Pessimism in Science and Industry

[YNM]

On 12/19/2016 at 8:50 AM, tater said:

Being ill prepared in the past doesn't compare to being well prepared now or then. Well prepared exploration and colonization of earth was trivial. Most died to disease, not deprivation. In the 1800s half the Europeans going to the Caribbean died of disease (say French troops during the Haitian revolution).

Perhaps you should try to look why the "settlers" won the wars against the indigenous people.

 

On-Topic :

It seems like the thread was going about "is it hard or not", for sure any new exploration will be hard, in fact much, much harder than what it was. Trip to the Moon can take a bit longer than a week, but a trip to Mars is a few 50x that.

I suppose humans are just caught in the wilderness of space, for now. Surely when there's a proper need, someone will go further on. Just like how rocketry have evolved from puny, unreliable ones to something enough to destroy a small country in a few decades, most development happening in the latter parts.

[magnemoe]

2 hours ago, Camacha said:

Car development has the same issue. Cars have become impossibly expensive to develop from scratch, so platforms get extensively reused and shared between companies. Meanwhile, trying new things does not happen anymore, as the risk is huge and the reward uncertain. Unfortunately, it is not just platforms that are expensive to develop. Almost all cars look like versions of other cars, with all the brands aiming for middle-of-the-road-models that appeal to everyone, but are not special to anyone either.

Multiple effects here, first its plenty of unusual cars, everything from sports cars to mobile homes of all types. 
Mass marked cars is another issue, an very competitive marked and the production lines are so optimized for large volume production you have an huge up front cost. Changing model typically require you to rebuild the lines. You also have to optimize the car in all sort of ways making it very expensive to design and set in production. I assume its cheaper to develop an new Ferrari model as you don't have to focus so much on keeping costs down while matching 8 competing cars on key points and make it easy to assemble.  

Cars also have other constrains, they need to be safe, aerodynamic and have good room inside while not being to large on outside. Finally they have to look good while not being to expensive. And as you say because of the cost you don't want to take risks in how they look. 
On the other hands cars has changed a lot during this time, far safer, pollute less while getting stronger engines. an gigantic list of features is standard today. 
Future will see more automation up to self driving cars, not sure how standard electrical cars will be but hybrids will end up as standard except them. 

[wumpus]

On 12/20/2016 at 2:10 AM, mikegarrison said:

In commercial aviation especially, the manufacturers are very conservative because it costs billions of dollars to kick off a new program. If it's a major failure, your entire company can be ruined. And if there is even one fatality from a new airplane type, all of a sudden you can have the public completely rejecting it. Just as bad as if it didn't work at all.

Try looking at all the "revolutionary" memory and storage technologies that have been researched over the years (typically involving "3d something").  Nearly all of them had great promise, and could ship power point slides showing just how huge an advantage they had over the current schemes.

But Moore's law was brutal.  If you had a "1000x" advantage over DRAM, you had better have your product in customer hands within 10 years or memory would have already passed you (and you really didn't expect the *shipping* product to have that 1000x advantage, did you?).  And hard drives were actually moving faster than that.  And while flash might be replacing hard drives, it is certainly due to superior speeds as they are at least ten times as expensive per byte and unlikely to ever catch up there.

I had great hopes that the Intel/Micron "3dxpoint" could at least shove its way into a intermediary buffer between DRAM and flash, but it looks like it just doesn't have the [latency] speed.  And that is a tech that made its way to the point of shipping samples.

[mikegarrison]

Right. That's the issue. Your revolutionary technology has to have an advantage over the alternative by the time it is ready to go into service. Too often in history people have thrown a lot of money into technologies because they were better than what existed at the time the new technology was kicked off. But by the time it was ready, their revolutionary technology was now a dud.

[todofwar]

27 minutes ago, mikegarrison said:

Right. That's the issue. Your revolutionary technology has to have an advantage over the alternative by the time it is ready to go into service. Too often in history people have thrown a lot of money into technologies because they were better than what existed at the time the new technology was kicked off. But by the time it was ready, their revolutionary technology was now a dud.

Of course, the question becomes which tech would have advanced faster. Let's say I build a graphene hard drive in a lab. It stores 10x more than the current market leader, and has a superior read out rate (I really don't know computer hard drive terms so bear with me). By the time I get it into a product, conventional hard drives have outpaced mine or made it so the benefit is no longer worth the added cost of scaling production to be able to match what can be made on existing machinery. But my graphene drive will improve in certain metrics at 1.05x the rate of conventional, and match improvement rates for the rest. Long term, it will be the better solution. So, is the market driving us to a local minimum in the end? It is (edit: potentially) a big issue with many technologies out there. 

