Author Topic: What does machine learning mean to you?  (Read 10759 times)

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
What does machine learning mean to you?
« on: September 22, 2016, 08:57:17 PM »
I went to this talk by an impressionable well spoken youngish academic economist who seems to mistaken warehouse labelling and basic human guided robotics from a couple decades ago for machine learning. He expressed little knowledge about science and technology, deep learning, data mining algorithms etc let alone machine learning. Also a lack of foresight on workforce trends like casual jobs, avoiding tax, being a manager, employment hierarchies, bitcoin trends, marketable tech vs prototypes, microfinancing in agrarian developing economy dynamics... just too many blindspots.

He seemed truly convinced 47% of jobs are at high risk of being lost to machines when that figure seems as arbitrary as 50% yes or no. I'm no expert at statistics but realistic when considering anomalies like the rise of casual jobs, non-taxable incomes, sub contacting, young people selling apps and making money via online, I'd forecast the rate of taxable full time positions falling to maybe 10-20% in place of casual, non taxeable income. Seems stupid to suggest 47% being lost to machines.

I don't think this chap has ever worked in a warehouse because only 1 person drives the forklift at a time due to safety regulations. If a machine is doing this it still needs someone to supervise this machine. In actuality since my uncle owns car factories, he had to employ more people after they installed robotics because of safety monitoring but of a different/ higher skillset. Instead of mechanics, engineers etc.

Machine driven warehouses require more expert monitoring. More safety regulations. Interesting he didn't even mention the fact machine driven warehouses have had to employ more ground staff in the warehouse but of a different skillset. That in actual fact companies had to employ more staff for robotic kind of warehouses. This chap was stuck on the industrial rev where machines replace panga cutters.

I'm wondering whether this stigma of academic economists being just history buff blabbers with no work experience to forecast or conduct quality research goes across the board. I noticed they make like a tenth in terms of income of what real world economists/stockbrokers make.

I wanna ask, what does machine learning mean to you?

Offline Empedocles

  • VIP
  • Enigma
  • *
  • Posts: 823
  • Reputation: 15758
Re: What does machine learning mean to you?
« Reply #1 on: September 23, 2016, 10:01:44 AM »
I went to this talk by an impressionable well spoken youngish academic economist who seems to mistaken warehouse labelling and basic human guided robotics from a couple decades ago for machine learning. He expressed little knowledge about science and technology, deep learning, data mining algorithms etc let alone machine learning. Also a lack of foresight on workforce trends like casual jobs, avoiding tax, being a manager, employment hierarchies, bitcoin trends, marketable tech vs prototypes, microfinancing in agrarian developing economy dynamics... just too many blindspots.

He seemed truly convinced 47% of jobs are at high risk of being lost to machines when that figure seems as arbitrary as 50% yes or no. I'm no expert at statistics but realistic when considering anomalies like the rise of casual jobs, non-taxable incomes, sub contacting, young people selling apps and making money via online, I'd forecast the rate of taxable full time positions falling to maybe 10-20% in place of casual, non taxeable income. Seems stupid to suggest 47% being lost to machines.

I don't think this chap has ever worked in a warehouse because only 1 person drives the forklift at a time due to safety regulations. If a machine is doing this it still needs someone to supervise this machine. In actuality since my uncle owns car factories, he had to employ more people after they installed robotics because of safety monitoring but of a different/ higher skillset. Instead of mechanics, engineers etc.

Machine driven warehouses require more expert monitoring. More safety regulations. Interesting he didn't even mention the fact machine driven warehouses have had to employ more ground staff in the warehouse but of a different skillset. That in actual fact companies had to employ more staff for robotic kind of warehouses. This chap was stuck on the industrial rev where machines replace panga cutters.

I'm wondering whether this stigma of academic economists being just history buff blabbers with no work experience to forecast or conduct quality research goes across the board. I noticed they make like a tenth in terms of income of what real world economists/stockbrokers make.

I wanna ask, what does machine learning mean to you?

That's a toughie, from what you've written.

Was he specifically talking about the "coming" AI singularity?

Suss out this article, maybe that's what the speaker was hinting at?

