1) The ability to highlight the area of a graph and paste it into a spreadsheet and having it show up as real data, not graphics.
Not a real killer app - we've had data mining / visualization / slice 'n' dice packages for over twenty years now. Sadly, none of them ever expand beyond a niche market because:
1) Most users can't interpret 2-D data (other than simple time series and quartile-type histograms.) Many people can't even interpret 2-D data (ask a person to explain a graph of unemployment claims data and you will be unpleasantly surprised.)
2) Most firms that examine complex, high-dimensional data (e.g. insurance companies, wall-street banks, economic think-tanks,) already have seriously sophisticated, domain-tailored tools. Wow, end-of-summer sales of pencils are up in sales district X - I wonder why? You don't think Staples already has some tools for correlation for back-to-school student buying with store-sales figures? Executives will greet this tool with a big yawn.
"You realized that citizens understand these simple economic principles. Yet our governments print fiant currencies like mad. I think it's reasonable for folks to try and do a little educating on the subject. Maybe some of this awareness will trickle up?"
Actually, our government has an exceedingly good understanding of these principles, its the citizens that generally don't. The governments knows that most citizens like a tax cut (Wow! free money for me!) yet are incapable of understanding what effect that will have on their future standard of living.
The Mises Institute starts with some simple truths (money is just the most convenient commodity,) and then reasons itself into the twilight zone. E.g. "prices can change from day to day, but today's new prices will be based on the prices of other things yesterday." Hmm, guess they forgot about things like futures markets that have only been around for a thousand years or so.
Their Iraq example is cute, but it's a very short term effect -- injecting new money into an economy inflates the money, it doesn't make the fiat currency a good long term holder of wealth -- the sophisticated investors will profit, the crooks will profit, the average person will lose money overall.
Actually, tetrachromatic vision DOES give you the ability to see colors that other people can't. We only represent the color-space as 3-dimensional because most people have 3 color receptors. If we were, for example, mantis shrimp, we would see color as a 10+ dimensional space.
Note that color vision is very different from hearing -- we can hear multiple tones at the same time - color vision is just the summation of various light energies captured by receptors attuned to different frequencies. Perfect color vision would essentially require an infinite number of tuned receptors.
Correct. You can't copyright a real word. You can't even trademark it: Microsoft has a trademark on "Microsoft Windows," not "Windows." No matter how much they'd like the generic word, they can't have it.
Economists at least have a lot of raw data from which to generate incorrect conclusions:) Almost no data is available on specific hospitals and doctors tho.
Actually, our delivery turned out to be quite exciting: my wife had placenta previa, and so labor would have been fatal. The C-section required cutting through all the major heart->placenta blood vessels. I was in the OR when this was done, and in was like a Sam Peckinpah movie. Everything turned out ok, but I was definitely worried when the chief surgeon asked for 2 pints of transfusion blood in the OR before the whole thing started.
Yes, and they also try to keep them alive. Consider this:
"The Chinese had thrown thousands of men against it, but the company held like a rock. The unit was cut off from its battalion, isolated deep in enemy territory; battered and bruised it held on, and when ammo and food ran out, the troupers lived off the land and used captured weapons. They built a barrier wall of tree branches, like the men at Valley Forge; the Chinamen came and the Chinamen went and the valient unit did not give an inch, until finally a U.S. tank force broke through and extricated it."
That was Lt. Doug Anderson's company in Korea. He was put up for the Medal of Honor, but didn't get it. If you think the men in that battle preferred eating grass to a ration that becomes edible with muddy water or urine, you should read a little military history.
Question the morals of those in charge, but thank every scientist who gives the grunts a way to get a meal better than "centipedes in chili sauce" when things go pear-shaped.
(the quote was from "About Face" by Lt Col Hackworth.)
Good luck trying to get data like this. The medical profession is the absolute worst for providing objective data. The basic reasons for this are:
Patient confidentially - hard to get raw data
Certification is binary - you're a doctor or not a doctor
Most funding for studies is from drug companies - little incentive to do broad studies
Massive survivorship bias in available data - if you try to treat very sick patients, you look like a murderer in the statistics
Only when the medical profession gets seriously insane (e.g. tonsilectomies for all,) do scientists really manage to investigate what's going on, and even then they are fought tooth and nail by the profession.
