I said 'many people don't consider problem-solving to be a sign of intelligence'. I did not say 'most'. The overwhelming majority do, and llike BMI, for most people, in most contexts, it's mostly accurate for what most people consider to be intelligence.
I think your analogy works very well. BMI is a fine overall statistic for plunking onto a chart in a power point presentation about how we might want to allocate more money to health education due to increasing teen health problems. But it easily fails when used for individuals, where there are many other more useful measures on hand. We should endeavor to not be like the dumbass doctor who looked at my young son's BMI number and did not actually look at him before pronouncing he should lose some weight. (My son is having a little difficulty learning to swim because he naturally sinks below the surface -- which means it is impossible for his body fat level to be unreasonable. A doctor astute enough to actually look at her patient could have easily figured that out.)
I agree. Most opinions about how terrible robots will forever be at driving vehicles vastly overestimate how competent hominids really are when behind the wheel in tricky situations, too. There are a lot of tells about potential problems that a computer could pick up consistently and then choose to slow down consistently, thereby both mitigating both the future potential accident as well as buying time to assess the situation correctly. Hominids routinely fail to slow down in ambiguous situations, in the vain hope that two eyeballs will figure it out in time if things go south.
Further, everyone knows what rampant inflation looks like, and why it leads to war and revolution. Nobody cared when it was runaway deflation, because holding a coin for two days doubled your wealth.
I'm pretty sure that we have no real idea what business looks like with deflationary currency. We can make educated assumptions. One thing is for sure, savings would be more beneficial than spending, and credit would largely be rare, rather than common.
Your "pretty sure" is simply wrong. In roughly the same historical period we saw an advanced economy get smacked by hyperinflation and fall into chaos then war, the USA itself was getting hammered by deflation. It was a wonderful time when American shoe factories laid off workers to starve in the streets while barefoot farmers were kicked off their farms, all to serve the idolatry of gold.
Gold fetishists have brainwashed certain people into believing that inflation is super scary while deflationary contractions are purely a theoretical problem. That is because they lie and some sheep are always eager to believe them.
Miners not prioritize transactions without a fee attached. The days when bitcoin transfers are free are years behind us.
There is another problem regarding bitcoin mining: the majority is in China where it is clear as mud if they are actually businesses that could survive under normal economic conditions. Even if someone gave me the most up to date asics in a new server, I could not pay even a quarter of my electrical bill running the thing. What are the costs that Chinese miners are paying? Or are they just stealing electricity by the consent of silent partners who are politically connected? Even if this extreme is not the case, China's electrical power costs are trending up, as they move away from coal fired plants and the demand for electricity continues to grow.
This all adds up to a hidden subsidy of bitcoin. What happens if the subsidy is removed? Transaction costs go up a lot, which will depress bitcoins usage by non-speculators.
I once had an identity theft incident, nearly 25 years ago, where a couple credit cards were taken over (mailing address changed, and new copies of credit cards shipped) and a few new credit card accounts opened. I caught the problem early enough that the damage was minor, more nuisance than financial.
But a little digging and simple deduction led me to this conclusion: someone gained access to my full name, mailing address, SSI, mother's maiden name, and multiple open credit card account numbers. Now where in the universe can this entire array of data be found in one place? My bank? Nope. University? Nope. Employer? Nope. Any one credit card company? Nope.
You guessed it! Credit reporting agencies.
Now it is theoretically possible that an energetic fraudster could gain a few tidbits and build from there. But the dirty little secret sitting in plain sight is the financial institutions themselves, particularly anyone and everyone who deals with credit card data, are the most likely places to leak. That has been true since forever, and it has not changed and will not change, until the laws are very different.
From my POV, most credit protection services are really a sick joke: "As your personal bank or personal credit card company, we would be happy to collect a monthly fee to mitigate the risks you bear due to our incompetence."
To some degree, I try to keep an open mind to engineering that seems to be the right tool for the problem on hand. But I do believe that this sloppy kind of engineering that is not much more than taking dirty data and making it somewhat less dirty will prove very easy to replace. Some day there will be an Amazon service where one business data analyst with an AI assistant can try out a hundred different correlation models in an afternoon, and deploy it out to a 1024 AWS cluster on a 1AM crontrigger while the information worker is having a fourth drink at the strip joint.