Edited December 21, 2016 by todofwar

[mikegarrison]

6 minutes ago, todofwar said:

Of course, the question becomes which tech would have advanced faster. Let's say I build a graphene hard drive in a lab. It stores 10x more than the current market leader, and has a superior read out rate (I really don't know computer hard drive terms so bear with me). By the time I get it into a product, conventional hard drives have outpaced mine or made it so the benefit is no longer worth the added cost of scaling production to be able to match what can be made on existing machinery. But my graphene drive will improve in certain metrics at 1.05x the rate of conventional, and match improvement rates for the rest. Long term, it will be the better solution. So, is the market driving us to a local minimum in the end? It is (edit: potentially) a big issue with many technologies out there. 

Well, if you really believe in it, you keep throwing money at it until it pays off or you run out of money. But beware the "second mover advantage".

[magnemoe]

5 hours ago, todofwar said:

Of course, the question becomes which tech would have advanced faster. Let's say I build a graphene hard drive in a lab. It stores 10x more than the current market leader, and has a superior read out rate (I really don't know computer hard drive terms so bear with me). By the time I get it into a product, conventional hard drives have outpaced mine or made it so the benefit is no longer worth the added cost of scaling production to be able to match what can be made on existing machinery. But my graphene drive will improve in certain metrics at 1.05x the rate of conventional, and match improvement rates for the rest. Long term, it will be the better solution. So, is the market driving us to a local minimum in the end? It is (edit: potentially) a big issue with many technologies out there. 

This, its often that the previous technology had reach an plateau, increasing performance became exponential expensive. Piston engines on planes is an example and they was replaced by jet engines. 
Exception is small private planes there cost is more important than performance. 
Hard drives, you had high performance hard drives once, up to 15k rmp to increase data rate and reduce delay, they also had overpowered magnets for moving the reading heads.
Flash killed them. 

Both flash and lcd displays is a bit interesting, flash killed the performance performance hard drive and has reduced hard drive to bulk storage, lcd killed the cathode displays totally. 
And both took an indirect route, they did not started out challenging hard drives and crt. 
lcd started as small low resolution displays in calculators, clocks and other displays. they replaced existing technology here and worked in improving the display until it was used in laptops, they could be more expensive and crappy than crt as you could not use crt in laptops. Then came the time you could fight crt on price and quality and soon replace them. 

Flash the same way, started with storing a few bit and only manage some hundred changes. worked well enough for storing presets and better than battery backup ram. next up was megabyte storage, feature phones with camera, early digital cameras, game console save sticks, early usb thumb drives. 
Size grows while performance increase until you got ssd. 

In short you need either an obvious benefit over something who has reached an plateau, or you need an path who can finance you all the way. 
Jet engines would probably came slower without ww2 and the cold war who pushed fast fighter jets. 

[mikegarrison]

1 hour ago, magnemoe said:

This, its often that the previous technology had reach an plateau, increasing performance became exponential expensive. Piston engines on planes is an example and they was replaced by jet engines. 
Exception is small private planes there cost is more important than performance.

This is not the real story. Or at least, not the whole real story.

Jets are more efficient than propellers at high speeds. Propellers are more efficient than jets at low speeds. As speeds increased, jets became the better option. But jets still had real problems at slow speeds (including takeoff and landing, where they really don't work all that well).

Then the turbofan was invented, which is sort of part-jet, part-propeller. And it fit very nicely into the speed range between jets and propellers.

So airplanes that are not designed to fly particularly fast use propellers (most general aviation, commuter turboprops, etc.). Planes that are designed to fly close to transonic speeds use turbofans. And planes that are designed to fly very fast use jets (or very low bypass turbofans).

[Jonfliesgoats]

A little note about jets:

The cost per ton per mile is lower than props because of the lower weight of the engines and higher altitudes with jets.  The savings in mass and higher altitude allow more payload to be carried much farther.  We typically fly around 84 to 90 percent the soeed of sound.  The speed is nice.

Turboprops are nice for short and medium haul for the same reason.  

[Jonfliesgoats]

Having scrolled through the thread again, there are a lot of good points made.  I'd like to remind people that my discussion about pessimism isn't about whether it's hard to go to space.  It will be hard and it will be risky.  I bring up pessimism because many folks, especially those only with access to pop-culture, are thinking this is impossibly hard.  Worse, they sometimes support pessimistic ideas with facts that are skewed to support pessimism.  

Mankind's expansion into the cosmos is entirely doable.  Pessimism allows us to feel better about abandoning high goals.  An idea that the world ends and hell starts at the Himalayas allows us to feel better about not venturing further from Greece (I expect some howls at this reference).  A cost-benefit-risk analysis, however, is good planning.  

With regard to fielding technology, there are some great points made as well.  There are lots of emerging techs that simply come to market too late.  Finding the best way forward is really tough.

I like everyone's comments and can't scroll and like fast enough to hit everyone's posts.