Deep Learning Is Going to Teach Us All the Lesson of Our Lives: Jobs Are for Machines

If you're not yet nauseous and want to understand a wee bit more on the singularity, here's one of the best articles I've ever read on the subject:

Part 1 - The AI Revolution: The Road to Superintelligence

Part 2 - The AI Revolution: Our Immortality or Extinction


Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #2 on: September 23, 2016, 02:15:24 PM »
No, it was just inflated figures with poor statistical methodology and poor quality data. He qualitatively rated whether jobs would be lost to machines in the near future rather than using predictive simulation methods which would require intense statistical coding and take longer to conduct. The method he may have used was on manipulating retrospect data typical of p-hacking which isn't the right method for predictive/forecast analytics. Some might consider this as manipulating figures for political mileage. It's hard to fault because he rated things himself and this is the reason why these qualitative methods aren't applied for predictions/forecasting. Even if his predictions turned out to be wrong, he can just turn around and say, like a politician, well then, lots of new jobs were created and that's something I didn't look at. . in actuality he could have and should have using appropriate statistical methods like predictive analytics. Real world economists and statisticians I suppose would see through the ruse. I haven't even bothered reading his paper because it smells like a typical weak analysis with bling bling outcomes not even measuring anything but reiterating a market trend- easy to replicate because it's just reproducing a null market artefact.

If one is gonna conduct a qualitative rating system and feed it into an algorithm, then that data needs to be sourced from direct causal pathways. So like getting data from actual companies and with a fine comb categorize which employees quit or were replaced because of new tech. Often with new tech it requires a larger e-learning dept, larger IT depart etc. because things get more efficient. Employees made redundant isn't so much because machines can do their job better, but because they don't want to undergo e-learning training and prefer a position with their current skillset. More younger factory workers welcome new training whereas seasoned ones prefer to work elsewhere. I could go on. I noticed academic economists rely too much on overall reports like WHO, Deloitte etc. because that's how they wrote essays as undergraduate students and they haven't yet graduated into the real world with real world experiences with real world contacts. Economists use personal connections to source data and data quality matters. Like with me, I source Big Pharma data because I have contacts higher up who trust me with it. I hear complaints particularly among academics about Big Pharma hiding data and so forth but there's so many shitty defensive grant hovelling academics with child like egos the size of thy kingdom come like this chap so I don't blame Big Pharma for erecting a wall on them.


Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #3 on: September 23, 2016, 02:54:21 PM »

Suss out this article, maybe that's what the speaker was hinting at?

Deep Learning Is Going to Teach Us All the Lesson of Our Lives: Jobs Are for Machines

Interesting but hogwash. People tend to humanize machine feedback even though backend it's just a bunch of networked databases categorizing stuff and generating the most broad response. I noticed society is at a period of machine learning quackery. Like swearing that fortune teller is really psychic. I also noticed the chap from the lecture was blown away by image detection software when in reality what they sell is fairy tale. Imaging scans are promising in deep learning because it has a natural neural network in a consistently controlled environment, thereby reliable data. That doesn't extend to image databases because those networks aren't there yet. It needs to have a template for simulation.

IBM Watson and the likes keep screaming Cognitive computing by somehow mirroring a computer with a human brain- neural networks. What makes them think an optimal machine brain has any resemblance at all to a human brain? To me when just looking at data, I imagine something more flat and evenly spaced like a server because data doesn't need to save on convoluted spatial densities like the human brain squished into a skull. This convolution could be why humans are so illogical, fuzzy memories, confused etc. Machine brain in my opinion should have equal electrical flows, spatial densities, and sufficient wiggle room to evolve. And I think an area that looks to be neglected is the machine learning eco-system. These areas are where things should be architecturally designed in such a way so machines can gain access to enough IoT sensors, collect as much data, conduct own relational cues and eventually grow and evolve. Humans will always be at the top of the food chain and create even more jobs because their brains evolve and get smarter with each successive generation. The same can't be said for machines because they get outdated and discarded every so couple of years.

