of obtaining help from local businesses already expert in that field) actually seems like a sensible idea --- or do you think a cop shouldn't get hungry after 14 hours stressful work?
Slashdot Mirror
Yes, it's a simplified toy model that may not perfectly describe the exact actions of a crow. However, within the model assumptions (which aren't unreasonable for real-world assumptions), the actions make sense. One component of the model is the time length tau that the critter "thinks ahead" to maximize entropy. Considering flying away to find food elsewhere requires taking more distant times into consideration; within a few-second "think ahead" horizon, there's nothing else for the crow to do than deal with accessible objects in its immediate environment. Another not unreasonable model assumption is that the crow's mental "world" at the time consists of only the stick, log, food, and itself; the crow isn't deeply pondering every other object it might encounter in the universe. Between the limited "think -ahead" time and the limited "objects" in the crow's mental universe, the stick-grabbing behavior does indeed follow the entropy-maximizing model.
A key component to the paper's model is that entropy maximization is not just through "local" maximization of the gradient, but total entropy maximization over an interval:
Inspired by recent developments [1–6] to naturally
generalize such biases so that they uniformly maximize
entropy production between the present and a future time
horizon, rather than just greedily maximizing instanta-
neous entropy production, we can also contemplate gener-
alized entropic forces over paths through configuration
space rather than just over the configuration space itself.
So, indeed, entropy maximization --- not just instantaneous entropy gradient maximization, which might "miss" solutions requiring passing through low-entropy-gradient regions to reach an even higher entropy final state --- is important. Of course, the result of maximizing entropy over a time interval is the maximization of the average gradient over the interval.
My "critter not going into a corner" example was based on the first toy model in the paper, a particle that drifts towards the center of a box when "driven" by entropy maximization. In some of the more "advanced" examples, there are more complex factors coming into play that may maximize entropy by "ending up in a corner," depending on how the problem is set up. However, if you read the paper (instead of just glancing at videos), the mathematical formalism that drives the model is all about maximizing entropy.
The point in the paper that addresses the "burn shit to be smart!" concept is that the "intelligence" is operating on a simplified, macroscopic model of the world, which doesn't pay attention to the microscopic entropy of chemical bonds (increased by setting stuff on fire). In this simplified "critter-scale" world, shorter-term entropy gain *is* the driving compulsion. The toy model "crow reaching food with a stick" example wasn't driven by the crow thinking "gee, if I don't eat now, I'll be dead next year, so I'd better do something about that." Instead, the problem was "solved" by the crow maximizing entropy a few seconds ahead --- e.g. it moves to reach the stick, because there are a lot more system states available if the stick can be manipulated instead of just lying in the same place on the ground. The "intelligent behavior" only needs to maximize entropy on the time-scale associated with completing the immediate task --- a few seconds --- rather than "long term" considerations about nutritional needs.
The idea is that the urge to resist entropy yields a competitive advantage and leads to intelligence.
Actually, the opposite: "intelligence" functions by seeking to maximize entropy. Note, however, we are talking about an approximate "macroscopic scale" entropy of large-scale objects in the system rather than the "microscopic" entropy of chemical reactions, so "intelligence" isn't about intentionally setting as many things as you can on fire ("microscopic" entropy maximization). So, the analogue statement to "all the gas molecules in a room won't pile up on one side" is "an intelligent critter won't want to get backed into a corner" --- in both cases, the "system" works to maximize "entropy," or the number of available future states (a lot less possibilities for where molecules can be if they're all confined to half the room; a lot less places where a critter can go if stuck in a corner).
The paper has nothing to do with the "design" and/or evolution of intelligence. It proposes a general mechanism by which "intelligent" brains may be able to "figure out" how to perform a wide variety of tasks (by maximizing future available states based on a simplified internal world model). Plain old evolutionary selective pressures would favor critters with brains good at carrying out this type of cognition.
For instance, on the planet Earth, man had always assumed that he was more intelligent than dolphins because he had achieved so much—the wheel, New York, wars and so on—whilst all the dolphins had ever done was muck about in the water having a good time. But conversely, the dolphins had always believed that they were far more intelligent than man—for precisely the same reasons.
-- Douglas Adams
Yes, what you're missing is the entire point of the paper. Here's my attempt at a quick summary:
Suppose you are a hungry crow. You see a tasty morsel of food in a hollow log (that you can't directly reach), and a long stick on the ground. The paper poses an answer to the question: what general mechanism would let you "figure out" how to get the food?