[kerbiloid]

3 hours ago, Jonfliesgoats said:

I bring up pessimism because many folks, especially those only with access to pop-culture, are thinking this is impossibly hard.

Probably because the humankind has already reached the threshold when understanding of "how it works" had gotten unachievable for a regular human's mind forces.

The current version of human had appeared ~200 000 years, with fire, spears, grass clothes and hasn't significantly changed since then.

About 40 000 years ago there happened a major upgrade (cognitive revolution) when they woke up and began documenting the world with painting of their hunting, dancing and other indecencies.
(As almost all modern people understand arts - or what they call so - we can presume that bad spectators have been eaten killed killed and eaten by aggrieved artsts and snobs.)

About 12000 years ago there happened a lesser major firmware upgrade: neolithic.
There appeared crafting, trading, building (i.e. perks like in Skyrim) and so on. But most of people became just peasants and herders, only a minority achieved higher levels, several persons per tribe.
And we still have this version of firmware.

In sweet medieval fantasy times most of people were just plowing, pasturing or fishing. Enough good to survive, but not so good to advance, a simple work for simple mind.
A blacksmith (especially), a miller, a healer were guru doing strange, obscured and scary things (so obviously warlocks), but with results still understandable for a regular peasant.
City crafts were advancing, and the most complicated of them were accessible for the minds of the very few, and mostly like experimental narratives, not as engineer's understanding.
All known science (and jurisprudence in addiion) could be studied in 10-20 years, giving PhD to the superiority complex..

In XVII..XIX an industrial revolution has taken place, and geographical studies have expanded. A spate of new information flooded all across.
But absolute majority were still peasants or simple workers, they were doing mostly the same as 12000 years ago, but with better instruments.
An experienced engineer still could understand any machine he could meet. Scientists divided into different sciences, as there was too much to learn all.

In mid-XX engineer knew what to do with specific theme he knows and would only laugh being asked about something from another field.
Scientific revelations which could be done using a school physics room and a bicycle workshop by a lone enthusiast, were already done. Science turned into assistants' lab routine.

Currently we have reached the point when nobody has detailed picture, he can see only a total silhouette.
Any human activity steadily moves in direction: "either be a guru, or just press a button the button will press herself."
And almost all humankind is still with 12000 year old version of firmware. Even engineers. 
They can understand that if drop a grain into ground and pour it with water, it will rot grow into a wheat, while they need to understand a jet functioning (nothing common with their firmware is adapted for).

So, the pessimism is enough predictable.

Edited December 22, 2016 by kerbiloid

[magnemoe]

5 hours ago, mikegarrison said:

This is not the real story. Or at least, not the whole real story.

Jets are more efficient than propellers at high speeds. Propellers are more efficient than jets at low speeds. As speeds increased, jets became the better option. But jets still had real problems at slow speeds (including takeoff and landing, where they really don't work all that well).

Then the turbofan was invented, which is sort of part-jet, part-propeller. And it fit very nicely into the speed range between jets and propellers.

So airplanes that are not designed to fly particularly fast use propellers (most general aviation, commuter turboprops, etc.). Planes that are designed to fly close to transonic speeds use turbofans. And planes that are designed to fly very fast use jets (or very low bypass turbofans).

True, propeller, jet and turbofan has their speed envelopes.  
I was mostly thinking of the shift from piston engines to turbines, here the turbines dominate totally except the small and cheap planes. 
Turbines has far higher trust to weight ratio and at least for higher effect engines they are more reliable. 

[Jonfliesgoats]

Nice points.  I certainly think they have some validity.  I am not entirely sure this pessimism is a new phenomenon, though.  True enough it is hard to grasp intangible risks and reward.  It is also hard for the pay person to understand nuances of those risks.  I think another key factor is our latent fear of change.  This fear serves us well, usually and we have evolved a sense of comfort from the familiar.  

I grew up in a trashy farm town.  Our great source of excitement was going to the local truck stop and watching people come and go while we got drunk on hidden stashes of booze.  

Many of us were convinced we weren't going to leave the local county because we simply weren't from the classes of people that do those things without enlisting.

[Jonfliesgoats]

3 hours ago, kerbiloid said:

Probably because the humankind has already reached the threshold when understanding of "how it works" had gotten unachievable for a regular human's mind forces.

The current version of human had appeared ~200 000 years, with fire, spears, grass clothes and hasn't significantly changed since then.

About 40 000 years ago there happened a major upgrade (cognitive revolution) when they woke up and began documenting the world with painting of their hunting, dancing and other indecencies.
(As almost all modern people understand arts - or what they call so - we can presume that bad spectators have been eaten killed killed and eaten by aggrieved artsts and snobs.)

