Because of one of the critical elements that actually make capitalism (kind of) work: competition.
If you have a cure and don't offer it, I will, and then you're screwed.
This tends to happen a lot in pharma because the actual drugs are discovered in publicly funded university labs, are taken through initial trials by spin offs, startups, and small biotech firms, then are bought up for terminal approval and marketing by the big guys. But anyone can come along and take that university research, tweak it a bit, and go through the process with a slightly different drug. That's why when one company comes out with some new class of drug, all the others have a very similar product shortly after.
Big pharma has been working hard to become an industry where there isn't any real competition so they don't have to worry about stuff like that. Perhaps Goldman-Sachs thinks they've arrived.
That's not really true. Most of the antidepressants that are properly prescribed do work, and have minimal and well known side effects. The problem is, big pharma has been coercing physicians into using antidepressants in ways that were never intended or tested.
I didn't list deficiency diseases because the OP probably considers those "treated" rather than "cured": you have to keep taking vitamin C to keep the scurvy at bay. If you have pellagra and someone gives you a supply of niacin, you might well consider it a cure though.
"Cure" is a slippy concept, and you can probably craft your definition to pedantically eliminate pretty much anything.
That depends what you consider a cure. Many types of cancer are curable. A considerable number of birth defects are curable with surgery. There are now some stem cell treatments for specific kinds of blindness.
It's probably even more dramatic than that. With all the profit packed into a couple of years, and with zero competitors during those two years, they didn't even have to advertise. Advertising is *the* major cost at a pharma company. Plus now they get to immediately reinvest the money that they would have earned at years 3-20.
The Goldman-Sachs analysis seems incredibly superficial.
The big players don't really find new drugs at all. Most of them have cut or eliminated their basic research units.
New drugs are usually discovered and isolated at public universities, then the initial stages of commercialization are done at a startup or various small (or less small, but more diversified) companies, then one of the major pharma companies will buy the rights and take it through the last stages of certification. Or maybe shelve it until a later time.
The 970 drives my 4 k screen reasonably well, but a 1070 would do better. I wouldn't make that upgrade for retail price, but for $200, sure.
Also, I do medical imaging and loading a full volume into memory and then manipulating it works in 4 GB but would work better in 8. It's handy to have a local machine with a decent card so you don't have to debug on the cluster all the time.
"Not long-term, but if you run a company and smell easy money, you adapt and chase it. That's just good business, if only for being able to survive... or beat your competition to the punch."
Chip design isn't easy money. Some people smelled easy money and whipped off some simple ASICs. They probably made back some fraction of a percent of what Nvidia pulls in.
If you're a smart business you try to supply some long term, sustainable demand. Nvidia has publicly said that they don't think cryptocurrency mining is that. They're enjoying lots of GPU demand, but they don't think it's worth alienating their core customers to serve. With Etherium moving to proof-of-stake voting, that's looking like it was probably a smart decision.
I think you're mistaken about miners demanding the best equipment too. Etherium mining works just as well on multiple slower GPUs as it does on one fast one (usually better, because of the way the algorithm is designed). So a smart miner will choose a good power/performance ratio. Games, on the other hand, generally benefit most from a single very powerful GPU, so gamers drive the high end.
Communication skills are pretty important for real computer science. Poor communication could certainly screw you up here, but so could a poor understanding of whatever algorithm you're supposed to describe. I've taught quite a few students, and many can rattle off an implementation of an algorithm without any problem but actually have no idea how it works.
To be clear, I said in my original post "if you're not a code monkey." If you're supposed to be learning to implement fairly straightforward requirements with known algorithms then this test is not so relevant to you, but you should also be in a software engineering or vocational coding program. If you're in computer science you need to demonstrate a deeper understanding than just being able to implement and use standard algorithms.
Nvidia at least sees the cryptocurrency thing as a flash in the pan. Miners have demanded a lot of high end GPUs all of a sudden, but they might not express that demand over decades like gamers do. They're certainly not driving GPU R&D.
