Is Science Slowing Down? - Politics Forum.org | PoFo

Wandering the information superhighway, he came upon the last refuge of civilization, PoFo, the only forum on the internet ...

Anything from household gadgets to the Large Hadron Collider (note: political science topics belong in the Environment & Science forum).

Moderator: PoFo The Lounge Mods

#14984155
I just read this rather interesting article, Is Science Slowing Down?, which tests the rather commonsensical view that science is moving forward ever faster and finds it wanting.

This is the standard presentation of Moore’s Law – the number of transistors you can fit on a chip doubles about every two years (eg grows by 35% per year). This is usually presented as an amazing example of modern science getting things right, and no wonder – it means you can go from a few thousand transistors per chip in 1971 to many million today, with the corresponding increase in computing power.

But BJRW have a pessimistic take. There are eighteen times more people involved in transistor-related research today than in 1971. So if in 1971 it took 1000 scientists to increase transistor density 35% per year, today it takes 18,000 scientists to do the same task. So apparently the average transistor scientist is eighteen times less productive today than fifty years ago. That should be surprising and scary.


The above quote is the angle of attack that the proposition is taking; that apparent exponential growth in scientific output is actually a game of diminishing returns masked by even greater exponential growth in inputs.

One interesting implication that I draw from it is that, excepting the AI wild card, the human population must maintain exponential growth in order to sustain continued scientific and technological development.
#14984166
SolarCross wrote:I just read this rather interesting article, Is Science Slowing Down?, which tests the rather commonsensical view that science is moving forward ever faster and finds it wanting.



The above quote is the angle of attack that the proposition is taking; that apparent exponential growth in scientific output is actually a game of diminishing returns masked by even greater exponential growth in inputs.

One interesting implication that I draw from it is that, excepting the AI wild card, the human population must maintain exponential growth in order to sustain continued scientific and technological development.


I currently work in the semiconductor industry. I agree and disagree with what's being said here. I would argue that science is just getting more complicated, more so than our ability to be productive is going down.

First, the progress of all scientific discovery is NOT entirely dependent on our ability to place more transistors in a given square area of silicon. Yes, the fact that it's become harder to do so will slow things down, but not to a crawl, at least not yet...

The reason it takes more research/work to get more transistors is simply because we are starting to hit what the industry calls "The wall". That is, we are getting close to the physical limit that we can shrink down silicon transistors. Current generation transistors are at the scale of 7nm right now. It's believed "the wall" will happen somewhere around 1nm. The generation after 7nm is 4nm. We might get another generation between 4nm and 1nm. Nonetheless, the silicon industry believes there is about 2-3 more generations left before we hit the wall. This will take about another 10-15 years or so to get to that point.

That said, there are massive improvements outside of die shrinking to be had. Some include micro-architecture improvements, die stacking, straining silicon, etc. etc. All of this can combat the fact that we aren't able to shrink transistors further. Last, we there are other forms of non-silicon based computing that could replace all this anyway.

Thus, I think computer processing power will continue to increase rapidly. It will take more researchers and engineers to do it. In that sense, sure, we can say the industry is less productive. Though, a more accurate way to put it is that the complexity of technology is increasing faster than we can simplify it.

Transistor shrinking was just the cheapest and easiest thing to do to get more performance from computer chips. There are many more ways to get more performance out of them that do not involve shrinking.
#14984329
ness31 wrote:Linking scientific progress to how much you can fit on a chip seems a bit silly :hmm:


That wasn't the only example but the subject isn't easy to quantify and moore's law is just an easy place to start with some actual numbers behind it. If you want to say it is an engineering project not a scientific one then you are missing that engineering is just the application of science, they are not separate things.
#14984331
I don’t really see how science or knowledge can slow down. They are accumulative, so we always have more as a basis for creating more. It just may be in different fields or directions. Until we start finding profound truths, then I don’t see a slow down.
#14984333
SolarCross wrote:That wasn't the only example but the subject isn't easy to quantify and moore's law is just an easy place to start with some actual numbers behind it. If you want to say it is an engineering project not a scientific one then you are missing that engineering is just the application of science, they are not separate things.


Pretty much what One Degree said.

True scientific advancement will come when we learn to live harmoniously with the natural world. But that isn’t easily jotted down so it becomes a bit hard to quantify.
#14984335
Rancid wrote:Transistor shrinking was just the cheapest and easiest thing to do to get more performance from computer chips. There are many more ways to get more performance out of them that do not involve shrinking.