I recently interviewed at a couple of the new fangled big data marketing startups that correlate piles of stuff to help target ads better, and they were continuously deploying up the wazoo. In fact, they had something like zero people doing traditional QA. It was not totally insane at all. But they did have a blaze attitude about deployments -- if stuff don't work in production they just roll back, and not worry about customer input data being dropped on the floor. Heck, they did not worry much about data losses even when things were working "correctly". To some people big data means no data really matters much, I guess.
That is a fine way to make money when these things are new and expectations are low, and there are easy ways to pick off a tiny slice of that googlenormous flow of ad money to achieve that IPO bliss. I could see AI helping out here.
But continuous delivery can easily suck for code that actually has to work so that people do not get fired or go bankrupt or even die. Continuous AI buzzwordology is not ready for this world.
It does not necessarily matter that the data is dirty. It only matters that globs of money can be spent a tiny bit more efficiently than the crappy way it was spent 20 years ago, to make this new kind of advertising exciting...to people who care about advertising.
For example, big automakers spent a few hundred million a year for advertising. They believe this kind of budget is in the right ballpark based on decades of experience. They are not naive. They do understand that individual advertising efforts within this big budget are flops, but what is going to be a flop is only clear in hindsight.
So we are literally talking about companies like GM looking at expanding their $350 million ad budget from last year to $360 million this year, and wondering whether to increase their "online" spending from $80 million to $90 million. So the data may be dirty but it could still be money better spent than a few more glossy magazine ads.
This. It is a good practice to send out an email after meetings to say "This is what I walked away with as our mutual understanding." Not so much because of illegal shenanigans, but because misunderstandings are common, even among honest and competent people. The honest and competent people appreciate your effort to keep you all on the same page, most especially when things are confusing. The dishonest will become tetchy.
You are being selective in your reading. Making the point that women tend to be neurotic was stupid, and a legitimate reason discontinue his employment as it erodes confidence in his ability to perform his job duties appropriately. Google does not have to prove "cause" here, merely an adequate reason that reasonably sustains a loss of confidence by his employer. That he may have made 99 other non-stupid arguments that are protected by CA law does not give him a legal blank check to sneak in a stupid (unprotected) argument that harms the business.
The problem with being a rambling fool is it only take one single poorly chosen sentence to annihilate his argument that Google did him wrong in the eyes of the court.
You are not connecting your own dots there. The property is a physical piece of land. If one single person can testify they crossed that land for years to get to the public beach without asking permission, that is sufficient grounds for a judge recognizing an easement. The particular usage of the land by the previous owners really does not matter. "Oh, we ran a business so we ignored the guy with the surfboard under his arm walking through our parking lot" is not a counterargument, but additional evidence that the easement exists. The specific usage of land in no way automatically obliterate the customs recognized under English Common Law.
Mod the parent up. It does not really matter what the title says. Reasonable evidence that there was ongoing access to the publicly owned beach is all that is necessary to prove the title was flawed.
Easements based on historical usage have been a thing in English Common Law for several centuries. In CA, like most US states, English Common Law precedents apply unless explicitly changed by legislative action.
Failure of the property owner to understand the law of the land is not a taking under the Constitution.
I think the argument that Google is making certain purported objectives harder has merit, which is not to say that I necessarily agree with that argument, only that I do respect it.
That Damore was _fired_, not for making overtly controversial remarks but for saying "Hey, maybe we should implement policies that have some scientific validity behind them" shows that discourse on the subject has reached Johnathan Swift big endian/small endian levels of ridiculousness.
How would you know? Damore made both controversial remarks and a call for better kinds of discourse. Both arguments happen to have been made in a very sloppy way. It is not obvious (to me) what Google's precise reasoning was here. They do not have to tell us, unless forced to under the bright lights of a courtroom.
Should you disagree, give quotations of the sections of his memo that say explicitly that he is against diversity and hostile to women.
As for hostile to women, we have his non-argument here:
Hiring practices which can effectively lower the bar for "diversity" candidates by decreasing the false negative rate
Reducing false negatives does not automatically change "the bar" in any particular direction. Speculating certain of your colleagues are inferior quality hires based on tortured (non)logic is arguably hostile.
Damore's logic is very sloppy in a number of places. That is why it gets labelled a rant, fairly or unfairly. Because he apparent joy in blabbering about things poorly connected to his alleged main point seems like a hint.
