faster better science

Discussions related to engineering and its applications. From civil and mechanical to aeronautic and robotics, etc.

faster better science

Postby Athena on April 28th, 2017, 12:20 pm 

Is it possible artificial intelligence can greatly improve the speed and accuracy of scientific development?

Science has outgrown the human mind and its limited capacities
https://www.theatlantic.com/science/arch...ed/524136/

EXCERPT: Science is in the midst of a data crisis. Last year, there were more than 1.2 million new papers published in the biomedical sciences alone, bringing the total number of peer-reviewed biomedical papers to over 26 million. However, the average scientist reads only about 250 papers a year. Meanwhile, the quality of the scientific literature has been in decline. Some recent studies found that the majority of biomedical papers were irreproducible.

The twin challenges of too much quantity and too little quality are rooted in the finite neurological capacity of the human mind. Scientists are deriving hypotheses from a smaller and smaller fraction of our collective knowledge and consequently, more and more, asking the wrong questions, or asking ones that have already been answered. Also, human creativity seems to depend increasingly on the stochasticity of previous experiences – particular life events that allow a researcher to notice something others do not. Although chance has always been a factor in scientific discovery, it is currently playing a much larger role than it should.

One promising strategy to overcome the current crisis is to integrate machines and artificial intelligence in the scientific process. Machines have greater memory and higher computational capacity than the human brain. Automation of the scientific process could greatly increase the rate of discovery. It could even begin another scientific revolution. That huge possibility hinges on an equally huge question: can scientific discovery really be automated? I believe it can, using an approach that we have known about for centuries. The answer to this question can be found in the work of Sir Francis Bacon, the 17th-century English philosopher and a key progenitor of modern science....

MORE: https://www.theatlantic.com/science/arch...ed/524136/
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon
vivian maxine liked this post


Re: faster better science

Postby Eclogite on April 28th, 2017, 2:58 pm 

I get a 404 when I try to access the article. That makes serious discussion of the topic impractical.

Edit: This appears to be the article here, from another url.
Eclogite
Forum Moderator
 
Posts: 1388
Joined: 07 Feb 2007
Location: Around and about


Re: faster better science

Postby Forest_Dump on April 28th, 2017, 4:19 pm 

Well, I couldn't access it either. However, from what I did managed to read:

At first I was astonished by the numbers. But then maybe not so much (assuming the numbers are correct). Some brief thoughts included wondering how much of this "publishing" is from the industry of producing scientists? With the publish or perish meme as well as the need to get credentials, publishing anything and everything, partiocularly given that the peer review process is so thankless while time consuming as well as rarely anonymous (who and how do editors choose as reviewers?) I can see how things could start to get out of hand particularly when there can be kinds fo competitions to get students through the process, into labs and institutions, etc. And then there is the demand to get products to market as fast as possible and a few good papers that no one has time or inclination to critique can translate into a lot of money... Bottom line, the scholarly and critical thinking ethos has certainly been hammered over the last few decades.

But still, a certain amount of automation may well assist to some degree. I don't think anyone would argue that computers haven't helped a lot with some things and who knows what the next generation of information technology could assist with. The danger, of course, is the same old one of blindly believing that technology, data processing, etc., can do it all. At some point we can only hope that some solid thinking by actual smart people (as opposed to just those who simply assert they are smart on the basis of mass-produced credentials including a few of those peer-reveewed papers that no one has read) will be keeping an eye on what is really happening with the "faster, better science" and those selling it.
User avatar
Forest_Dump
Resident Member
 
Posts: 8773
Joined: 31 Mar 2005
Location: Great Lakes Region
Athenavivian maxine liked this post


Re: faster better science

Postby vivian maxine on April 28th, 2017, 5:04 pm 

Athena, good to bring something like this to attention for at least consideration. AI is good but just one question about AI. Is it capable of judging? There is a difference between remembering what has already been said and or done and judging where the author might be going astray in his/her thinking. Or, heavens to Betsy, judging the quality of the writing for clarity and comprehension. Just wondering. Ideas?
vivian maxine
Resident Member
 