Yes, exactly!
I understand that the IT technology merely depends on those 1s and 0s ...
I propose to start using 2s as well, that should open up the field !
#14984336
SolarCross wrote:I just read this rather interesting article, Is Science Slowing Down?, which tests the rather commonsensical view that science is moving forward ever faster and finds it wanting.



The above quote is the angle of attack that the proposition is taking; that apparent exponential growth in scientific output is actually a game of diminishing returns masked by even greater exponential growth in inputs.

One interesting implication that I draw from it is that, excepting the AI wild card, the human population must maintain exponential growth in order to sustain continued scientific and technological development.


Like others have said, the future is not grim, there are possible alternatives to just increasing transistor + not all scientific progress is a direct linear relation with computational power. But what I do not think would ever be the answer would be to just keep making more people as a solution. For once, you can just add more computers. Second there is a decent chance that in the next 50-60 years we might lose a significant portion of the workforce (almost certainly in industrialized countries) to automation and AI. That could be your source of extra people if you do end up needing it :lol: having more babies is not the answer I think unless we become an interplanetary species, in which case, it might be the answer.
#14984338
Ter wrote:
Yes, exactly!
I understand that the IT technology merely depends on those 1s and 0s ...
I propose to start using 2s as well, that should open up the field !


Funny you say that. I think quantum computing is supposed to do something like that.

It would make more sense to expand out to 0,1,2,3 not just 0,1,2. Powers of two work easier in computers. However, 0,1,2,3 is just a two bit number.

00
01
10
11

That's 0,1,2,3
#14984341
:roll:

out of 0s and 1s you can make any number. The neurons in your head are technically binary as well as they either fire or don't fire which is logically the same as a computer's true or false, 0 or 1. Neurons are just slower, smaller and more energy efficient.

Let's not get hung up on moore's law the article's contention is that the slow down applies far more widely than that.
#14984356
SolarCross wrote::roll:

out of 0s and 1s you can make any number. The neurons in your head are technically binary as well as they either fire or don't fire which is logically the same as a computer's true or false, 0 or 1. Neurons are just slower, smaller and more energy efficient.

Let's not get hung up on moore's law the article's contention is that the slow down applies far more widely than that.

This is ture. However, the brain and its neurons have a workaround for it. Every neuron end up being innervated (receiving connection from other neurons) by many neurons and they have a "threshold" for firing. This threshold can be met by either spatial summation (multiple neurons innervating the same neuron fire at the same time and meet the treshold by "summation of potential) or by temporal summation (if one neuron fires many times quickly). Long story short, it is a very complex process which augments the simple "0" and "1" nature of a single neuron exponentially and can make for many, many mistakes that would be catastrophic in a traditional silicon chip. Not really relevant to the conversation but interesting of how different these two types of computers work.
User avatar
By Ter
#14984363
XogGyux wrote:This is ture. However, the brain and its neurons have a workaround for it. Every neuron end up being innervated (receiving connection from other neurons) by many neurons and they have a "threshold" for firing. This threshold can be met by either spatial summation (multiple neurons innervating the same neuron fire at the same time and meet the treshold by "summation of potential) or by temporal summation (if one neuron fires many times quickly). Long story short, it is a very complex process which augments the simple "0" and "1" nature of a single neuron exponentially and can make for many, many mistakes that would be catastrophic in a traditional silicon chip. Not really relevant to the conversation but interesting of how different these two types of computers work.

I was just going to say that but @XogGyux somehow stole my thoughts and posted it before I could.
#14984367
Yes, my neurons also take issue with being labeled ‘binary’. They would like you all to know that they are ‘pan-neuronic’ and would :D thus like to be addressed as so
#14984377
ness31 wrote:Yes, my neurons also take issue with being labeled ‘binary’. They would like you all to know that they are ‘pan-neuronic’ and would :D thus like to be addressed as so

What pronoun your neurons prefer?
He/She/It/They/Ze? something else?
#14984417
SolarCross wrote:Let's not get hung up on moore's law the article's contention is that the slow down applies far more widely than that.


Yes, a slow down is possible. Given how scientific progress and technology is becoming more complex. This is where the jobs of the future will be!

However, who's to say there might not be some giant discovery that accelerates progress again? Moreover, it's generally understood that scientific and technological progress does not happen gradually. Usually there is one big discovery/development that happens which ushers in a new era of progress. Then thing slow down, until the next big thing.

Basically science works in paradigm shifts. For example, going from Newtonian mechanics to relativity.

I do not think Gene was a communist. He was a p[…]

This took the fucking biscuit by the way https://[…]

The Popular Vote...

(shrugs) If you say so, I disagree and have laid o[…]

Speaking of bias, these media and fact checking s[…]