When my wife decided to go the minimal intervention childbirth route, we talked to a lot of hospitals: all gave us glossy brochures explaining how they were really good. We demanded, and got, the C-section rate tables from them as well. The common reason for a C-section is "failure to progess," i.e. the child ain't coming out quick enough. So, I wanted to know how good each hospital was: "how many failure-to-progress c-sections happen on Friday and Monday vs. Saturday and Sunday?"
The nurse talking to us couldn't understand why I might even ask the question. When I explained, she said they didn't have that data and was surprised that I might even consider that their c-section decisions weren't completely objective.
We had the kid (and I love her,) but another data point is hidden from the public.
The dictator/terrorist doesn't want a cheap Tomahawk, he wants a 200 mile range missile that can hit its target. He wants easy launch by a relatively untrained crew.
Against a technologically inferior foe, he's ahead of the game.
Against a superior foe (e.g. Israel), he wants them spending jet money to intercept WW2 tech. Build cheap, launch many, hope a few hit. Even if all are intercepted, you win if the cost to intercept is much bigger than the cost to launch.
Frankly his pulse jet propulsion is not going to give you anything close to the range of the tubofan that real cruise missles use. The will also tend to be slow. Max speed of around 400 knots. Not likley to be stealthy. Pretty much an updated WWII V-1.
Still, a very scary weapon if you could build and launch at around $20K a pop. Add a GPS, make it ground-hugging, and you have a weapon that is 100 times as cost effective as a Scud.
Something that just about any college grad could whip up.
Something just about any college grad could *understand*. Something that 1 in 5 college students could design given a year. Something that no college student could "whip up."
Engineering is very different from theory. Given a spec for a missile, I could "whip up" a design in 20 minutes or so: payload, range, target manuevability, maximum weight, maximum diameter, prep-time, etc, dictate the basic design (shape, fuel load, control surfaces, etc.) Ask five designers, and they'll all whip up roughly the same design. "Whipping up" the actual missile is a completely different proposition - expect 10+ man-years to produce something that is reliable, manufacturable, and cost-efficient.
Someone with all-around expertise in building pulse-jets could head a project that would start producing serious prototypes really fast. This is technology transfer at its scariest.
A simple deep-well pump would be a motor at the wellhead that drives a loop of rope that reaches to just below the water. Attach some buckets to the rope, add a sluice at the head, and you are done.
Note that one definition of "pump" is "a device that raises water." We are so used to high-power pumps that we tend to think of them as things that pressurize and accelerate fluids using blades, etc.
I realize that McDonald's is a corporation, and as such, all right thinking Slashdot posters must hate them with spittle-spewing passion, but doesn't the lady in question have any fault in the matter, no matter how small?
She definitely had some fault in the matter: she should have used a table or something when adding the milk and sugar to her coffee.
But the common thread is that the lady in question placed a relatively flimsy container of hot liquid in her lap. The flimsiness of the container or exact temperature of the liquid is beside the point.
Well, from a legal standpoint, neither the flimsiness of the container or the exact temperature are beside the point. McDonalds knew about the flimsiness and the ability of the very hot coffee to cause extreme burns. A reasonable person would not expect the cup to collapse or the coffee to cause third degree burns.
So, both parties were at fault to some degree. Given the woman initially only negotiated for her costs (the medical bills,) and not "pain and suffering" or punitive damages, you have to wonder what on earth was going through McDonalds' minds when they decided to play hardball. Any court would find them at least partially liable for the harm: one phone call to check that she is not a habitual lawsuit filer, then have your insurance pay it, give blessings that it's just the medical bills, and move on.
So you are claiming that buying a cup of coffee is analogous to playing Russian roulette?
I'm going to add sugar to my drink. Obviously I have a one in six chance of dying now. Milk? I practically deserve to die at this point - 11/36 people die every day from this high-risk activity.
You all do know theres a web site about bogus legal cases names after the woman in this case, right? http://www.stellaawards.com/
Well, I read the court's actual decision, but thanks for the pointer to a subscription only site that declares that it "means to be entertaining." Did you read the original documents?
Exactly what, then, did McDonald's do wrong? Did it exhibit "willful, wanton, reckless or malicious conduct" -- the standard in New Mexico for awarding punitive damages?
Willful - yes. Wanton - yes. Reckless - yes. Malicious - no. It simply cut corners on serving coffee. The damages it had to pay where less than the profits it earned in one hour of operation. Hardly something to get excited about.