The one thing I agreed with the speaker was on the lack of perceptual knowledge. This I believe it because only of late are IoT sensors being deployed to collect data in different environments and thereby churn out more data so potential patterns can be investigated.

Offline Empedocles

  • VIP
  • Enigma
  • *
  • Posts: 823
  • Reputation: 15758
Re: What does machine learning mean to you?
« Reply #4 on: September 23, 2016, 04:19:30 PM »
Then the only thing I could advise you is to read Nick Bostrom's Superintelligence: Paths, Dangers, Strategies

Quote
Superintelligence asks the questions: what happens when machines surpass humans in general intelligence? Will artificial agents save or destroy us? Nick Bostrom lays the foundation for understanding the future of humanity and intelligent life. The human brain has some capabilities that the brains of other animals lack. It is to these distinctive capabilities that our species owes its dominant position. If machine brains surpassed human brains in general intelligence, then this new superintelligence could become extremely powerful?possibly beyond our control. As the fate of the gorillas now depends more on humans than on the species itself, so would the fate of humankind depend on the actions of the machine superintelligence.

But we have one advantage: we get to make the first move. Will it be possible to construct a seed Artificial Intelligence, to engineer initial conditions so as to make an intelligence explosion survivable? How could one achieve a controlled detonation?

This profoundly ambitious and original audiobook breaks down a vast track of difficult intellectual terrain. After an utterly engrossing journey that takes us to the frontiers of thinking about the human condition and the future of intelligent life, we find in Nick Bostrom's work nothing less than a reconceptualization of the essential task of our time.

It's a very good read.

Offline Nefertiti

  • Moderator
  • Enigma
  • *
  • Posts: 11329
  • Reputation: 26106
  • Shoo Be Doo Be Doo Oop
Re: What does machine learning mean to you?
« Reply #5 on: September 23, 2016, 04:59:21 PM »
I agree with veri about jobs - every revolution or disruption takes away and creates new jobs in equal measure. I never buy the Hollywood story about rise of the machines. The gorilla analogy is wrong because humans did not create gorillas and other species unlike machines which are not a species. Superintelligence should cure AIDS and cancer and other endless human problems. I don't foresee any takeover or anything near it. Machines like robots will remain TOOLS for humans to use.
♫♫ They say all good boys go to heaven... but bad boys bring heaven to you ~ song by Julia Michaels

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #6 on: September 23, 2016, 05:14:22 PM »
Smells like science fiction. I think I'll stick with Foucault, Hegel etc.

This is the kind of quackery I'm talking about. There isn't even an agreed definition of intelligence.

Algorithms don't experience the world in an organic, evolutionary context. The only thing to fear is some dictator abusing surveillance for some purported greater good. Machines notoriously breakdown, go err, glitch, moreso with complex code. The damn wires can fry. Machines don't sleep, dream, and mimic organic conditions necessary for intelligence to evolve.

The only way for an entity to be super intelligent is to have the capacity to make mistakes and evolve to a higher mode of social and emotional discourse. A sense of self-awareness and self-reflection, a sense of time, things that can't be coded into bits. Things that have been debated since civilization, and perhaps a reason why machines like humans have limits. Different kind of limits but nevertheless limits like the Tower of Babel and numerous ambitious feats built by humans.

Just because someone has photographic memory, doesn't mean their super intelligence surpasses average intelligence or they're perceived as powerful and a threat to democracy. There are too many dynamics involved which moot these quackery notions. Super intelligence may equally be judged as retarded if there isn't an appreciation. This is what I mean by machine learning eco-system. Even the most wide spread eco-systems intelligent or not aren't engineered to understand power dynamics. That being, there'll always be a power hierarchy regardless of super intelligence. There'll always be race/tribal factions, there'll always be a human touch to disrupt the status quo regardless of how dangerous a super intelligent eco-system can get. It's like the colonialists throughout history. They thought they git the better of native populaces with industrial prowess but soon realized they needed to work with the natives to make the eco-system work. Then realized it wasn't cost effective so set the natives free etc. Etc

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #7 on: September 23, 2016, 05:26:35 PM »
The problem with my Empy is that he's so into scifi and gorillas taking over the world like a dreamy scientist with idealized notions their clunky hands are gonna change/destroy world, cataclysmic proportions watch out. The world or history I should say move in smaller increments than the human mind battling with itself day in day out at the fact it's running out of time death looming psychological/protection mechanism. We project mysticism in all manner of speak to man made objects/projects/architects when we lack pragmatic skills in extrapolating unknowns. Then we make up super intelligent super power enemies like vampires, aliens and now robots, even though there hasn't been a single iota if evidence aliens, gorillas or even just tractors are super intelligent.