It basically is an argument against what he perceives as so-called "political correctness" in pursuit of gender equity. He tries to be more clever than that but that's what he's saying. He claims to be a "classical liberal" but it's pretty clear that he is not based on his arguments. He makes population arguments and then assumes these must apply on the individual level.
And that is why I am having zero sympathy for him. Bringing in the discussion of population statistics muddies his own supposedly non-sexist argument. He could have simply stated "For whatever reason, from the larger world, the pipeline of excellent engineers that shows up at our door happens to be skewed. Google should treat these special individuals as individuals." That could have cut out a third of the piece, and gotten down quickly to specific practices he wants to question.
Even when he is trying to make his point, he adds junk like...
Hiring practices which can effectively lower the bar for "diversity" candidates by decreasing the false negative rate
Reducing the false negatives is somehow bad because? Amending recruiting practices for some candidates might be "unfair" by some personal ethical definition of the concept, but it is clear as mud whether what he describes actually lowers the bar. Heck, it might even raise both the average quality and number of female hires, for all we know.
Nearly all of our wasted time comes from trying to navigate a context switch.
Even if true, it may not matter in way anyone should care. Because your employer cares about your team's overall productivity, not your productivity. Good software companies in the 21st century are not infatuated with engineers churning out huge piles of code, but with a merely good amount of code productivity that meets the real needs.
Code that is built wrong, in either the wrong way or the towards the wrong goal, often has near zero or even negative value, once you factor in the opportunity costs.
If you are working on a problem that is genuinely without ambiguous requirements, and if all the people who dependencies on your work know exactly what to expect, and if you really are completely competent to get it done without input from anyone else, if all that is true, then, yes, distracting you is very expensive.
But when there are lots of ambiguities, then many of those "distractions" are useful events increasing other people's productivity. Certainly it is possible for that to get way out of hand, but it is not automatically true that open office spaces will cause such.
80% of your education from primary school to a master's degree could have been self-taught from books, once you mastered the basics of reading. So, your point makes sense, but it is not exactly a surprise that sophisticated computer programs can be a very effective replacement for books.
Getting back to your main point, what you suggest can be done, but it does run into two significant bumps in the road: (1a) It is much more expensive to do this well than is generally realized, and (1b) the world is already flooded with mediocre teaching software with more on its way, which is pushing back the necessary effort to do these things right, and (2) for all it is quite possible to learn material from computers, it is also very much easier to cheat. For example: Khan Academy is an excellent resource, but it has failed to gain acceptance as a "real" school that can certify students in the manner employers are used to (which is actually a pretty low bar).
Active learning is actually a harder form of learning in that in requires more work from both student and teacher.
That is why the mediocre are so certain that active learning does not work -- they have never seen it really tried because their only experience as a student suffers the limitation of being who they are.
Furthermore, poorly implemented active learning can easily be worse than the traditional lecture. The smart ones do the real work, while half their teammates snooze or hold text message conversations on their phone, and then copy the work of their betters.
Part of the problem is active learning is touted as a "replacement" for lectures. It is really an entirely different way to organize teaching, where the disappearance of lectures is an inevitable but incidental result. The point is not to get rid of lectures, but to really put the nose to the grindstone with what is most likely to work.
I basically with. I think the "more work" part keeps getting lost in the shuffle.
But it's higher education's damn job to figure out how to teach, not just how to hold group discussions after the fact. I don't get the impression anyone is doing that job at this school.
It is only very recently that there were any expectations that university lecturers be good at teaching. A century ago, they were expected to be fairly current and interesting. Lectures were necessary because good quality textbooks often did not exist, and even more rarely were current ones available. And if they existed anywhere, you might have to learn German or French to read them -- that would be your problem.
If you did not study hard as possible before lectures, you might be lost. Regardless you and your study buddies would be expected to scramble to cobble together useful knowledge from your notes. The professor was probably happy to answer intelligent questions, or could be paid to tutor you if the quality of your questions did not recommend you as a possible future collaborator.
In a sense, this kind of non-lecturing is a throwback to the old style that the likes of Alan Turing went through. Since there are good textbooks and online material in most topics, it is no longer necessary to lecture in the classic style, and the professor is available to coach the students towards the kind of discussions that students once did amongst themselves.
