What has happened to Slashdot? I saw the post title and expected to see a big dose of the "Andy Groves Fallacy" in the comments section. Instead I found multiple well reasoned critiques of in silico drug discovery. This place has changed.
Which do you think is more likely? Distribution of profits to unemployed people, or distribution of profits to wealthy C-level executives and investors?
For the most part in this case: Neither. As robots/AI decrease the cost of a good or service, the good or service sells for less.
Actually what Springer and Elsevier (the publishers of the majority of science journals) could do is explain that in the case of papers coming from offending organizations, a translator would have to be hired and asked to directly contact and vet the suggested peer reviewers. This would be at the expense of the institution/authors.
If you would like to use this as evidence for a sweeping and universal hypothesis, shouldn't you at least have a brief look at the retraction notice that the story is based on?
From a first glance, all of the papers come from China (click on affiliations to see which institutes they work at). China and India have been notorious as paper mills for decades. So why are you extrapolating this to work from countries where peer reviewers are,a, very easy for editors to directly find and contact, and,b, can usually be communicated with directly without the need for a translator?
Because this type of fraud is rampant in China and India (look at the author lists in the retraction note: all at Chinese institutions) and much more rare in the US and Europe.
"Bin Picking" requires recognizing which objects are which, and what their orientation ("pose") is. Then plan a way to move to collect them.
For big retail operations, wouldn't requiring the wholesalers to package their items so that they are easy to pick and place be easier? A QR code, standardized grab points, etc. It won't work for all operations (like reshelving clothes after they have been tried on), but it would be easy for anything that currently comes in a blister pack.
In all seriousness, why are clinical trials so expensive? Volunteers, a doctor with a sack of pills, a lab technician and a statistician are all you need, no?
In essence, but things get complicated pretty quickly. You don't have to pay the volunteers (though that happens frequently enough in phase I), but you do have to recruit them. A lot of them: phase III trials can have thousands of patients. That means lots of money to the medical practices that are treating people with that condition. Then you have to screen them to make certain that they fall into the groups you want to look at, which means lots of tests. a "sack of pills" can mean a weekly injection of a monoclonal antibody (which are still pretty expensive to manufacture) at a doctor's office. Monitoring during a trial can mean lots of tests, including CT scans and other imaging. And clinical trials can be looong. For things like cardiovascular conditions or Alzheimers you have to run them for years before the results start to show. Then there's the paperwork. If there's ever a situation where, yeah, it IS important to make certain everything was documented correctly, by the right people, and then stored in a secure manner, it's a clinical trial.
Bristol-Meyers Squib has a 75% margin on their drugs. And almost 30% return on equity.
They like to blame R&D but one Summer I worked at one of their research labs. It was a very very nice place. Parts could have been from a country club. The head of the place helicoptered in from NY every morning - which is all considered R&D "costs". The cafeteria food was 5-star but cost as much as a McDonald's meal.
The only sucky part was the animal section.
I miss that place.
Hate to break it to you but you are talking about the CHEAP end of R&D costs: preclinical. The vast majority of spending is in phase II and III clinical trials.
While I think these are likely a rip-off, you do also need to consider R&D costs and for rarer disorders the cost spread over a smaller number of dosages.
Actually you don't. The price is set by finding the maximum value for (price per dose) x (doses sold per month). R&D is a sunk cost at that point. All else being equal, it doesn't matter whether they spent $100M or $10B, the price will still be the same.
The various drug companies spend 3 1/2 x as much on advertising and marketing as research, with MOST basic research paid for by governments and sold for pennies or given away for free
Welfare starts at the top
As a percentage of revenue Pharma spends more on R&D than almost any other industry.
And furthermore, not all research is reproducible.
Quite true, but what non-reproducible research is necessary for writing regulations? Even when the data is incidents and accidents that no one wants to reproduce, the research itself is reproducible by redoing the analysis.
Whether or not research proves it to be the "best" diet for most people, the Mediterranean diet has been recommended and researched for ~50 years now. So not exactly a craze.
Like me and many users of libraries, Gildas marks the place from which he takes a book and carefully reshelves it when he is done, saving the library staff reshelving work. The algorithm missed his book and now it is shredded or moldering in a distant storage facility.
Most academic libraries put up pretty explicit notices saying why patrons shouldn't reshelve the books.
A branch library justifies its expenditure of public funding by being useful to it's community. To a first approximation, If more people check out new mass market fluff than old mass market fluff then recycling older titles is useful. Branch libraries are just that: branches. Almost all have access to state or regional interlibrary loan for rare titles. If the goal is to have older titles onsite so that people browsing will come across them: shuffling the rarer books between branch libraries every so often would be better than trying to have a "complete" collection at each site. Regular users of the library would have new-to-them titles to browse every few months.