It was substantially hotter than coffee served at other restaurants in the area. McDonalds had also been repeatly warned that the coffee was being served too hot for the cups they served it in (some would just collapse based on the heat.)
You're conflating two different ideas: criminal law and contracts. Killing people falls under criminal law, requirements for graduation falls under contract law.
To take your (extreme) case of serial killers, the courts in the USA have often held that an unenforced law is unenforable. If no one has been prosecuted under a law for many years, the law is usually considered obsolete. E.g. if cars fall under the same rules as horses, failing to tether your car to a post when you go shopping is not going to get you convicted of a crime. The courts are highly suspicious of selective prosecution using old laws that are still on the books.
Non-criminal behaviour requires prompt action if a party wants to preserve its rights. If you keep a dog in a "no pets" apartment, but I ignore your action, you get to keep doing it. If I ignore you using my front lawn as a shortcut to the bus-stop, yet take no action, after a while it becomes a right of way. If I let you camp in my back yard for a year, it's your dwelling.
Filing suit seems reasonable here: it lets the plagurist find out what the school knew, and when it knew it. If there was bad faith, he has a good chance of winning, if not, then end of suit. Bear in mind that the school's role is more than just the awarder of a diploma, it's also meant to educate him. If it failed to educate him in the subtilties of copying, then it failed.
It's a valid point to say that people convicted of a felony shouldn't be allowed to vote.
I find this statement very confusing: if the government convicts you of a crime, it also removes your ability to have a say in what constitutes a crime? How can this possibly be valid (i.e. reflect the will of the people?)
Striping a rapist of his right to vote seems to accomplish little: no politician is running on the right to rape platform. The rapist will probably vote just like everyone else.
Striping a low-level drug dealer of his right to vote means no politician need consider liberalizing the drug laws: the constituency has already plea-bargained away its voting power.
As you note, this "valid point" disenfranchises the blacks, and also the poor. This is neither valid nor good.
Print statements are a great tool, especially on large pieces of software maintained/enhanced by many people. Once you've debugged your problem, you just #ifdef out the prints, and check the code back into version control.
When the next poor programmer comes along, trying to fix/find a bug in that code, he a) can #ifdef the prints back on and quickly get debugging output about the important events taking place in his run, and b) read the code and see where the hairy bits are, because they tend to be the sections most heavily littered with debugging print calls.
Fancy debugger IDEs just don't support this preservation of institutional knowledge.
The chance of an alien microbe getting a toehold on Earth is pretty much zero. Most every niche is filled by a lifeform that has four billion years worth of ancestors that didn't die before they breed.
If a Martian microbe shows up, it gets eaten in 30 seconds by some terrestrial super-optimised (for earth) bug. The martian bug's super radiation protection, cold-protection, etc, just means it has misallocated resources for the terrestrial environment.
Notice that people get infected by bugs that have evolved attacking animals similar to people (e.g. primates, mammals, some birds.) The nasty ones come from animals similar to us. We have little to fear from reptile bug, less from plant bugs, and nothing to fear from things that attack fungi. Martian bugs would be like tourists from Iowa trying to infect New York City.
Huh? Why quote US trademark law when Lindows has already won in the US. Microsoft didn't trademark "windows," they trademarked "Microsoft Windows." The "windows" part is a generic term that had prior use. Under US law, no amount of market share can convert a generic term into a trademarkable term.
Non-flashy is easier to define: stuff that could actually produce useful results in a reasonable time-frame. Flashy is anything not in that set.
In the early 80s, Yale seemed to have a few good avenues of non-flashy research and engineering going:
1) T - a solid, interesting Scheme-like language. 2) Hardware - essentially the prototype ELI machines. 3) OSes - Lisp environments vs DecSystem20 vs Unix. 4) Graphics/Realtime stuff - mostly due to Apollos. 5) Networked computers - the proto-internet was exciting. 6) User modelling - i.e. what does a user think he's doing?
Everything else was basically flashy or pointless junk. The worst offender was Schank - a producer of brittle demos that got research dollars and did zero to advance CS. Actually, Yale probably set back CS by its contribution to the AI winter of the mid to late 80s.
Slashdot Mirror
1) The ability to highlight the area of a graph and paste it into a spreadsheet and having it show up as real data, not graphics.