About 12000 years ago there happened a lesser major firmware upgrade: neolithic.
There appeared crafting, trading, building (i.e. perks like in Skyrim) and so on. But most of people became just peasants and herders, only a minority achieved higher levels, several persons per tribe.
And we still have this version of firmware.

In sweet medieval fantasy times most of people were just plowing, pasturing or fishing. Enough good to survive, but not so good to advance, a simple work for simple mind.
A blacksmith (especially), a miller, a healer were guru doing strange, obscured and scary things (so obviously warlocks), but with results still understandable for a regular peasant.
City crafts were advancing, and the most complicated of them were accessible for the minds of the very few, and mostly like experimental narratives, not as engineer's understanding.
All known science (and jurisprudence in addiion) could be studied in 10-20 years, giving PhD to the superiority complex..

In XVII..XIX an industrial revolution has taken place, and geographical studies have expanded. A spate of new information flooded all across.
But absolute majority were still peasants or simple workers, they were doing mostly the same as 12000 years ago, but with better instruments.
An experienced engineer still could understand any machine he could meet. Scientists divided into different sciences, as there was too much to learn all.

In mid-XX engineer knew what to do with specific theme he knows and would only laugh being asked about something from another field.
Scientific revelations which could be done using a school physics room and a bicycle workshop by a lone enthusiast, were already done. Science turned into assistants' lab routine.

Currently we have reached the point when nobody has detailed picture, he can see only a total silhouette.
Any human activity steadily moves in direction: "either be a guru, or just press a button the button will press herself."
And almost all humankind is still with 12000 year old version of firmware. Even engineers. 
They can understand that if drop a grain into ground and pour it with water, it will rot grow into a wheat, while they need to understand a jet functioning (nothing common with their firmware is adapted for).

So, the pessimism is enough predictable.

 

[PB666]

http://retractionwatch.com/2016/12/22/journal-cleans-house-retracting-6-cancer-papers-plagiarism/

Yep. Editor fail to create an adequate pre-review process. Only about a decade and a half behind the times.

Skeptical, at one point we had an 85% rejection rate with about a 60% summary rejection rate.

I've made more scientist groan than just about anyone you will ever meat, but again if you can become a full professor in some University and BRICA country with just two mediocre to trivial publications . . . . .. .

 

 

[wumpus]

9 hours ago, kerbiloid said:

About 12000 years ago there happened a lesser major firmware upgrade: neolithic.
There appeared crafting, trading, building (i.e. perks like in Skyrim) and so on. But most of people became just peasants and herders, only a minority achieved higher levels, several persons per tribe.
And we still have this version of firmware.

This tends to be assumed (although plenty assume paleolithic), but it is rather hard to verify.  We can verify that some humans have genes that evolved (or at least spread) during the neolithic era (I carry genes which allow me to be lactose tolerant, that wouldn't exist in the paleolithic era).  To be honest, I'm not sure that evolutionary pressures in a medieval village were all that different from a neolithic one.

There was a shocking amount of change (at least in the west) from 1800-1900.  While technology since then has been spectacular, most modern people could probably get by (the technology if not the culture/politics) in 1900, but would be lost in 1800.  

[kerbiloid]

11 minutes ago, wumpus said:

There was a shocking amount of change (at least in the west) from 1800-1900.  While technology since then has been spectacular, most modern people could probably get by (the technology if not the culture/politics) in 1900, but would be lost in 1800.  

Just because in 1800 they would use much more primitive tools and live in worse conditions unusual for them.
But mentally they would be more comfortable with lo-tech, rather than hi-tech.

Hi-tech is just nothing common with what the human was adapted for millenia.

Edited December 22, 2016 by kerbiloid

[magnemoe]

1 hour ago, wumpus said:

This tends to be assumed (although plenty assume paleolithic), but it is rather hard to verify.  We can verify that some humans have genes that evolved (or at least spread) during the neolithic era (I carry genes which allow me to be lactose tolerant, that wouldn't exist in the paleolithic era).  To be honest, I'm not sure that evolutionary pressures in a medieval village were all that different from a neolithic one.

There was a shocking amount of change (at least in the west) from 1800-1900.  While technology since then has been spectacular, most modern people could probably get by (the technology if not the culture/politics) in 1900, but would be lost in 1800.  

Lactose tolerance evolved, not sure of other outside of immune system who probably changed a lot.

Changes was larger from 1900 to 2000 than from 1800 to 1900. Yes 1800 most was farmers while 1900 they was factory workers so it was an giant social change. 
But you could take an farmer from 1800 and put in on an farm in 1900 and he would understand it.
Same with an 1800 to 1900 kitchen. Lots of new cool stuff but most of it is understandable. 
1900 to 2000 would be totally alien. 
Wonder how 2100 will look like? Good chance the shock would be less than 1900 to 2000.