I agree with veri about jobs - every revolution or disruption takes away and creates new jobs in equal measure. I never buy the Hollywood story about rise of the machines. The gorilla analogy is wrong because humans did not create gorillas and other species unlike machines which are not a species. Superintelligence should cure AIDS and cancer and other endless human problems. I don't foresee any takeover or anything near it. Machines like robots will remain TOOLS for humans to use.

Offline MOON Ki

  • Moderator
  • Enigma
  • *
  • Posts: 2667
  • Reputation: 5780
Re: What does machine learning mean to you?
« Reply #8 on: September 23, 2016, 05:55:07 PM »
Then the only thing I could advise you is to read Nick Bostrom's Superintelligence: Paths, Dangers, Strategies

I don't know if you have ever heard of the Alcor Life Extension Foundation:   http://www.alcor.org/  It's an organization that, for a fee, freezes people until the day when technology has improved enough to bring them back to old state. They have been doing roaring business for decades.  Their fees have gone down quite a bit, but it still costs much less to freeze just the head than to freeze the whole body:  http://www.alcor.org/BecomeMember/scheduleA.html

Back in the 1980s, they got into trouble when they were accused of homicide.   Apparently, if just the head is to be frozen, then it is best to cut it off right after death or, even better, just before death---to avoid information loss.     The "trouble" was that someone had reported to the sheriff that they had done the latter, which would indeed be homicide.     A little circus then ensued, with the sheriff trying to find the head and Alcor constantly moving it.    But Alcor never denied cutting off the head, so the view was formed that their might be a basis for a charge of homicide.   Alcor was asked to explain itself.

Alcor had this to say: Even if they had cut off the head at the alleged time, there was no basis for a charge of homicide.    That might seem absurd, but it wasn't entirely so; their argument relied on how "death" was to be understood.   Consider typical definitions:

Quote
homicide: the deliberate and unlawful killing of one person by another;

killing: an act of causing death, especially deliberately;

death: the action or fact of dying or being killed;

One can see the circularity; so the matter of just what "death" is became an issue.  All agreed that it could be defined as "a permanent loss of life".  Alcor then argued that Dora couldn't possibly be dead, because her loss of life was only temporary.   To prove  the latter, they assembled an impressive array of scientists, one of whom was Hans Moravec, a leading AI scientist at Carnegie Mellon.

Mr. Moravec was simply supposed to present an argument to the effect that, yes, Dora would some day return to life.   That he did.    But he couldn't help himself, and didn't stop there.   He went on to argue that Alcor's business was quite crude: returning human remains to life was probably a waste of time, given that biological materials would just deteriorate again.   He had a much better idea: Computer technology was improving so fast that one day it would be possible to download the contents of the human brain (into something like a really large USB stick today).   Materials engineering was also improving to such an extent that it would one day be possible to build an artificial body that could last near-forever.   All one then needed to do was upload the saved brain contents into the incredible body and get a much better product.    That would be a post-human (Homo Sapiens v2.0).   

Even with that, Moravec and his friends were just getting warmed up.    They had other neat ideas, such as near-instant travel: Suppose a post-human in Nairobi wanted to travel to London.   All he would do is arrange for a body rental in London and then transmit the contents of his brain (e.g. via the internet or radio waves or ...).     (This, of course, raised other problems that could arise from having multiple copies of a person running around.)