Not a real killer app - we've had data mining / visualization / slice 'n' dice packages for over twenty years now. Sadly, none of them ever expand beyond a niche market because:
1) Most users can't interpret 2-D data (other than simple time series and quartile-type histograms.) Many people can't even interpret 2-D data (ask a person to explain a graph of unemployment claims data and you will be unpleasantly surprised.)
2) Most firms that examine complex, high-dimensional data (e.g. insurance companies, wall-street banks, economic think-tanks,) already have seriously sophisticated, domain-tailored tools. Wow, end-of-summer sales of pencils are up in sales district X - I wonder why? You don't think Staples already has some tools for correlation for back-to-school student buying with store-sales figures? Executives will greet this tool with a big yawn.
"You realized that citizens understand these simple economic principles. Yet our governments print fiant currencies like mad. I think it's reasonable for folks to try and do a little educating on the subject. Maybe some of this awareness will trickle up?"
Actually, our government has an exceedingly good understanding of these principles, its the citizens that generally don't. The governments knows that most citizens like a tax cut (Wow! free money for me!) yet are incapable of understanding what effect that will have on their future standard of living.
The Mises Institute starts with some simple truths (money is just the most convenient commodity,) and then reasons itself into the twilight zone. E.g. "prices can change from day to day, but today's new prices will be based on the prices of other things yesterday." Hmm, guess they forgot about things like futures markets that have only been around for a thousand years or so.
Their Iraq example is cute, but it's a very short term effect -- injecting new money into an economy inflates the money, it doesn't make the fiat currency a good long term holder of wealth -- the sophisticated investors will profit, the crooks will profit, the average person will lose money overall.
Actually, tetrachromatic vision DOES give you the ability to see colors that other people can't. We only represent the color-space as 3-dimensional because most people have 3 color receptors. If we were, for example, mantis shrimp, we would see color as a 10+ dimensional space.
Note that color vision is very different from hearing -- we can hear multiple tones at the same time - color vision is just the summation of various light energies captured by receptors attuned to different frequencies. Perfect color vision would essentially require an infinite number of tuned receptors.
Correct. You can't copyright a real word. You can't even trademark it: Microsoft has a trademark on "Microsoft Windows," not "Windows." No matter how much they'd like the generic word, they can't have it.
Economists at least have a lot of raw data from which to generate incorrect conclusions :) Almost no data is available on specific hospitals and doctors tho.
Actually, our delivery turned out to be quite exciting: my wife had placenta previa, and so labor would have been fatal. The C-section required cutting through all the major heart->placenta blood vessels. I was in the OR when this was done, and in was like a Sam Peckinpah movie. Everything turned out ok, but I was definitely worried when the chief surgeon asked for 2 pints of transfusion blood in the OR before the whole thing started.
Yes, and they also try to keep them alive. Consider this:
"The Chinese had thrown thousands of men against it, but the company held like a rock. The unit was cut off from its battalion, isolated deep in enemy territory; battered and bruised it held on, and when ammo and food ran out, the troupers lived off the land and used captured weapons. They built a barrier wall of tree branches, like the men at Valley Forge; the Chinamen came and the Chinamen went and the valient unit did not give an inch, until finally a U.S. tank force broke through and extricated it."
That was Lt. Doug Anderson's company in Korea. He was put up for the Medal of Honor, but didn't get it. If you think the men in that battle preferred eating grass to a ration that becomes edible with muddy water or urine, you should read a little military history.
Question the morals of those in charge, but thank every scientist who gives the grunts a way to get a meal better than "centipedes in chili sauce" when things go pear-shaped.
(the quote was from "About Face" by Lt Col Hackworth.)
Only when the medical profession gets seriously insane (e.g. tonsilectomies for all,) do scientists really manage to investigate what's going on, and even then they are fought tooth and nail by the profession.
When my wife decided to go the minimal intervention childbirth route, we talked to a lot of hospitals: all gave us glossy brochures explaining how they were really good. We demanded, and got, the C-section rate tables from them as well. The common reason for a C-section is "failure to progess," i.e. the child ain't coming out quick enough. So, I wanted to know how good each hospital was: "how many failure-to-progress c-sections happen on Friday and Monday vs. Saturday and Sunday?"
The nurse talking to us couldn't understand why I might even ask the question. When I explained, she said they didn't have that data and was surprised that I might even consider that their c-section decisions weren't completely objective.