There is a small subset of superteachers who are heads and shoulders above the rest. And the main difference is that they are tireless in trying new things to further improve their teaching methods.
Guess who volunteers for the small scale research into new teaching methods?
Motivate these superteachers and give them sticks and dirt and frayed string, and they will figure out a way to make an effective lesson. Give them weird new teaching materials, and they will figure out a way to make an effective lesson.
Educational research is expensive. The new ideas may be sound. The new ideas may work in the hands of superteachers. But how to implement these new ideas on a larger scale with non-superteachers is an entirely separate problem from the original small scale research, one that little grant money is given for relative to the daunting size of the problem.
While fundraising by college deans and presidents, scrambling for grants by professors, and lubricating alum donations with feel good events like sports becomes more and more and more important, of course they will figure out how to cut corners on the little stuff that does not keep the money flowing in.
It is not malice. It is not stupidity. It is not even greed.
It is all about having a business plan that actually works.
We expect these giant educational institutions to be run like smart businesses and then complain when they are run like smart businesses. I am not saying that you or I should like this trend, only we should recognize why it is happening.
The cost of living is so high in Silicon Valley, it is too expensive to start a company there. However you can start a company in the newer business incubators say in Upstate NY, or in a rural town trying to attract technology firms.
No. It is too expensive for startups with certain kinds of business plans. To varying degrees, that has been true since the 90s, when there were some people here on/. declaring that $350k houses were pricing everyone out, and surely everyone with a family would rush to Nowhere, USA for the higher standard of living.
Not every business benefits from being in or near the Valley. That has always been true.
One important advantage of the Valley is it is possible to hire up (relatively) quickly, and get experienced talent for key positions. That just will not happen if you set yourself up in a cornfield, and rely on coaxing graduates from University of Illinois at Urbana-Champaign.
Slow organic growth by hiring a fresh grad or three can make for a fine business. But VCs are not interested in you, and you would be foolish to even ask for their money.
I would strongly disagree. Without a precise weight and assessment of metabolism and kidney function and salt & water intake of the individual, dosages are all ballpark guesses. 5% might matter in a one in a million situation, but the reality is we are prescribed dosages that are usually only + or - 50% from optimal all the time.
Slashdot Mirror
I said 'many people don't consider problem-solving to be a sign of intelligence'. I did not say 'most'. The overwhelming majority do, and llike BMI, for most people, in most contexts, it's mostly accurate for what most people consider to be intelligence.
I think your analogy works very well. BMI is a fine overall statistic for plunking onto a chart in a power point presentation about how we might want to allocate more money to health education due to increasing teen health problems. But it easily fails when used for individuals, where there are many other more useful measures on hand. We should endeavor to not be like the dumbass doctor who looked at my young son's BMI number and did not actually look at him before pronouncing he should lose some weight. (My son is having a little difficulty learning to swim because he naturally sinks below the surface -- which means it is impossible for his body fat level to be unreasonable. A doctor astute enough to actually look at her patient could have easily figured that out.)
I agree. Most opinions about how terrible robots will forever be at driving vehicles vastly overestimate how competent hominids really are when behind the wheel in tricky situations, too. There are a lot of tells about potential problems that a computer could pick up consistently and then choose to slow down consistently, thereby both mitigating both the future potential accident as well as buying time to assess the situation correctly. Hominids routinely fail to slow down in ambiguous situations, in the vain hope that two eyeballs will figure it out in time if things go south.
Further, everyone knows what rampant inflation looks like, and why it leads to war and revolution. Nobody cared when it was runaway deflation, because holding a coin for two days doubled your wealth.
I'm pretty sure that we have no real idea what business looks like with deflationary currency. We can make educated assumptions. One thing is for sure, savings would be more beneficial than spending, and credit would largely be rare, rather than common.
Your "pretty sure" is simply wrong. In roughly the same historical period we saw an advanced economy get smacked by hyperinflation and fall into chaos then war, the USA itself was getting hammered by deflation. It was a wonderful time when American shoe factories laid off workers to starve in the streets while barefoot farmers were kicked off their farms, all to serve the idolatry of gold.
Gold fetishists have brainwashed certain people into believing that inflation is super scary while deflationary contractions are purely a theoretical problem. That is because they lie and some sheep are always eager to believe them.
Miners not prioritize transactions without a fee attached. The days when bitcoin transfers are free are years behind us.