Still, it was admitted that it would be a while before we had post-humans.     But in the meantime, we could still upgrade to a trans-human ("transitory human", i.e. Homo Sapiens v1.5).  That's the sort of thing people like Nick Bostrom are talking about: go to his webpage and scroll down to "Transhumanism": http://www.nickbostrom.com/ 

The industry is now quite busy, and there are "Transhuman Associations" all over the place, e.g.

http://www.uktranshumanistassociation.org/

http://transhumanism.org/index.php/wta/hvcs/

To get back to the question of machines taking over from humans: no need to worry; we will all have been upgraded to post-humans.
« Last Edit: September 23, 2016, 06:57:35 PM by MOON Ki »
MOON Ki  is  Muli Otieno Otiende Njoroge arap Kiprotich
Your True Friend, Brother,  and  Compatriot.

Offline Empedocles

  • VIP
  • Enigma
  • *
  • Posts: 823
  • Reputation: 15758
Re: What does machine learning mean to you?
« Reply #9 on: September 23, 2016, 06:14:06 PM »
Less than 50 years ago, it was complete sci-fi that we would be here, in different corners of the world, chatting like this. It was unimaginable, except to the fans of series like Star Trek TOS etc.

Yet today, the naysayers are using sci-fi technology to laugh about what may come about in the future. On the other hand, the leading minds who helped open up the internet and the computer revolution are themselves taking the possibility of the rise of the machines seriously.:


.

Maybe they're wrong.

But I'd rather err on the side of caution.



Offline Empedocles

  • VIP
  • Enigma
  • *
  • Posts: 823
  • Reputation: 15758
Re: What does machine learning mean to you?
« Reply #10 on: September 23, 2016, 06:22:25 PM »
Then the only thing I could advise you is to read Nick Bostrom's Superintelligence: Paths, Dangers, Strategies

I don't know if you have ever heard of the Alcor Life Extension Foundation: http://www.alcor.org/BecomeMember/scheduleA.html  It's an organization that, for a fee, freezes people until the day when technology has improved enough to bring them back to life; they have been doing roaring business for decades.  Their fees have gone down quite a bit, but it still costs much less to freeze just the head than to freeze the whole body.

Back in the 1980s, they got into trouble when they were accused of homicide.   Apparently, if just the head is to be frozen, then it is best to cut it off right after death or, even better, just before death---to avoid information loss.     The "trouble" was that someone had reported to the sheriff that they had done the latter, which would indeed be homicide.     A little circus then ensued, with the sheriff trying to find the head and Alcor constantly moving it.    But Alcor never denied cutting off the head, so the view was formed that their might be a basis for a charge of homicide.   Alcor was asked to explain itself.

Alcor had this to say: Even if they had cut off the head at the alleged time, there was no basis for a charge of homicide.    That might seem absurd, but it wasn't entirely so; their argument relied on how "death" was to be understood.   That was critical, given how things are defined:

Quote
homicide: the deliberate and unlawful killing of one person by another;

killing: an act of causing death, especially deliberately;

death: the action or fact of dying or being killed;

One can see the circularity; so the matter of just what "death" is became an issue.  All agreed that it could be defined as "a permanent loss of life".  Alcor then argued that Dora couldn't possibly be dead, because her loss of life was only temporary.     To prove  the latter, they assembled an impressive array of scientists, one of whom was Hans Moravec, a leading AI scientist at Carnegie Mellon.

Mr. Moravec was simply supposed to present an argument to the effect that, yes, Dora would some day return to life.   That he did.    But he couldn't help himself, and didn't stop there.   He went on to argue that Alcor's business was quite crude: returning human remains to life was probably a waste of time, given that biological materials would just deteriorate again.   He had a much better idea: Computer technology was improving so fast that one day it would be possible to download the contents of the human brain (into something life a really large USB stick today).   Materials engineering was also improving to such an extent that it would one day be possible to build an artificial body that could last near-forever.   All one then needed to do was upload the saved brain contents into the incredible body and get a much better product.    That would be a post-human (Homo Sapiens v2.0).   

Even with that, Moravec and his friends were just getting warmed up.    They had other neat ideas, such as near-instant travel: Suppose a post-human in Nairobi wanted to travel to London.   All he would do is arrange for a body rental in London and then transmit the contents of his brain (e.g. via the internet or radio waves or ...).     (This, of course, raised other problems that could arise from having multiple copies of a person running around.)