We had the kid (and I love her,) but another data point is hidden from the public.
The dictator/terrorist doesn't want a cheap Tomahawk, he wants a 200 mile range missile that can hit its target. He wants easy launch by a relatively untrained crew.
Against a technologically inferior foe, he's ahead of the game.
Against a superior foe (e.g. Israel), he wants them spending jet money to intercept WW2 tech. Build cheap, launch many, hope a few hit. Even if all are intercepted, you win if the cost to intercept is much bigger than the cost to launch.
Frankly his pulse jet propulsion is not going to give you anything close to the range of the tubofan that real cruise missles use. The will also tend to be slow. Max speed of around 400 knots. Not likley to be stealthy. Pretty much an updated WWII V-1.
Still, a very scary weapon if you could build and launch at around $20K a pop. Add a GPS, make it ground-hugging, and you have a weapon that is 100 times as cost effective as a Scud.
Something that just about any college grad could whip up.
Something just about any college grad could *understand*. Something that 1 in 5 college students could design given a year. Something that no college student could "whip up."
Engineering is very different from theory. Given a spec for a missile, I could "whip up" a design in 20 minutes or so: payload, range, target manuevability, maximum weight, maximum diameter, prep-time, etc, dictate the basic design (shape, fuel load, control surfaces, etc.) Ask five designers, and they'll all whip up roughly the same design. "Whipping up" the actual missile is a completely different proposition - expect 10+ man-years to produce something that is reliable, manufacturable, and cost-efficient.
Someone with all-around expertise in building pulse-jets could head a project that would start producing serious prototypes really fast. This is technology transfer at its scariest.
Maybe the English-speaking reporter read the first entry at dictionary.com or opened his Oxford English Dictionary and read:
Pump (pronounciation and derivation)
Machines for raising water were in ancient and medieval use...
A simple deep-well pump would be a motor at the wellhead that drives a loop of rope that reaches to just below the water. Attach some buckets to the rope, add a sluice at the head, and you are done.
Note that one definition of "pump" is "a device that raises water." We are so used to high-power pumps that we tend to think of them as things that pressurize and accelerate fluids using blades, etc.
One atmosphere of pressure is about 10 meters of water. You can't pump water any higher than that. I smell exaggeration.
Um, you can't suck water higher than 10 meters. How to you think water gets to the toilets on the top floor of a skyscraper?
I realize that McDonald's is a corporation, and as such, all right thinking Slashdot posters must hate them with spittle-spewing passion, but doesn't the lady in question have any fault in the matter, no matter how small?
She definitely had some fault in the matter: she should have used a table or something when adding the milk and sugar to her coffee.
But the common thread is that the lady in question placed a relatively flimsy container of hot liquid in her lap. The flimsiness of the container or exact temperature of the liquid is beside the point.
Well, from a legal standpoint, neither the flimsiness of the container or the exact temperature are beside the point. McDonalds knew about the flimsiness and the ability of the very hot coffee to cause extreme burns. A reasonable person would not expect the cup to collapse or the coffee to cause third degree burns.
So, both parties were at fault to some degree. Given the woman initially only negotiated for her costs (the medical bills,) and not "pain and suffering" or punitive damages, you have to wonder what on earth was going through McDonalds' minds when they decided to play hardball. Any court would find them at least partially liable for the harm: one phone call to check that she is not a habitual lawsuit filer, then have your insurance pay it, give blessings that it's just the medical bills, and move on.
As has been noted, in the case in question:
1) the cup wasn't paper.
2) the car wasn't moving.
So you are claiming that buying a cup of coffee is analogous to playing Russian roulette?
I'm going to add sugar to my drink. Obviously I have a one in six chance of dying now. Milk? I practically deserve to die at this point - 11/36 people die every day from this high-risk activity.
(1) That was hyperbole.
Ooh, no, sorry. Thanks for playing.
You all do know theres a web site about bogus legal cases names after the woman in this case, right? http://www.stellaawards.com/
Well, I read the court's actual decision, but thanks for the pointer to a subscription only site that declares that it "means to be entertaining." Did you read the original documents?
Exactly what, then, did McDonald's do wrong? Did it exhibit "willful, wanton, reckless or malicious conduct" -- the standard in New Mexico for awarding punitive damages?