Or it could be the mouse models of Alzheimers that are based on overexpression of amyloid precursor protein don't really reflect the pathology found in humans.
http://www.alzforum.org/news/r...
WTF are you on? How is not doing and/or providing proper maintenance and insurance have anything to do with supply and demand?
But it is everything to do with a free market.
Also the medalion/token system for Taxi model isn't about raising prices. It's about traffic congestion
Please divide the lobbying dollars (and outright cash bribes) given for the purpose of swaying taxi medallion numbers into two pots: one for people interested in traffic congestion, the other for people interested in taxi profitability. Which pot overflows?
Road deaths have been falling for decades, Since 1972, deaths per capita have halved and deaths per mile quartered. This is not a valid area for extraordinary efforts.
Actually that indicates it absolutely is a valid area for extraordinary efforts: it's an area where increased investments in safety consistently result in... more safety. Traffic deaths are now on the rise. Finding a way to get people to ignore their phones while driving would probably have a good ROI too.
It would probably be better to have the appliances themselves do the load balancing as opposed to the circuits. As long as we are going to end up with an internet of things, have the house stagger when the AC and refrigerator compressors start up and turn off the electric oven/range and water heaters for a few seconds whenever a big motor starts up as well. That way when someone hacks your house they can burn out your batteries, inverter, HVAC, and refrigerator by turning on everything in the house at the same time!
In all, more than 50 gigabits of data were relayed over the past 15 months to-
When talking about data transfer like this and assuming you are not trying to sell me something, is there really a good reason to use bits instead of bytes?
It's prudent to look ahead in time to try to predict dangers and other bad situations, but you also need to keep track of the probabilities.
OT, but that's pretty much my response to people who say they are buying a handgun for self defense when, after considering demographics, the odds of someone in their household using the handgun to commit suicide or spousal homicide are higher than the odds of their being shot by someone outside the home.
Shouldn't we attend to the big dangers first?
Absolutely. For the vast majority of readers of this website, attending to the big dangers first pretty means getting all of your cardiovascular risk factors under control before spending dollar one on a gun for self defense.
Slashdot Mirror
What has happened to Slashdot? I saw the post title and expected to see a big dose of the "Andy Groves Fallacy" in the comments section. Instead I found multiple well reasoned critiques of in silico drug discovery. This place has changed.
Which do you think is more likely? Distribution of profits to unemployed people, or distribution of profits to wealthy C-level executives and investors?
For the most part in this case: Neither. As robots/AI decrease the cost of a good or service, the good or service sells for less.
Unless you leave earth, its still limited (if you leave earth, its virtually unlimited)
Unless you have a magic 500% efficient propulsion system up your sleeve it is safe to say If you leave Earth everything is extremely limited.
Robots are getting pretty good at building walls ...
Actually what Springer and Elsevier (the publishers of the majority of science journals) could do is explain that in the case of papers coming from offending organizations, a translator would have to be hired and asked to directly contact and vet the suggested peer reviewers. This would be at the expense of the institution/authors.
If you would like to use this as evidence for a sweeping and universal hypothesis, shouldn't you at least have a brief look at the retraction notice that the story is based on?
https://link.springer.com/arti...
From a first glance, all of the papers come from China (click on affiliations to see which institutes they work at). China and India have been notorious as paper mills for decades. So why are you extrapolating this to work from countries where peer reviewers are ,a, very easy for editors to directly find and contact, and ,b, can usually be communicated with directly without the need for a translator?
Because this type of fraud is rampant in China and India (look at the author lists in the retraction note: all at Chinese institutions) and much more rare in the US and Europe.
"Bin Picking" requires recognizing which objects are which, and what their orientation ("pose") is. Then plan a way to move to collect them.
For big retail operations, wouldn't requiring the wholesalers to package their items so that they are easy to pick and place be easier? A QR code, standardized grab points, etc. It won't work for all operations (like reshelving clothes after they have been tried on), but it would be easy for anything that currently comes in a blister pack.
In all seriousness, why are clinical trials so expensive? Volunteers, a doctor with a sack of pills, a lab technician and a statistician are all you need, no?
In essence, but things get complicated pretty quickly. You don't have to pay the volunteers (though that happens frequently enough in phase I), but you do have to recruit them. A lot of them: phase III trials can have thousands of patients. That means lots of money to the medical practices that are treating people with that condition. Then you have to screen them to make certain that they fall into the groups you want to look at, which means lots of tests. a "sack of pills" can mean a weekly injection of a monoclonal antibody (which are still pretty expensive to manufacture) at a doctor's office. Monitoring during a trial can mean lots of tests, including CT scans and other imaging. And clinical trials can be looong. For things like cardiovascular conditions or Alzheimers you have to run them for years before the results start to show. Then there's the paperwork. If there's ever a situation where, yeah, it IS important to make certain everything was documented correctly, by the right people, and then stored in a secure manner, it's a clinical trial.