Still, it was admitted that it would be a while before we had post-humans.     But in the meantime, we could still upgrade to a trans-human ("transitory human", i.e. Homo Sapiens v1.5).  That's the sort of thing people like Nick Bostrom are talking about: go to his webpage and scroll down to "Transhumanism": http://www.nickbostrom.com/ 

The industry is now quite busy, and there are "Transhuman Associations" all over the place, e.g.

http://www.uktranshumanistassociation.org/

http://transhumanism.org/index.php/wta/hvcs/

To get back to the question of machines taking over from humans: no need to worry; we will all have been upgraded to post-humans.

You're right, the AI industry is really interesting and keeps changing on a daily basis.

I remember when I first read a story by Larry Niven (A World Out of Time) and got interested in Alcor. Still open to debate so I really can't say anything about it.

I'm sure you know of Ray Kurzweil who is a massive proponent of Trans humanism, although personally I think his timeline is a wee bit optimistic.

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #11 on: September 23, 2016, 06:31:24 PM »
Fascinating case, thanks for sharing.

It reinforces the notion some scientists are permanently lost in fantasy. Those Carnegie fellas should consider publishing comic books where cryonics are often featured and since most of what academics do is publish nonsense anyway. Cryonics don't work. It's called being frozen to death. A cult thing. Freezing temperatures have decay rates. Why would anyone want to return back to being a baby? Circle of life should be respected. I'd imagine whatever spirit comes back to that body wouldn't be Dora.

Best chance is to summon a deceased spirit or invent spectacles to see parallel dimensions like the spirit world because there is evidence of such invisible life forces. I think such research efforts are hampered on the supernatural as a deterrent against mass suicides.

At present science has managed to teleport a beam of light. Maybe it could work for humans oneday and if so when it breaks apart our atoms then why not transport to other dimensions like the afterlife.

Offline MOON Ki

  • Moderator
  • Enigma
  • *
  • Posts: 2667
  • Reputation: 5780
Re: What does machine learning mean to you?
« Reply #12 on: September 23, 2016, 06:42:54 PM »
I remember when I first read a story by Larry Niven (A World Out of Time) and got interested in Alcor. Still open to debate so I really can't say anything about it.

For what it's worth, they now take clients outside the USA---for an extra $10K.    That might seem steep, but consider the costs of having someone fly to Nairobi, sit by your bedside and cut off your head at the right time, and then freeze it for a hurried trip back to the USA.

Quote
I'm sure you know of Ray Kurzweil who is a massive proponent of Trans humanism, although personally I think his timeline is a wee bit optimistic.

Yes.  Haven't read much of him lately, but I did note his optimism. 
MOON Ki  is  Muli Otieno Otiende Njoroge arap Kiprotich
Your True Friend, Brother,  and  Compatriot.

Offline MOON Ki

  • Moderator
  • Enigma
  • *
  • Posts: 2667
  • Reputation: 5780
Re: What does machine learning mean to you?
« Reply #13 on: September 23, 2016, 06:53:22 PM »
Fascinating case, thanks for sharing.
It's called being frozen to death.

Depend on what you mean by "death".    Alcor would include you among  those who seriously misunderstand what they do.  The popular notion is that they freeze dead people, to be later restored to life.    Far from it.  According to them, what they do is

Quote
The Alcor Life Extension Foundation is the world leader in cryonics, cryonics research, and cryonics technology. Cryonics is the science of using ultra-cold temperature to preserve human life with the intent of restoring good health when technology becomes available to do so.

They respect the law and accept the notion of "legal death", but at best that is to be understood as a "temporary loss of life".

Quote
I'd imagine whatever spirit comes back to that body wouldn't be Dora.

I take it you mean "the old Dora".      Merely restoring the original was exactly the sort of thing Moravec considered crude: Why bother with that when you could  modify the downloaded brain contents and get an even better person?

So, on:

Quote
Why would anyone want to return back to being a baby?