Willful - yes. Wanton - yes. Reckless - yes. Malicious - no. It simply cut corners on serving coffee. The damages it had to pay where less than the profits it earned in one hour of operation. Hardly something to get excited about.
It was substantially hotter than coffee served at other restaurants in the area. McDonalds had also been repeatly warned that the coffee was being served too hot for the cups they served it in (some would just collapse based on the heat.)
So, yes, too fucking hot sums it up well.
You're conflating two different ideas: criminal law and contracts. Killing people falls under criminal law, requirements for graduation falls under contract law.
To take your (extreme) case of serial killers, the courts in the USA have often held that an unenforced law is unenforable. If no one has been prosecuted under a law for many years, the law is usually considered obsolete. E.g. if cars fall under the same rules as horses, failing to tether your car to a post when you go shopping is not going to get you convicted of a crime. The courts are highly suspicious of selective prosecution using old laws that are still on the books.
Non-criminal behaviour requires prompt action if a party wants to preserve its rights. If you keep a dog in a "no pets" apartment, but I ignore your action, you get to keep doing it. If I ignore you using my front lawn as a shortcut to the bus-stop, yet take no action, after a while it becomes a right of way. If I let you camp in my back yard for a year, it's your dwelling.
Filing suit seems reasonable here: it lets the plagurist find out what the school knew, and when it knew it. If there was bad faith, he has a good chance of winning, if not, then end of suit. Bear in mind that the school's role is more than just the awarder of a diploma, it's also meant to educate him. If it failed to educate him in the subtilties of copying, then it failed.
It's a valid point to say that people convicted of a felony shouldn't be allowed to vote.
I find this statement very confusing: if the government convicts you of a crime, it also removes your ability to have a say in what constitutes a crime? How can this possibly be valid (i.e. reflect the will of the people?)
Striping a rapist of his right to vote seems to accomplish little: no politician is running on the right to rape platform. The rapist will probably vote just like everyone else.
Striping a low-level drug dealer of his right to vote means no politician need consider liberalizing the drug laws: the constituency has already plea-bargained away its voting power.
As you note, this "valid point" disenfranchises the blacks, and also the poor. This is neither valid nor good.
Print statements are a great tool, especially on large pieces of software maintained/enhanced by many people. Once you've debugged your problem, you just #ifdef out the prints, and check the code back into version control.
When the next poor programmer comes along, trying to fix/find a bug in that code, he a) can #ifdef the prints back on and quickly get debugging output about the important events taking place in his run, and b) read the code and see where the hairy bits are, because they tend to be the sections most heavily littered with debugging print calls.
Fancy debugger IDEs just don't support this preservation of institutional knowledge.
The chance of an alien microbe getting a toehold on Earth is pretty much zero. Most every niche is filled by a lifeform that has four billion years worth of ancestors that didn't die before they breed.
If a Martian microbe shows up, it gets eaten in 30 seconds by some terrestrial super-optimised (for earth) bug. The martian bug's super radiation protection, cold-protection, etc, just means it has misallocated resources for the terrestrial environment.
Notice that people get infected by bugs that have evolved attacking animals similar to people (e.g. primates, mammals, some birds.) The nasty ones come from animals similar to us. We have little to fear from reptile bug, less from plant bugs, and nothing to fear from things that attack fungi. Martian bugs would be like tourists from Iowa trying to infect New York City.
Huh? Why quote US trademark law when Lindows has already won in the US. Microsoft didn't trademark "windows," they trademarked "Microsoft Windows." The "windows" part is a generic term that had prior use. Under US law, no amount of market share can convert a generic term into a trademarkable term.
"Watson, I need you."
"Watson, tell me what you're wearing."
Non-flashy is easier to define: stuff that could actually produce useful results in a reasonable time-frame. Flashy is anything not in that set.
In the early 80s, Yale seemed to have a few good avenues of non-flashy research and engineering going:
1) T - a solid, interesting Scheme-like language.
2) Hardware - essentially the prototype ELI machines.
3) OSes - Lisp environments vs DecSystem20 vs Unix.
4) Graphics/Realtime stuff - mostly due to Apollos.
5) Networked computers - the proto-internet was exciting.
6) User modelling - i.e. what does a user think he's doing?
Everything else was basically flashy or pointless junk. The worst offender was Schank - a producer of brittle demos that got research dollars and did zero to advance CS. Actually, Yale probably set back CS by its contribution to the AI winter of the mid to late 80s.