/sci/ - Science & Math

>>6294658
>Most of them are incompetent
[Citation Needed]

>their work can be easily be automated
[Citation Needed]

>>6294658
*Diagnosis* can be easily automated. You still need someone to check what the symptoms are.

>>6294676
When you know the symptom everyone can get a diagnosis.
The problem are the fuck ton of diseases with the same symptoms and the fact that not everyone experiments the same symptoms with the same disease.
Now things like surgeries, those could be automated.

>>6294665
citations are also needed to prove otherwise, thus rendering your post invalid

>this is part 2 of a guest series written by a rich californian technocrat
>in part 1 he laid the groundwork by describing how california's artificial "intelligence" products are ready to replace people's jobs despite embarrassing accuracy on trivial tasks for average human beings
>it's more than curve fitting, we swear
>in part 3 he will describe how california's products are ready to replace teachers and further privatize education

like no shit he's a technology investor he's going to tell you we should invest in technology. these are THE 2 big areas that california has decided to monetize next because there's obviously a huge amount of funding for healthcare atm and MOOCs have opened up the opportunity to further legislate privatization into education via austerity efforts

>>6294684
like you might as well have linked a dipshit article on how "causation is dead and correlation is sufficient" you fucking google shill

>>6294665
>Most of them are incompetent
You can get into medschool with an art degree. With that alone should be sufficient, but also the disciple consist in juts memorizing books and no human alive can learn all the information that has come out in the recent years.
An automaton would be more efficient than any human.

>>6294681
>surgeries, those could be automated
You've just (tried) to say that diagnosis is difficult because everyone is different, and then went on to say surgeries are easy? Because everyone is identical on the inside right?

>>6294685
No, the burden of proof is on the claim maker. I don't have to prove you wrong, you have to prove you're right.

>>6294698
>You can get into medschool with an art degree
You can, but how many have? Also, MD's with an art degree went through the same schooling everyone else did in med school, what would prevent them from performing at the same level?

>no human alive can learn all the information that has come out in the recent years.
Very true. Computers should definitely help doctors, but completely replacing a highly trained MD with google probably wouldn't work out too well.

>>6294745

surgeries are about as hard as automated driving (a little bit harder but not by much)

Whether a job can be automated depends upon how much creatitvity is involved.

There are certainly many clinical functions that can be automaateed.

I don't believe they will be fully replaced for a while however; not because of limitations in technology (surgeries are already conducted from remote locations by doctors thru use of robotics).

But because of the numer of variables that must be considered for any given illness we recognize today.
Our knowledge of pathologies which affect humans remains very incomplete. And the human body contains many closely interconnected ad complex biologic relationships. Medical science is still relatiely primitive in its reductionist approach to diagnostics and treatment, given that fundamental causes of many diseases are not completely understood.

I don't think we can establish reliable algorythms that take into account all necessary measurands for treatment of any given condition. Even if we could account for all the nuances human intellect allows over AI at this stage, patientts will feel as though their condition may be unique and will want a human to work with. They will not feel confident in a 'one size fits all' algorythm because they will feel it lacks empathy and has incomplete information.

>>6294685

nah nigger fuck you either you cite, reformat, or get back to x.

>>6294698
>no human alive can learn all the information that has come out in the recent years.

You could say that about any STEM field, though. Nobody knows everything offhand. They don't need to. They just need to know where to look.

>>6294759
> the burden of proof is on the claim maker
so since you don't bother with proving most of them are competent, you don't claim such a thing.
Good to know I was right. They are mostly incompetent

>>6294681
Do you even Watson? Diagnosis already has been automated.

>>6294794
Fine, if you need me to prove that the majority of doctors are qualified to be doctors (which is the logical assumption) here:
http://www.ama-assn.org/apps/listserv/x-check/qmeasure.cgi?submit=PQRS2014

That site ranks doctors and facilities based on different measures.

>>6294813
What are you talking about. Proving =/= claiming. Nice circular logic, though.

>>6294760
what's your evidence for that belief

>>6294658
Most doctors, except surgeons but they're gonna be replaced as well, are fucking useless.

CS is automating everything and eliminating all other jobs.

If you're not in CS:

- You will never have a job for life
- You will never become rich
- You will always worry the next machine or AI that will do job better than you
- Your education will be useless
- You will never live in utopia
- Even math proofs are being automated and nothing is sacred

All because CS people and all these CS billionaires want more money for themselves.