No need for that.    You could return as the person your parents always hoped you would be.    In fact, prior to the brain-downloading and permanent-body-construction, you could submit detailed specifications for the new-and-improved you want to be returned.    Or have your loved ones specify the you (new-and-improved) they want back ... Alcor now accepts "third-party arrangements", for an extra $25K or $50K.
MOON Ki  is  Muli Otieno Otiende Njoroge arap Kiprotich
Your True Friend, Brother,  and  Compatriot.

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #14 on: September 23, 2016, 08:22:50 PM »
It's quackery advocated by fringe cults out to make money. Its banned in advanced directives and shunned by reputable medical communities.

The baby analogy is about life cycles. There is a time to be a baby, time to be a teen, time to die and just move on.

Does SciFi even address the nature of existence, death, truth, time, those higher concepts fundamental to being human? The best it can offer is the notion of extending life but to bring back the dead, these are age old unwritten codes in the universe around before science, religion or history. It's in the dirt, the stars, that would require permanent time reversals and that isn't possible in theory and not even with machines that are coded in time stamps unless one can time travel forwards while still alive. The closest thing to reversing death is to prolong life or travel forwards through time. Not the idealistic heresy of consistently bringing back to life dead things. Even by simple logic it doesn't make sense. There are too many inherent problems concerning zombies aka post-human that's not in line with recorded theories by the best minds in history since Socrates.

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #15 on: September 23, 2016, 08:31:07 PM »
Minds like Hawkins should stick to what they're specialized in: numbers. Some of the stuff he says is like Ben Carson running for president: naive and clueless.

Offline MOON Ki

  • Moderator
  • Enigma
  • *
  • Posts: 2667
  • Reputation: 5780
Re: What does machine learning mean to you?
« Reply #16 on: September 23, 2016, 09:26:08 PM »
The baby analogy is about life cycles. There is a time to be a baby, time to be a teen, time to die and just move on.

I got that.  What you don't seem to appreciate is that trans-humanism and post-humanism is also about choices.      There will be those who at 40 or 80 or 100 decide that they have had enough.   They'd like to check out right now.  So long, and thanks for the fish.    And then there are those who at 105 might feel that they are just getting started and could do with another 10,000 years.  A great deal of trans-humanism and post-humanism is about unwanted, undesirable, unwelcome, avoidable .... "death".   (Note the quotation marks.) Surely, there couldn't be a better example of "technology empowering people".   So it's unhelpful to think about a time for this and a time for that ... where did it get that gloomy fellow in Ecclessiastes?

Quote
Does SciFi even address the nature of existence, death, truth, time, those higher concepts fundamental to being human?

First, we aren't talking about ScFi; the guys in the Alcor case, for example, have all had detailed scientific proposals for the "going forward".   Second, these are vague, ill-defined, and unhelpfully backward notions.   I have mentioned "death" above and Alcor's view on "legal" death.     Supposedly, the law is one one of the hallmarks of civilization.     But if you look at how law in most place defines "death", then, to the extent that it does, it's "when a government-approved type (e.g. registered doctor) declares it to be so".  But how does he/she really know that the person on the slab is really dead?  And even by those "standards", don't we know of a case where a person was declared dead only to have him or her sit up and ask for a cup of tea?

A starting point for trans-humanism and post-humanism is that this primitive notion of dying is exactly what is stopping the species homo sapiens from achieving its real potential.   Why die at all, if you don't care to?   Free yourself!

Quote
There are too many inherent problems concerning zombies aka post-human that's not in line with recorded theories by the best minds in history since Socrates.

Well, we have better minds now, and they are working on it, which is why the future looks so bright: Die when you are ready to.  Don't  be the person you are when you could be the person you imagined you would or could be.   

Think about this: There was time when the core of procreation involved a man clubbing a woman on the head and then dragging her back to the cave for some in+out.    But today one doesn't even have to be around to have progeny---sperm, egg, embryo, freezing.    The next step should involve  an artificial womb and the use of technology to cut down from 9 months to 1 month.
MOON Ki  is  Muli Otieno Otiende Njoroge arap Kiprotich
Your True Friend, Brother,  and  Compatriot.