There is another problem regarding bitcoin mining: the majority is in China where it is clear as mud if they are actually businesses that could survive under normal economic conditions. Even if someone gave me the most up to date asics in a new server, I could not pay even a quarter of my electrical bill running the thing. What are the costs that Chinese miners are paying? Or are they just stealing electricity by the consent of silent partners who are politically connected? Even if this extreme is not the case, China's electrical power costs are trending up, as they move away from coal fired plants and the demand for electricity continues to grow.
This all adds up to a hidden subsidy of bitcoin. What happens if the subsidy is removed? Transaction costs go up a lot, which will depress bitcoins usage by non-speculators.
I once had an identity theft incident, nearly 25 years ago, where a couple credit cards were taken over (mailing address changed, and new copies of credit cards shipped) and a few new credit card accounts opened. I caught the problem early enough that the damage was minor, more nuisance than financial.
But a little digging and simple deduction led me to this conclusion: someone gained access to my full name, mailing address, SSI, mother's maiden name, and multiple open credit card account numbers. Now where in the universe can this entire array of data be found in one place? My bank? Nope. University? Nope. Employer? Nope. Any one credit card company? Nope.
You guessed it! Credit reporting agencies.
Now it is theoretically possible that an energetic fraudster could gain a few tidbits and build from there. But the dirty little secret sitting in plain sight is the financial institutions themselves, particularly anyone and everyone who deals with credit card data, are the most likely places to leak. That has been true since forever, and it has not changed and will not change, until the laws are very different.
From my POV, most credit protection services are really a sick joke: "As your personal bank or personal credit card company, we would be happy to collect a monthly fee to mitigate the risks you bear due to our incompetence."
Well said.
To some degree, I try to keep an open mind to engineering that seems to be the right tool for the problem on hand. But I do believe that this sloppy kind of engineering that is not much more than taking dirty data and making it somewhat less dirty will prove very easy to replace. Some day there will be an Amazon service where one business data analyst with an AI assistant can try out a hundred different correlation models in an afternoon, and deploy it out to a 1024 AWS cluster on a 1AM crontrigger while the information worker is having a fourth drink at the strip joint.
I recently interviewed at a couple of the new fangled big data marketing startups that correlate piles of stuff to help target ads better, and they were continuously deploying up the wazoo. In fact, they had something like zero people doing traditional QA. It was not totally insane at all. But they did have a blaze attitude about deployments -- if stuff don't work in production they just roll back, and not worry about customer input data being dropped on the floor. Heck, they did not worry much about data losses even when things were working "correctly". To some people big data means no data really matters much, I guess.
That is a fine way to make money when these things are new and expectations are low, and there are easy ways to pick off a tiny slice of that googlenormous flow of ad money to achieve that IPO bliss. I could see AI helping out here.
But continuous delivery can easily suck for code that actually has to work so that people do not get fired or go bankrupt or even die. Continuous AI buzzwordology is not ready for this world.
It does not necessarily matter that the data is dirty. It only matters that globs of money can be spent a tiny bit more efficiently than the crappy way it was spent 20 years ago, to make this new kind of advertising exciting...to people who care about advertising.
For example, big automakers spent a few hundred million a year for advertising. They believe this kind of budget is in the right ballpark based on decades of experience. They are not naive. They do understand that individual advertising efforts within this big budget are flops, but what is going to be a flop is only clear in hindsight.
So we are literally talking about companies like GM looking at expanding their $350 million ad budget from last year to $360 million this year, and wondering whether to increase their "online" spending from $80 million to $90 million. So the data may be dirty but it could still be money better spent than a few more glossy magazine ads.
This. It is a good practice to send out an email after meetings to say "This is what I walked away with as our mutual understanding." Not so much because of illegal shenanigans, but because misunderstandings are common, even among honest and competent people. The honest and competent people appreciate your effort to keep you all on the same page, most especially when things are confusing. The dishonest will become tetchy.
You are being selective in your reading. Making the point that women tend to be neurotic was stupid, and a legitimate reason discontinue his employment as it erodes confidence in his ability to perform his job duties appropriately. Google does not have to prove "cause" here, merely an adequate reason that reasonably sustains a loss of confidence by his employer. That he may have made 99 other non-stupid arguments that are protected by CA law does not give him a legal blank check to sneak in a stupid (unprotected) argument that harms the business.