>>6294857

>>6294658
CS majors do not belong on /sci/
Medicine/Biology does not belong on /sci/

GTFO!

>>6294658
>ITT: people who don't know how a hospital/surgery/diagnosis works
never change /sci/

>>6294862
>Something is done on computers
>Obviously all credit should go to undergrad CS tards

>>6294808
>Read something on wikipedia
>Assume the application and functionality are limitless

This is why CS majors are mocked on /sci/

>>6294936
>implying other fields have the skills necessary to make any of this shit happen.
>implying CS people don't comprise 99% of the workforce of companies that are making this shit happen
>implying that CS people aren't at the top of all of the important companies like Google etc

sour grapes anon… sour grapes are your problem. stay delusional.

>>6294862
yeah but once computer scientists develop a general AI, it will undoubtedly have the capacity to be better at coding than humans.

>>6294963
Computer science is just applied math. If math proofs are able to be COMPLETLY automated, then so is any CS related progress.

Although you have no proof of everything relevant (as in shit we dont already know) is able to be developed the way you say it can, even if you are right, you are just the guys that are doing the work so the rest of us can kick back and relax (and in general employ ourselves in other shit). Not saying that is a bad thing (assuming you are right, for which you have provided no proof), I am saying you are the one taking the cost up the butt.

Good Job, CSfags.

>>6294992
>Computer science is just applied math.


uh oh.. sure thing kid. that's a completely meaningless and vapid statement.

point is that a math graduate has none of the skills that CS people have and are necessary to create these technologies.

>>6294963
>STAY MAD
>U JELLY
>BUTHURRTT :DDD
>THANKS FOR THE FREE MONEY
Truly subhuman

>>6294760
Like, naw dude. 4th year med student here. Robots are great for helping surgeons. we love 'em. But what does a robot do when the patient goes into cardiac arrest in the middle of a procedure? what happens if a wall of a blood vessel ruptures and the patient starts bleeding out on the table and oh shit cardiac arrest and fuck pulmonary aspiration? Having *people* there to deal with shit like that won't be replaced in the near or even distant future. And what about trauma surgery? Some dude comes in w/ multiple gsw's, shit torn up all over the place, there isn't a robot around that can deal with that clusterfuck in time to save someone's life. 100 years from now? who the fuck knows. But until you *actually* know everything that goes on in the OR, keep your ignorance to yourself. It's embarrassing.

>>6295032
Does not change the fact that even if what you say is true (btw you still cant prove that every possible knowledge generating endeavour can be automated), said automated approach will be able to develop CS related inovations faster and better than CSfaggies. So either you are in the same position as everyone else (only that you are the only only one doing the grunt work) or you are full of shit. Your choice.

>>6295055
>>6295055

So much jelly.

>>6295052
>>6295055

NEETs… NEETs everywhere.

>>6295063
And I will consider your necessity to insult and resort to fallacious thinking as your graceless way of admitting defeat. Stay mad.

>>6295066
Jokes on you. I inherited a shitton of money and I currently study suff out of pure preference.

>>6295067
>Stay mad

I'm not the one who's mad. Quit projecting. I fucking love the filed that I'm in. This past summer my salary as an intern was higher than yours will be 5 years form the time you graduate.

Keep on hating and stay mad! I'm sure your own condition will improve by hating on others. Lulz.

>>6295071
>salary
Pleb.

>>6295070
>I'm a millionaire lie retort


Oh yeah… hahahahaha

>>6295078
Believe it or not, does not change the fact of my previous post. Thanks for either taking it in the butt for the rest of us or alternatively being full of shit instead of us.

>>6295054
Pretty much this.

My brother is doing Med and I'm a SoftEng which happens to be involved in healthcare solutions and does that shit is a mess. When you say a 100 years, I just say HA! When we have near perfect voice recognition, natural language processing, artificial vision, fuzzy logic AIs and all the right sensors then we would be 100 years from automating that shit

>>6294808
A man comes into the ER with sever stomach pain. He states it's been going on for weeks. He feels better at night when he's lying down to sleep. The pain starts again about when he gets to work in the morning. He's been tested for everything from ulcers to cancer to you name it. He's not a stressed out guy. Not an alcoholic, none of that. Nothing wrong. The machines don't see shit. But his problem is one that any attentive doc would figure out. What's your diognosis? What would a robot do in that instance?