Ethereum's proof of work is based on directed graphs and apparently it is a memory and bandwidth hog so it works pretty well on a GPU but you can't really just make a single ASIC that can tear through hashes like you can with bitcoin. You can certainly make an ASIC that does the work, but you also have to have lots of memory interfaced to that ASIC through a high speed interconnect. So you're really talking about designing a custom computer rather than just an ASIC with a simple interface.
Thanks for the GPU tip. Looks like I'll be upgrading my 970 soon.
That would be interesting. Mr. Miyagi would suggest that you'd learn better by being forced to go look it up repeatedly. There is some research that suggests you do learn best when things are difficult, but not impossibly difficult.
I was just discussing with the always insightful Angel'o'sphere the possibility that the proliferation of IDEs with autocompletion is at least partially responsible for some of the poor interfaces in libraries and languages. Java apparently has at least three Date classes, all in different places!
I think your examples nicely make the point. I don't know if the commonality of IDEs is responsible for the plethora of Date classes in Java, but I can't imagine it helps.
"If you DON'T know the language, having autocomplete make suggestions can help as you guess your way through it."
Agreed. Which is a great reason not to use them. It's like looking at your fingers when you type. It helps when you're a crappy typist, and when you're learning you can fool yourself into thinking it makes you better, but you'll never learn properly if you do.
Well, there's a good reason. Perhaps the use of IDEs encourages poor library design. WHY are there so many date classes you can't remember where they are?
It's kind of like effective theories in physics. What is magnetism? Magnetism is the force acting perpendicular to the direction of motion of a charged particle.
I suggested to someone the other day that the ideal Python IDE was a text editor and a terminal. Choose an editor with syntax colouring if you must. The look I got. But what about the autocompletion??
I've never actually met an IDE where the autocompletion didn't piss me off.
The student gets a paper notebook and a pencil, and is paired with a partner who knows nothing about programming. The student has to explain an algorithm to the partner. At the end of the exam the partner has to independently write down their understanding of the algorithm, with diagrams.
The best test of whether you understand something is how well you can teach someone else. And unless you're a code monkey, this is probably what you're going to spend half your life doing anyway, whether it's explaining to PHBs or grant review committees. My sister makes custom leather book coverings. She's going to make me one for the cheap notebook I keep in my bag at all times for this exact task.
See, I was with you up to the example. Words do have meanings. Strategy is a word, and it has a meaning. It means "a plan of action or policy designed to achieve a major overall aim."
It doesn't mean deciding who you are (navel gazing?), or whatever the hell the article thinks it means (as far as I can tell the author doesn't actually settle on a definition).
The quote in the summary alone was a good WTF. First sentence, yup, can believe that. Second sentence, business types find words scary? Okay. From there on... WTF?
I imagine it's kind of like having a spell checker. The spell checker isn't perfect, (e.g. accepting waist instead of waste) so you have to be paying some attention when you write, but it will catch and prevent lots of your errors. The two of you working together can generally do a better job more easily than you a person would alone.
Assuming you're talking about the 2011 attack, the perpetrator was sentenced to 21 years, at the end of which his sentence could be extended by five years at a time, indefinitely, until he is determined to no longer be dangerous.
What do you think the purpose of imprisonment is? Research shows that beyond a point, harsher sentences (including the death penalty) don't really serve as much of a deterrent. The "mass murderer only got X years" expresses a desire for vengeance, which is natural, but not particularly useful: it might make some people feel better for a short period of time, but doesn't really contribute much else, and has some serious downsides. Justice systems focused on harm prevention and rehabilitation have proven to be the most effective, and in such a system the sentence the 2011 killer got makes a lot of sense.
Norway has a low crime rate, consistently ranks among the best places in the world to live, has a relatively small rate of incarceration and massacres like the one you mentioned are very rare. It seems they're doing something right.
Slashdot Mirror
Because of one of the critical elements that actually make capitalism (kind of) work: competition.
If you have a cure and don't offer it, I will, and then you're screwed.