The problem with being a rambling fool is it only take one single poorly chosen sentence to annihilate his argument that Google did him wrong in the eyes of the court.
You are not connecting your own dots there. The property is a physical piece of land. If one single person can testify they crossed that land for years to get to the public beach without asking permission, that is sufficient grounds for a judge recognizing an easement. The particular usage of the land by the previous owners really does not matter. "Oh, we ran a business so we ignored the guy with the surfboard under his arm walking through our parking lot" is not a counterargument, but additional evidence that the easement exists. The specific usage of land in no way automatically obliterate the customs recognized under English Common Law.
Mod the parent up. It does not really matter what the title says. Reasonable evidence that there was ongoing access to the publicly owned beach is all that is necessary to prove the title was flawed.
Easements based on historical usage have been a thing in English Common Law for several centuries. In CA, like most US states, English Common Law precedents apply unless explicitly changed by legislative action.
Failure of the property owner to understand the law of the land is not a taking under the Constitution.
I think the argument that Google is making certain purported objectives harder has merit, which is not to say that I necessarily agree with that argument, only that I do respect it.
That Damore was _fired_, not for making overtly controversial remarks but for saying "Hey, maybe we should implement policies that have some scientific validity behind them" shows that discourse on the subject has reached Johnathan Swift big endian/small endian levels of ridiculousness.
How would you know? Damore made both controversial remarks and a call for better kinds of discourse. Both arguments happen to have been made in a very sloppy way. It is not obvious (to me) what Google's precise reasoning was here. They do not have to tell us, unless forced to under the bright lights of a courtroom.
Should you disagree, give quotations of the sections of his memo that say explicitly that he is against diversity and hostile to women.
As for hostile to women, we have his non-argument here:
Hiring practices which can effectively lower the bar for "diversity" candidates by decreasing the false negative rate
Reducing false negatives does not automatically change "the bar" in any particular direction. Speculating certain of your colleagues are inferior quality hires based on tortured (non)logic is arguably hostile.
Damore's logic is very sloppy in a number of places. That is why it gets labelled a rant, fairly or unfairly. Because he apparent joy in blabbering about things poorly connected to his alleged main point seems like a hint.
It basically is an argument against what he perceives as so-called "political correctness" in pursuit of gender equity. He tries to be more clever than that but that's what he's saying. He claims to be a "classical liberal" but it's pretty clear that he is not based on his arguments. He makes population arguments and then assumes these must apply on the individual level.
And that is why I am having zero sympathy for him. Bringing in the discussion of population statistics muddies his own supposedly non-sexist argument. He could have simply stated "For whatever reason, from the larger world, the pipeline of excellent engineers that shows up at our door happens to be skewed. Google should treat these special individuals as individuals." That could have cut out a third of the piece, and gotten down quickly to specific practices he wants to question.
Even when he is trying to make his point, he adds junk like...
Hiring practices which can effectively lower the bar for "diversity" candidates by decreasing the false negative rate
Reducing the false negatives is somehow bad because? Amending recruiting practices for some candidates might be "unfair" by some personal ethical definition of the concept, but it is clear as mud whether what he describes actually lowers the bar. Heck, it might even raise both the average quality and number of female hires, for all we know.
Nearly all of our wasted time comes from trying to navigate a context switch.
Even if true, it may not matter in way anyone should care. Because your employer cares about your team's overall productivity, not your productivity. Good software companies in the 21st century are not infatuated with engineers churning out huge piles of code, but with a merely good amount of code productivity that meets the real needs.
Code that is built wrong, in either the wrong way or the towards the wrong goal, often has near zero or even negative value, once you factor in the opportunity costs.
If you are working on a problem that is genuinely without ambiguous requirements, and if all the people who dependencies on your work know exactly what to expect, and if you really are completely competent to get it done without input from anyone else, if all that is true, then, yes, distracting you is very expensive.
But when there are lots of ambiguities, then many of those "distractions" are useful events increasing other people's productivity. Certainly it is possible for that to get way out of hand, but it is not automatically true that open office spaces will cause such.
80% of your education from primary school to a master's degree could have been self-taught from books, once you mastered the basics of reading. So, your point makes sense, but it is not exactly a surprise that sophisticated computer programs can be a very effective replacement for books.