Many cognitive models might approach this by assuming the crow has a big table of "knowledge" that it can logically manipulate to deduce an answer: "stick can reach food from entrance to log," "I can get stick if I go over there," "I can move stick to entrance of log," => "I can reach food." This paper, however, proposes a much more general and simple model: the crow lives by the rule "I'll do whatever will maximize the number of different world states my world can be in 5 seconds from now." By this principle, the crow can reach a lot more states if it can move the stick (instead of the fewer states where the stick just sits in the same place on the ground), so it heads over towards the stick. Now it can reach a lot more states if it pokes the food out of the hole with the stick, so it does. And now, it can eat the tasty food.
The paper shows a few different examples where the single "maximize available future states" principle allows toy models to "solve" various problems and exhibit behavior associated with "cognition." This provides a very general mechanism for cognition driving a wide variety of behaviors, that doesn't require the thinking critter to have a giant "knowledge bank" from which to calculate complicated chains of logic before acting.
You can tell this is a physicist's paper. It lacks spherical cows, but only because the toy models were set up in 2D. So, instead, we get a crow, chimpanzee, or elephant approximated by circular disks.
Disney's sequel-making experts know a lot about cloning soulless monstrosities; I'm sure they'll get the details correct.
Those were economists, not scientists. When your job is to be an apologist for the wealthy, accuracy and statistical rigor has little to do with the standards by which your work is judged.
Exactly. "Statistics for people who don't understand math" courses are counterproductive and actually dangerous to the sciences. They churn out people who treat statistical analysis as a magical black box --- they've memorized which incantation sequences to type into some calculator or statistical analysis software package without the least understanding of what they're doing. The result is folks churning out research analysis applying all sorts of sophisticated-sounding statistical methods in inapplicable and utterly wrong ways. If calculus and linear algebra are above your comprehension level, then your understanding of statistics will do nothing but put you in the "knows just enough to be dangerous" category where you're too ignorant to tell that you're incompetent.
You're a moron.
Anyway, a simple example of what "wasted" data would mean: suppose you are measuring some small signal with a background of similar or larger magnitude. You can measure signal+background or just background. So, you tell your lab tech to spend the next couple weeks measuring the signal and background, and at the end of the time they come back and say "I measured the signal+background for 100 hours, interspersed with 10 hours of background data." D'oh! If the signal is smaller than the background, then the statistical sensitivity of 100h Sig+Bkg and 10h Bkg is barely better than 10 hours of each; you should have measured ~55 hours of each. Your signal measurements are statistically crippled by uncertainty in the background, and you've wasted a lot of lab time for suboptimal results. You needed people with expertise on all parts of the experiment --- what the expected signal and noise components are, how to allocate time for maximally useful statistics --- to be in on the planning.
No, "double blind" doesn't mean the people performing an experiment don't know what hypothesis they are testing for or how the data is being analyzed. In fact, precisely determining and understanding the protocol for data analysis *before* doing the work is a critical component of proper "blind" experiments. "Blinding" means that the experimental protocol is designed so the researchers will not be able to tell which way particular results will skew the outcome while doing the experiment, not that they are ignorant of what they're doing.
Blinded protocol example: researcher 'A' randomly fills numbered bottles with either a solution being studied or an identical-appearing control substance, recording which bottle contains which solution. Researcher 'B' takes the bottles, not knowing which contains which contents, sprays them on a series of petri dishes, and measures how stuff grows on each one. After 'B' collects the data, 'A' reveals what was in each bottle, so analysis (according to a rigorously predetermined procedure) can indicate what effects the solution under study had compared to the control. Blinding has nothing to do with 'A' and/or 'B' being ignorant about what their hypotheses or analysis methods are.
Math may come second, but it does need to come. If Einstein had just been the guy who went around saying "dude, everything is like totally relative! Cosmic space-time bendy-warp, all-one time-cube, dude!," and expecting someone else to fill in the mathematical formalism, I doubt he'd be all that famous now. Einstein was able to write down his insights as tensor calculus equations --- that's why he's remembered as a famous scientist, not an incoherent ranting quack.
Collecting data without having a darn good grasp of how the data analysis works is a great way to waste a huge amount of time and money collecting mostly useless data. It may not be the same person doing both, but the data-collector definitely needs to be intimately "in the loop" about how their experimental work impacts uncertainties in the final analysis.
Intuition and the part of math that involves being good at grinding through lengthy, dense calculations without making sign errors don't have to be the same person. However, a strong and intuitive sense of what math is capable of (which requires advanced mathematical education) do need to go together for scientific productivity. Otherwise, it's just like the techno-incompetent manager asking engineers to implement his "brilliant" physically impossible designs.
The first scene, in the first movie, is a slo-mo shot of Jar-Jar Binks getting his head sliced off with a lightsaber.