This tends to happen a lot in pharma because the actual drugs are discovered in publicly funded university labs, are taken through initial trials by spin offs, startups, and small biotech firms, then are bought up for terminal approval and marketing by the big guys. But anyone can come along and take that university research, tweak it a bit, and go through the process with a slightly different drug. That's why when one company comes out with some new class of drug, all the others have a very similar product shortly after.
Big pharma has been working hard to become an industry where there isn't any real competition so they don't have to worry about stuff like that. Perhaps Goldman-Sachs thinks they've arrived.
That's not really true. Most of the antidepressants that are properly prescribed do work, and have minimal and well known side effects. The problem is, big pharma has been coercing physicians into using antidepressants in ways that were never intended or tested.
I didn't list deficiency diseases because the OP probably considers those "treated" rather than "cured": you have to keep taking vitamin C to keep the scurvy at bay. If you have pellagra and someone gives you a supply of niacin, you might well consider it a cure though.
"Cure" is a slippy concept, and you can probably craft your definition to pedantically eliminate pretty much anything.
That depends what you consider a cure. Many types of cancer are curable. A considerable number of birth defects are curable with surgery. There are now some stem cell treatments for specific kinds of blindness.
It's probably even more dramatic than that. With all the profit packed into a couple of years, and with zero competitors during those two years, they didn't even have to advertise. Advertising is *the* major cost at a pharma company. Plus now they get to immediately reinvest the money that they would have earned at years 3-20.
The Goldman-Sachs analysis seems incredibly superficial.
The big players don't really find new drugs at all. Most of them have cut or eliminated their basic research units.
New drugs are usually discovered and isolated at public universities, then the initial stages of commercialization are done at a startup or various small (or less small, but more diversified) companies, then one of the major pharma companies will buy the rights and take it through the last stages of certification. Or maybe shelve it until a later time.
The 970 drives my 4 k screen reasonably well, but a 1070 would do better. I wouldn't make that upgrade for retail price, but for $200, sure.
Also, I do medical imaging and loading a full volume into memory and then manipulating it works in 4 GB but would work better in 8. It's handy to have a local machine with a decent card so you don't have to debug on the cluster all the time.
"Not long-term, but if you run a company and smell easy money, you adapt and chase it. That's just good business, if only for being able to survive... or beat your competition to the punch."
Chip design isn't easy money. Some people smelled easy money and whipped off some simple ASICs. They probably made back some fraction of a percent of what Nvidia pulls in.
If you're a smart business you try to supply some long term, sustainable demand. Nvidia has publicly said that they don't think cryptocurrency mining is that. They're enjoying lots of GPU demand, but they don't think it's worth alienating their core customers to serve. With Etherium moving to proof-of-stake voting, that's looking like it was probably a smart decision.
I think you're mistaken about miners demanding the best equipment too. Etherium mining works just as well on multiple slower GPUs as it does on one fast one (usually better, because of the way the algorithm is designed). So a smart miner will choose a good power/performance ratio. Games, on the other hand, generally benefit most from a single very powerful GPU, so gamers drive the high end.
Communication skills are pretty important for real computer science. Poor communication could certainly screw you up here, but so could a poor understanding of whatever algorithm you're supposed to describe. I've taught quite a few students, and many can rattle off an implementation of an algorithm without any problem but actually have no idea how it works.
To be clear, I said in my original post "if you're not a code monkey." If you're supposed to be learning to implement fairly straightforward requirements with known algorithms then this test is not so relevant to you, but you should also be in a software engineering or vocational coding program. If you're in computer science you need to demonstrate a deeper understanding than just being able to implement and use standard algorithms.
Nvidia at least sees the cryptocurrency thing as a flash in the pan. Miners have demanded a lot of high end GPUs all of a sudden, but they might not express that demand over decades like gamers do. They're certainly not driving GPU R&D.
Ethereum's proof of work is based on directed graphs and apparently it is a memory and bandwidth hog so it works pretty well on a GPU but you can't really just make a single ASIC that can tear through hashes like you can with bitcoin. You can certainly make an ASIC that does the work, but you also have to have lots of memory interfaced to that ASIC through a high speed interconnect. So you're really talking about designing a custom computer rather than just an ASIC with a simple interface.