Offline MOON Ki

  • Moderator
  • Enigma
  • *
  • Posts: 2667
  • Reputation: 5780
Re: What does machine learning mean to you?
« Reply #17 on: September 23, 2016, 10:00:21 PM »
I'd imagine whatever spirit comes back to that body wouldn't be Dora.

While poking around Bostrom's website, I found what might be a (partial?) answer for you.    Go to the paper entitled "Why I Want to be a Posthuman When I Grow Uphttp://www.nickbostrom.com/

Section 6 is on "personal identity": after upgrading to a "post-human" are you the same person that you were before?

(Note that his notion of what a constitutes a post-human is the sort of thing Moravec would consider crude and limited.   It is closer to Homo Sapiens Version 1.5.)
MOON Ki  is  Muli Otieno Otiende Njoroge arap Kiprotich
Your True Friend, Brother,  and  Compatriot.

Offline veritas

  • Enigma
  • *
  • Posts: 3353
  • Reputation: 4790
Re: What does machine learning mean to you?
« Reply #18 on: September 24, 2016, 08:32:42 AM »
What you don't seem to appreciate is that trans-humanism and post-humanism are labels with no history, evidence or data to back up it's existence. They're labels referred to by SciFi cults and occasionally dreamy confused academics. Trans- post- are in reference to collective phases to describe a particular movement like post-modernism. However trans- isn't used for this because there's no agreed upon condition for transitional states when it comes to movements.

Existence, death, time etc. have volumes of theorists devoted to it. Just because there are technological revolutions today, doesn't mean any of these achievements have added to the knowledge database. Science is mostly hit and miss discoveries procedurally/observation, has it advanced human knowledge? Just because I can turn on electric lights, doesn't mean I understand the property of lights better. Probably better to observe a naked flame.

Choice isn't some individual self thing like psychologists and the media like to parrot. Choice isn't just privy to humans. Choice is embedded in the whole architect, eco-system, it's contextual, embedded in time, history etc. Extrapolated with destiny and other heuristics too complex to shoulder by just individuals alone.

Euthanasia is controversial but I'd say wise to er on the side of caution. Often when one considers euthanasia they are sick and often times don't want to trouble family members or don't want to be in pain anymore. It's illegal for many reasons because it's no different to murder. On the flipside choosing to take one's life is called suicide and not some responsible decision. When one says "why I want to be posthuman when I grow up" they are suffering from delusional psychosis... idealizing suicide, driven away by romanticized SciFi notions that are downright false. But hey, for publishers, it sells. They are terrified of death or are sick of their shitty life- that's the underbelly- no amount of BS about I want to be an alien changes their detrimental mental state. One should face their fears like death and peacefully accept it.

IVF freeze sperm and eggs but only for a short while. These tissues are constantly updated or abused I might say by replacing stale eggs with healthy stranger's eggs. Sperm banks, DNA centers, and so forth are racked with problems, hard to lay accountable because not many can tell if there's been a manipulation unless their child was black to white parents or something that obvious.

It's rather dreamy to think one can remain immortal like Dorian Grey for so and so years without consequences. The best science could do at this point is to prolong youth. I think they've achieved this somewhat at the tissue level but a whole body would require constant organ harvesting and other nefarious means that they do in places like France. When it gets to the latter, it becomes an ethical issue, a human rights violation, unequal privilege, classism, elitism, and a modern form of hedonism, uncharacteristic of a healthy society. Encouraging this culture as per history has shown aka Hitler's academics & Aryan race- decimate nations, destroys civilisations.

Stop reading SciFi, it's not healthy.

Offline Empedocles

  • VIP
  • Enigma
  • *
  • Posts: 823
  • Reputation: 15758
Re: What does machine learning mean to you?
« Reply #19 on: September 24, 2016, 09:49:49 AM »
What you don't seem to appreciate is that trans-humanism and post-humanism are labels with no history, evidence or data to back up it's existence. They're labels referred to by SciFi cults and occasionally dreamy confused academics.

Stop reading SciFi, it's not healthy.


Excuse me but teeheehee.  :D

500 years ago, what history, evidence or data to back up