Unfortunately, the second scene has the camera view swooping through the door marked "sekrit cloning lab" into a room filled with tens of thousands of mechanical pods. Lids on the pods slide open in unison, as the camera zooms in to the blank soulless gaze of a Jar-Jar clone. Scrolling title text rolls from the bottom of the screen, receding to a vanishing point:
STAR
WARS
EPISODE VII
Rise of the Jar-Jarmy
Looks more like trollbait than any one political ideaology.
Have you seen the political ideologies this country generates? All too often, the reality transcends attempts at parody --- the lobbyist pundits brought in by the media as "serious" viewpoints are more sick and incoherent than the foulest internet troll.
I'm not denying there's no harm, but it's nothing near to a whole day's economic output lost. A careful analysis of actual societal costs and benefits is necessary. A few people might suffer or die from medical issues (in life-or-death emergency situations, however, an ambulance will still come, and probably faster than if the streets were clogged with snowbound vehicles); on the other hand, many lives might be saved from traffic accidents as people skidded through unusually dangerous road conditions to get to/from work. True, some people might get stuck at the airport --- assuming flights would be running anyway, instead of also all being canceled for the weather.
Is it a "slippery slope" for government to declare a special "holiday" when cost-benefit analyses indicate significant overall savings for society? If you're a total libertarian government-should-do-nothing type, then you won't be happy with this. On the other hand, this is something the "free market" will not fix on its own: employers, especially, will be too happy to offload risks onto their employees, and demand people come in to work (with the employee taking the added risk of personal injury or death for the employer's benefit) --- the same type of safety issues that unregulated markets tend to fail spectacularly at. The economic criteria and cost-benefit analyses for making such decisions should be made public and transparent (and subject to voter modification should people disagree with the government technocrats' decision), but I don't think it's out-of-place (or a slippery slope to fascism) for the government to regulate "general public welfare" safety concerns like road travel during dangerous weather conditions.
Very few companies of any sort are running at 100% capacity all the time, and limited in their output by the number of hours in the day --- basically, whenever someone hits this point, it's profitable for them (or a competitor) to expand operations, until economic output is limited by how many units (of whatever one is producing; goods or services) one can sell at positive margins instead of how much can be produced at maximum capacity in a day. At worst, you have to spend a tiny bit more on overtime pay (a small fraction of total production costs) to catch up. Basically nothing public consumer-facing will cause overall economic loss, because even if customers don't return to the same place they'll still spend their money somewhere. And most businesses selling to other businesses have fixed order sizes, that they can make up with higher production rates (or a few extra hours) on other days. And again, most money not spent in one place still gets spent in another --- net economic damage only occurs when something is actually destroyed (e.g. food that spoils during the "missing" day) rather than shuffled around to another time/place.
And I'm sure a local food service establishment could have supplied food too. Oh wait, that's what happened. Aside from the "derp, cops love donuts!" funny factor of the particular choice of vendor, what happened seems like a perfectly fine and practical use of available resources. If the cops had called in a military convoy to do their coffee runs, many of the same folks complaining about what did happen would be whining instead about the massive waste of taxpayer money to send a squadron of hummers with MREs instead of just using local food vendors.
I don't know how intrusive the emergency drills on your campus are, but the ones I've seen really don't stop much productive work. It's not "everyone cower under their desk in the fetal position and tremble in silence until the all-clear is announced," but "stay in the room doing what you were doing anyway, and hold off on any trips into the hall/outdoors that you can wait on." No one gets tackled by the campus police for needing to get to the bathroom, or puts their experiments on hold, or stops their lecture, or quits working on homework.
In retrospect, it's interesting that the bomber didn't kill more people when they actually had the chance. During their escape, they held up a convenience store and stole a car --- without shooting the robbery victims. An interesting artifact of human psychology, even at its most twisted: the terrorists willing to blow up random strangers weren't willing to look a shopkeeper or driver in the eye and shoot them; in panicked flight and personal contact with potential victims, they showed far more restraint and respect for human life than their premeditated impersonal cold-blooded murders just hours before.
So, making sure officers on an intensive stakeout can get a quick calorie boost is a bad thing (herp derp, cops and donuts!)? Would it have been better to tell the police force "everyone go grab lunch in a far-away safe location, and be back in two hours"? Or, should the police force maintain their own mobile food preparation facilities, so they're ready to bring their own snack trucks to the site of once-in-a-few-years major operations? Keeping a close-by source of quick-energy-rush stay-awake food available (by the simple and easy method of obtaining help from local businesses already expert in that field) actually seems like a sensible idea --- or do you think a cop shouldn't get hungry after 14 hours stressful work?