Thanks for the GPU tip. Looks like I'll be upgrading my 970 soon.
That would be interesting. Mr. Miyagi would suggest that you'd learn better by being forced to go look it up repeatedly. There is some research that suggests you do learn best when things are difficult, but not impossibly difficult.
I was just discussing with the always insightful Angel'o'sphere the possibility that the proliferation of IDEs with autocompletion is at least partially responsible for some of the poor interfaces in libraries and languages. Java apparently has at least three Date classes, all in different places!
I think your examples nicely make the point. I don't know if the commonality of IDEs is responsible for the plethora of Date classes in Java, but I can't imagine it helps.
"If you DON'T know the language, having autocomplete make suggestions can help as you guess your way through it."
Agreed. Which is a great reason not to use them. It's like looking at your fingers when you type. It helps when you're a crappy typist, and when you're learning you can fool yourself into thinking it makes you better, but you'll never learn properly if you do.
Well, there's a good reason. Perhaps the use of IDEs encourages poor library design. WHY are there so many date classes you can't remember where they are?
It's kind of like effective theories in physics. What is magnetism? Magnetism is the force acting perpendicular to the direction of motion of a charged particle.
I suggested to someone the other day that the ideal Python IDE was a text editor and a terminal. Choose an editor with syntax colouring if you must. The look I got. But what about the autocompletion??
I've never actually met an IDE where the autocompletion didn't piss me off.
The ultimate CS test:
The student gets a paper notebook and a pencil, and is paired with a partner who knows nothing about programming. The student has to explain an algorithm to the partner. At the end of the exam the partner has to independently write down their understanding of the algorithm, with diagrams.
The best test of whether you understand something is how well you can teach someone else. And unless you're a code monkey, this is probably what you're going to spend half your life doing anyway, whether it's explaining to PHBs or grant review committees. My sister makes custom leather book coverings. She's going to make me one for the cheap notebook I keep in my bag at all times for this exact task.
See, I was with you up to the example. Words do have meanings. Strategy is a word, and it has a meaning. It means "a plan of action or policy designed to achieve a major overall aim."
It doesn't mean deciding who you are (navel gazing?), or whatever the hell the article thinks it means (as far as I can tell the author doesn't actually settle on a definition).
Because they used all the sports terms already?
The quote in the summary alone was a good WTF.
First sentence, yup, can believe that. Second sentence, business types find words scary? Okay. From there on... WTF?
"which was supposed to compare before AP vs after AP but 2/3 of the cars in the study didn't have any 'before AP' miles."
That's not really a problem if you know something beyond stats 100. Auto steer plus auto braking... starting to sound like auto pilot hey?
I imagine it's kind of like having a spell checker. The spell checker isn't perfect, (e.g. accepting waist instead of waste) so you have to be paying some attention when you write, but it will catch and prevent lots of your errors. The two of you working together can generally do a better job more easily than you a person would alone.
Homer thinks you've been damaged by poor educational opportunities. And no, that's not Homer Simpson.
Assuming you're talking about the 2011 attack, the perpetrator was sentenced to 21 years, at the end of which his sentence could be extended by five years at a time, indefinitely, until he is determined to no longer be dangerous.
What do you think the purpose of imprisonment is? Research shows that beyond a point, harsher sentences (including the death penalty) don't really serve as much of a deterrent. The "mass murderer only got X years" expresses a desire for vengeance, which is natural, but not particularly useful: it might make some people feel better for a short period of time, but doesn't really contribute much else, and has some serious downsides. Justice systems focused on harm prevention and rehabilitation have proven to be the most effective, and in such a system the sentence the 2011 killer got makes a lot of sense.
Norway has a low crime rate, consistently ranks among the best places in the world to live, has a relatively small rate of incarceration and massacres like the one you mentioned are very rare. It seems they're doing something right.