>>6295085
>more lies

please, tell us more about your wealth. is this why you're wasting your life on shitposting on 4chan? hahahah

>>6295092
yeah i was being optimistic about the 100 years. I don't like to make tech predictions past 50 years or so. I mean, who the fuck knows....it would take an amazing conglomeration of shit to get rid of a good surgical team.

>>6295100
Hey, I was even going on about my wealth (despite, you know, having more wealth would give me plenty of time to "waste"...and I do). I said that it does not matter if you believe me, the facts I stated in my previous posts (>>6294992 and >>6295055) already fucked you up for me.

>>6294936
lol, Watson is already being used at real hospitals and is better than doctors. stay buttmad, your job will be replaced soon enough.

>>6295124
>I said that it does not matter if you believe me

Of course it matters. It shows how ridiculous your lies are getting and how envy has driven you insane. Seek help before you go on a violent rampage. It's not too late.

>>6295133
see:
>>6295096

>>6295143
How should I know what a computer would do in that situation? (Or you for that matter)

>>6294658
its entirely possible to have robotic doctors and surgeons (we already have robotic surgeons). however, people still need to go to school and get educated to have the knowledge to do research and develop better science

>>6295146
the moral of that story was that the guy was wearing his pants and his belt too tightly. There's no diagnostic test to figure that out. People are stupid, they do stupid shit, and they lie.

>>6294963
>>implying CS people don't comprise 99% of the workforce of companies that are making this shit happen
>people

Those "people" are absolutely not CS majors.

>>implying other fields have the skills necessary to make any of this shit happen.

Coding Java GUIs, learning vim/emacs, plagiarize code, placing all the work on one team member, and cheating on algorithms exams "skills" are worthless in the real world. Your undergrad degree is worth it's weight in shit.

>>6295164
>Those "people" are absolutely not CS majors.

Of course not, they're all pure mathematicians and theorethical physicists, rite?

>bullshit topic

Another idiot who has no idea about CS curriculum. Are you as rich as the other idiot or ate you just ignorant?

>>6295138
Even if I was an envious and insane man, would that make any less true what I said? You keep using ad hominems and fail to see how you are the joke here. Please read again my postrs untill you realize your sad sad position.

Also...enjoy your salary hahahahahaha

>>6295171
>Thinking undergrads do any cutting edge work
>Thinking those were topics and not the only things CS failures remember after leaving school

>>6294658
to cure your autism

>>6295032
>skills
>that CS people have

Pissing in bottles?

>>6295096
His wife is drugging him in his sleep?

>>6295194
> thinking that cutting edge work is what's needed right now
> thinking that we don't need algorithm kiddies to build all our shit

>>6295251
that would have shown up in a tox screen.
see:
>>6295157

>>6295245
>autistic as fuck

you can repost that pic million times and it still won't make it true.

enjoy unemployment and yes, you're projecting.

Machines will not replace doctors in the next 50 years, however they can reduce the need for doctors by a huge amount. Which is a good thing, as we have a worldwide shortage.

You still need real people for those times when the expert system fails to find the problem. But for 95% of the cases, a well programmed machine can determine the correct treatment and save the doctors' time.

>>6294658
true dat, most of the time after they leave there looking at flow charts.....this this and this he might have this well nope ok then it might be this well nope......yeah computers could do alot better

>>6295360
yeah realtalk machine learning is important to the future of medicine and medical is where all the CS people are looking next and if you want to monetize machine learning here is the One Golden Rule: you need an application scenario that's ok with 5-10% failure rates

the meaning and importance of failure rate is really different among different scenarios. there are scenarios where 90% is trivial to achieve in some fundamental sense but humans achieve 99.9..% (facial recognition is one). then there are things that humans suck at just as bad as machine learners, and there are things that humans can do much better but it's not worth the time and cost when 90% is sufficient

the prevalence of ML right now has more to do with advances in accessability and "how to" then advances in AI itself. AI is still pretty primitive

OP

In the states the care you receive for your symptoms is often decided by your insurance company, that company simply checks to see if your claims (doctors report of claims) meet minimum requirements for the requested treatment, if they do not then said requested treatment (requested by your physician) is denied.... My brother works for a provider signing off or denying doctors requests for procedures/tests based purely on paperwork provided to him, he can do 100's a day (yes he is a doctor, internal medicine)

All in all I agree doctors are with OP doctors are functions the same as your mechanic only usually smarter

>>6294763
>Medical science is still relatiely primitive in its reductionist approach to diagnostics and treatment, given that fundamental causes of many diseases are not completely understood.