Bristol-Meyers Squib has a 75% margin on their drugs. And almost 30% return on equity.
They like to blame R&D but one Summer I worked at one of their research labs. It was a very very nice place. Parts could have been from a country club. The head of the place helicoptered in from NY every morning - which is all considered R&D "costs". The cafeteria food was 5-star but cost as much as a McDonald's meal.
The only sucky part was the animal section.
I miss that place.
Hate to break it to you but you are talking about the CHEAP end of R&D costs: preclinical. The vast majority of spending is in phase II and III clinical trials.
While I think these are likely a rip-off, you do also need to consider R&D costs and for rarer disorders the cost spread over a smaller number of dosages.
Actually you don't. The price is set by finding the maximum value for (price per dose) x (doses sold per month). R&D is a sunk cost at that point. All else being equal, it doesn't matter whether they spent $100M or $10B, the price will still be the same.
The various drug companies spend 3 1/2 x as much on advertising and marketing as research, with MOST basic research paid for by governments and sold for pennies or given away for free Welfare starts at the top
As a percentage of revenue Pharma spends more on R&D than almost any other industry.
And furthermore, not all research is reproducible.
Quite true, but what non-reproducible research is necessary for writing regulations? Even when the data is incidents and accidents that no one wants to reproduce, the research itself is reproducible by redoing the analysis.
Whoosh.
Whether or not research proves it to be the "best" diet for most people, the Mediterranean diet has been recommended and researched for ~50 years now. So not exactly a craze.
Like me and many users of libraries, Gildas marks the place from which he takes a book and carefully reshelves it when he is done, saving the library staff reshelving work. The algorithm missed his book and now it is shredded or moldering in a distant storage facility.
Most academic libraries put up pretty explicit notices saying why patrons shouldn't reshelve the books.
A branch library justifies its expenditure of public funding by being useful to it's community. To a first approximation, If more people check out new mass market fluff than old mass market fluff then recycling older titles is useful. Branch libraries are just that: branches. Almost all have access to state or regional interlibrary loan for rare titles. If the goal is to have older titles onsite so that people browsing will come across them: shuffling the rarer books between branch libraries every so often would be better than trying to have a "complete" collection at each site. Regular users of the library would have new-to-them titles to browse every few months.
Or it could be the mouse models of Alzheimers that are based on overexpression of amyloid precursor protein don't really reflect the pathology found in humans. http://www.alzforum.org/news/r...
WTF are you on? How is not doing and/or providing proper maintenance and insurance have anything to do with supply and demand?
But it is everything to do with a free market.
Also the medalion/token system for Taxi model isn't about raising prices. It's about traffic congestion
Please divide the lobbying dollars (and outright cash bribes) given for the purpose of swaying taxi medallion numbers into two pots: one for people interested in traffic congestion, the other for people interested in taxi profitability. Which pot overflows?
Road deaths have been falling for decades, Since 1972, deaths per capita have halved and deaths per mile quartered. This is not a valid area for extraordinary efforts.
Actually that indicates it absolutely is a valid area for extraordinary efforts: it's an area where increased investments in safety consistently result in ... more safety. Traffic deaths are now on the rise. Finding a way to get people to ignore their phones while driving would probably have a good ROI too.
if time delay is an issue then "as needed" means continuously - and thus no time delay so long as you have enough stations broadcasting.
It would probably be better to have the appliances themselves do the load balancing as opposed to the circuits. As long as we are going to end up with an internet of things, have the house stagger when the AC and refrigerator compressors start up and turn off the electric oven/range and water heaters for a few seconds whenever a big motor starts up as well. That way when someone hacks your house they can burn out your batteries, inverter, HVAC, and refrigerator by turning on everything in the house at the same time!
In all, more than 50 gigabits of data were relayed over the past 15 months to-
When talking about data transfer like this and assuming you are not trying to sell me something, is there really a good reason to use bits instead of bytes?
Speaking of UBI and incentivizing productivity ... how to handle the "having kids" issue?
It's prudent to look ahead in time to try to predict dangers and other bad situations, but you also need to keep track of the probabilities.
OT, but that's pretty much my response to people who say they are buying a handgun for self defense when, after considering demographics, the odds of someone in their household using the handgun to commit suicide or spousal homicide are higher than the odds of their being shot by someone outside the home.
Shouldn't we attend to the big dangers first?
Absolutely. For the vast majority of readers of this website, attending to the big dangers first pretty means getting all of your cardiovascular risk factors under control before spending dollar one on a gun for self defense.