Getting back to your main point, what you suggest can be done, but it does run into two significant bumps in the road: (1a) It is much more expensive to do this well than is generally realized, and (1b) the world is already flooded with mediocre teaching software with more on its way, which is pushing back the necessary effort to do these things right, and (2) for all it is quite possible to learn material from computers, it is also very much easier to cheat. For example: Khan Academy is an excellent resource, but it has failed to gain acceptance as a "real" school that can certify students in the manner employers are used to (which is actually a pretty low bar).
Active learning is actually a harder form of learning in that in requires more work from both student and teacher.
That is why the mediocre are so certain that active learning does not work -- they have never seen it really tried because their only experience as a student suffers the limitation of being who they are.
Furthermore, poorly implemented active learning can easily be worse than the traditional lecture. The smart ones do the real work, while half their teammates snooze or hold text message conversations on their phone, and then copy the work of their betters.
Part of the problem is active learning is touted as a "replacement" for lectures. It is really an entirely different way to organize teaching, where the disappearance of lectures is an inevitable but incidental result. The point is not to get rid of lectures, but to really put the nose to the grindstone with what is most likely to work.
I basically with. I think the "more work" part keeps getting lost in the shuffle.
But it's higher education's damn job to figure out how to teach, not just how to hold group discussions after the fact. I don't get the impression anyone is doing that job at this school.
It is only very recently that there were any expectations that university lecturers be good at teaching. A century ago, they were expected to be fairly current and interesting. Lectures were necessary because good quality textbooks often did not exist, and even more rarely were current ones available. And if they existed anywhere, you might have to learn German or French to read them -- that would be your problem.
If you did not study hard as possible before lectures, you might be lost. Regardless you and your study buddies would be expected to scramble to cobble together useful knowledge from your notes. The professor was probably happy to answer intelligent questions, or could be paid to tutor you if the quality of your questions did not recommend you as a possible future collaborator.
In a sense, this kind of non-lecturing is a throwback to the old style that the likes of Alan Turing went through. Since there are good textbooks and online material in most topics, it is no longer necessary to lecture in the classic style, and the professor is available to coach the students towards the kind of discussions that students once did amongst themselves.
There is a small subset of superteachers who are heads and shoulders above the rest. And the main difference is that they are tireless in trying new things to further improve their teaching methods.
Guess who volunteers for the small scale research into new teaching methods?
Motivate these superteachers and give them sticks and dirt and frayed string, and they will figure out a way to make an effective lesson. Give them weird new teaching materials, and they will figure out a way to make an effective lesson.
Educational research is expensive. The new ideas may be sound. The new ideas may work in the hands of superteachers. But how to implement these new ideas on a larger scale with non-superteachers is an entirely separate problem from the original small scale research, one that little grant money is given for relative to the daunting size of the problem.
While fundraising by college deans and presidents, scrambling for grants by professors, and lubricating alum donations with feel good events like sports becomes more and more and more important, of course they will figure out how to cut corners on the little stuff that does not keep the money flowing in.
It is not malice. It is not stupidity. It is not even greed.
It is all about having a business plan that actually works.
We expect these giant educational institutions to be run like smart businesses and then complain when they are run like smart businesses. I am not saying that you or I should like this trend, only we should recognize why it is happening.
The cost of living is so high in Silicon Valley, it is too expensive to start a company there. However you can start a company in the newer business incubators say in Upstate NY, or in a rural town trying to attract technology firms.
No. It is too expensive for startups with certain kinds of business plans. To varying degrees, that has been true since the 90s, when there were some people here on /. declaring that $350k houses were pricing everyone out, and surely everyone with a family would rush to Nowhere, USA for the higher standard of living.
Not every business benefits from being in or near the Valley. That has always been true.
One important advantage of the Valley is it is possible to hire up (relatively) quickly, and get experienced talent for key positions. That just will not happen if you set yourself up in a cornfield, and rely on coaxing graduates from University of Illinois at Urbana-Champaign.
Slow organic growth by hiring a fresh grad or three can make for a fine business. But VCs are not interested in you, and you would be foolish to even ask for their money.
I would strongly disagree. Without a precise weight and assessment of metabolism and kidney function and salt & water intake of the individual, dosages are all ballpark guesses. 5% might matter in a one in a million situation, but the reality is we are prescribed dosages that are usually only + or - 50% from optimal all the time.