Posts: 2837
Joined: 01 Aug 2014


Re: faster better science

Postby Athena on May 5th, 2017, 12:42 pm 

Sorry about the link problem. I have tested this one here and now, and it worked for me.

https://www.theatlantic.com/science/arc ... ed/524136/

In another forum the first posted links worked just fine. Strange. I do not think I will go into computer science, because it shuts my brain down.
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon


Re: faster better science

Postby Athena on May 5th, 2017, 1:31 pm 

vivian maxine » April 28th, 2017, 3:04 pm wrote:Athena, good to bring something like this to attention for at least consideration. AI is good but just one question about AI. Is it capable of judging? There is a difference between remembering what has already been said and or done and judging where the author might be going astray in his/her thinking. Or, heavens to Betsy, judging the quality of the writing for clarity and comprehension. Just wondering. Ideas?


Well, my experience in science forums is humans have prejudiced judgments, and they make the computer programs, so maybe the computers will not do any better, but they might do the doing faster?

When it comes to how our computers work and how the internet works, and quantum physics, smile, my brain just shuts down. But my mother was a key puncher. A key puncher sat at a machine that made holes in cards and these cards could be read by computers. We could do the same thing with a knitting needle and cards with holes in them. So I imagine the computer would not have judgment as humans do, but like the old keypunch cards, the computer could categorize the information. Once the information was categorized, some of it would be trashed and some of it would get human attention. Related information would be in the same pile and easier for a human to find.

Today's AI may go beyond what the old keypunch card reading computers did, and that is both exciting and frightening. But the subject is more than this. As this science advances, what appears to be complex may become very simplest. DNA contains a lot of information and not all of it is expressed. Perhaps our complex understanding of some things is like that? The more we learn, the more we might realize the information is a lot of chatter, and what is really important to know is quite simple?

When I was a kid, I was always inventing things, and my inventions were never the simple thing that did the job much better than what I was designing. A computer program could perhaps get to the nitty gritty much faster than our creative minds? A lot of information becoming fewer basic facts.
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon


Re: faster better science

Postby Athena on May 5th, 2017, 1:37 pm 

Forest_Dump » April 28th, 2017, 2:19 pm wrote:Well, I couldn't access it either. However, from what I did managed to read:

At first I was astonished by the numbers. But then maybe not so much (assuming the numbers are correct). Some brief thoughts included wondering how much of this "publishing" is from the industry of producing scientists? With the publish or perish meme as well as the need to get credentials, publishing anything and everything, partiocularly given that the peer review process is so thankless while time consuming as well as rarely anonymous (who and how do editors choose as reviewers?) I can see how things could start to get out of hand particularly when there can be kinds fo competitions to get students through the process, into labs and institutions, etc. And then there is the demand to get products to market as fast as possible and a few good papers that no one has time or inclination to critique can translate into a lot of money... Bottom line, the scholarly and critical thinking ethos has certainly been hammered over the last few decades.

But still, a certain amount of automation may well assist to some degree. I don't think anyone would argue that computers haven't helped a lot with some things and who knows what the next generation of information technology could assist with. The danger, of course, is the same old one of blindly believing that technology, data processing, etc., can do it all. At some point we can only hope that some solid thinking by actual smart people (as opposed to just those who simply assert they are smart on the basis of mass-produced credentials including a few of those peer-reveewed papers that no one has read) will be keeping an eye on what is really happening with the "faster, better science" and those selling it.


I think Newton got more credit than he should have gotten because it controlled the flow of information and who got credit for it. A nice thing about computers is they will not have this human agenda and as more and more people are involved, a small group we loose control. Science will become increasingly democratized.
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon


Re: faster better science

Postby vivian maxine on May 5th, 2017, 2:04 pm 

Thank you, Athena. I repeat a saying that is as old as computers. "The information in a computer is only as accurate as what a human put into it." That isn't the exact wording but has the same meaning. You'll have heard it, I'm sure. It was the response to "It has to be right. It's in the computer." So, AI judgment will be as good as is the judgment of the human who programmed it.
vivian maxine
Resident Member
 
Posts: 2837
Joined: 01 Aug 2014


Re: faster better science

Postby Athena on May 6th, 2017, 12:24 pm 

vivian maxine » May 5th, 2017, 12:04 pm wrote:Thank you, Athena. I repeat a saying that is as old as computers. "The information in a computer is only as accurate as what a human put into it." That isn't the exact wording but has the same meaning. You'll have heard it, I'm sure. It was the response to "It has to be right. It's in the computer." So, AI judgment will be as good as is the judgment of the human who programmed it.


Except the talk of AI is that it can have better judgment than humans because a self-correcting system can be installed. Or more important, AI can not make the thinking errors common to humans. Daniel Kahneman has done several youtubes explaining our thinking and why we make so many errors. I think what he has to say about how our brains work is fascinating, and it can lead to the idea that in some ways computers can do better because they can not be tricked as easily as our brains can be tricked into having wrong information. Of course this also why computers can not have human judgment. Sometimes the right answer is the wrong answer for a human being because of the difference between us and computers.
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon


Re: faster better science

Postby vivian maxine on May 6th, 2017, 1:21 pm 

I know what you are saying. I just find it very hard to believe. AI has to be programmed by humans. Like the computer (is it an AI?) what the human puts in will come out.

Of course, the way this world is panning out, I'll probably be proven totally wrong --- they "proved" to the Wright brothers that they could not fly; gravity, you know --- but I'd not be happy being directed by a machine. :-)
vivian maxine
Resident Member
 
Posts: 2837
Joined: 01 Aug 2014


Re: faster better science

Postby Heavy_Water on May 6th, 2017, 7:02 pm 

But we've been using computers for over two decades now as a tool to aid with research, number crunching, and writing and publishing peer reviewed writings.

So from where does the claim that we use a smaller percentage of available knowledge come from? With the aforementioned computers, and the information super highway online, we are now able to glean more data and research than ever before.

In those opinion had the quality​ of our work diminished? I think it's never been better. Can you prove me wrong?

Remember why the internet was created? LOL. And by whom.

Scientists.


Wishing to exchange information.

We have only built on that for the past three decades.

And now have access to more info than ever before in history.

Computing power, processing speed, memory. All increased by levels of magnitude since eniac. Your smart phone had more memory.

Therefore, we have assess to all that. And use it for a never seen before utilization of reams of knowledge . Amounts unheard of, and indeed at SciFi levels not so very long ago.

Peer review is more competitive and discerning than ever. As the justification for funding is tight. Therefore, a need if not a requirement for quality and accuracy.

I basically vehemently disagree with the entire premise of the article. And in fact think it could not be more wrong. I believe the exact opposite of it's basic premise is the real truth.
User avatar
Heavy_Water
Banned User
 
Posts: 61
Joined: 02 May 2017
Location: Texas


Re: faster better science

Postby Braininvat on May 6th, 2017, 7:40 pm 

We must have read different articles. The one I read was about the uses of text mining, with the aid of natural-language algorithms, something that's widely accepted as a coming thing. The amounts of data are now so vast that the Baconian method outlined could practically be used with the help of fast processors. Hard to disagree with that.
User avatar
Braininvat
Forum Administrator
 
Posts: 6580
Joined: 21 Jan 2014
Location: Black Hills


Re: faster better science

Postby Athena on May 7th, 2017, 4:27 pm 

vivian maxine » May 6th, 2017, 11:21 am wrote:I know what you are saying. I just find it very hard to believe. AI has to be programmed by humans. Like the computer (is it an AI?) what the human puts in will come out.

Of course, the way this world is panning out, I'll probably be proven totally wrong --- they "proved" to the Wright brothers that they could not fly; gravity, you know --- but I'd not be happy being directed by a machine. :-)


Here is why we should not trust our judgment. Understanding this would make it clear why sometimes a machine has better judgment.

https://www.youtube.com/watch?v=mWaIE6u3wvw
Athena
Banned User
 
Posts: 1936
Joined: 22 Apr 2012
Location: Eugene, Oregon



Return to Engineering & Electronics

Who is online

Users browsing this forum: No registered users and 5 guests