This. Ppl think we're advanced, but actually medicine is still "medieval" in terms of diagnostics and treating very common illnesses.

I'm laughing my ass off when I hear all these uber-transhumanists/singulatarians etc talking about the future as the pie in the sky when we can't even cure common diseases like liver disease, arthritis, and others. Not to mention that the drugs used to treat them have so many secondary effects, basically if you have a liver disease and need to take anti-inflamatory drugs against arthritis problems, you're screwed, because a long-term intake of anti-inflamatory medication messes with your liver.

That's the state of medicine today...

>>6295400
What is the name of that comic?

>>6295639
Saturday Morning Breakfast Surreal

>>6295639
Smbc

Not that I think they should be replaced by AI

But alot of GP's are fucking useless (they ask symptoms, completely disregard this and then order a barage of tests, then make a wrong diagnosis and send you to a specialist) particularly foreign ones from countries with shit tier education.

>>6295708
Not only that but the medical science is changing daily and new research is coming out all the time and a GP just simply cannot keep up with all of it. Even a specialist can't know everything about his field of practice and can't integrate all of the best practices and make decisions as well as a machine.

That's why Watson is going to decimate medicine.

http://bits.blogs.nytimes.com/2012/10/30/i-b-m-s-watson-goes-to-medical-school/?_php=true&_type=blogs&_r=0&pagewanted=all

I'd rather get diagnosed and get a suggestion from Watson than from a real doc… after Watson goes through a lot of testing of course.

>>6295483

Bio-medical science is progressing quite well though.

>>6294685
your and idiot

look here

http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCwQFjAA&url=http%3A%2F%2Fstpk.cs.rtu.lv%2Fsites%2Fall%2Ffiles%2Fstpk%2Fmateriali%2FMI%2FArtificial%2520Intelligence%2520A%2520Modern%2520Approach.pdf&ei=lJjYUvjTB-Of2QWo04HoBw&usg=AFQjCNGWuwp4bDR-YTsUKSSmKHPmcC7cPA&sig2=BHk7ufBPssRTx79YYc5lYw

>>6295715

Doctors earn experience though. A /good/ doctor can make very crafty deductions based on past patients, medical knowledge, patient history etc

>>6295054

engineer here

any automated process will be precise enough to not have this sort of thing happen in the first place, so explain why there needs to be so many people involved in a surgery when it can be an automated process that's safer for the patient. automated systems don't suffer from fatiuge, don't need to eat or sleep, and will be that much more repeatable that a turdlord surgeon with an overinflated ego

>>6295754
>(c) 1995
>using a 20 years old book as reference for one of the fastest evolving field of science
That book doesn't even MENTION Support Vector Machines, (Principal) Component Analysis, Makrov Chains / MDF or Sparse Representation, which have been some of the "hot" topics of machine learning in the last 5-10 years.

>>6295755
so can AI. and AI can create even better deductions.

http://www.rawstory.com/rs/2013/02/12/ai-system-diagnoses-illnesses-better-than-doctors/

>An artificial intelligence system developed by researchers at Indiana University can diagnose illnesses and prescribe courses of treatment significantly better than a human doctor, the university said Monday.

>Using a computerized decision making processes similar to IBM’s wiz computer “Watson” that won the game show “Jeopardy,” researchers plugged in big medical data sources and tasked it to simulate treatment outcomes for 500 patients, most of whom suffered from clinical depression and at least one other chronic condition, like high blood pressure or diabetes.

>Using data from actual patient-doctor treatment sessions, computer science assistant professor Kris Hauser and Ph.D. student Casey C. Bennett compared real-life outcomes to simulated treatment regiments and found their computer was nearly 42 percent better at diagnosing illnesses and prescribing effective treatments than human doctors.

Human doctors are going to be a dying breed in less than 15 years.

The general topic of this thread is statistical vs. clinical decisions. There is a wealth of evidence on that which shows that mostly algorithms are better and when they are not, they are equal. In only very few cases are humans better.

Society has still not adopted to these findings, even tho they make for HUGE savings in all kinds of areas.

Key paper: http://www.ncbi.nlm.nih.gov/pubmed/10752360

The process of making judgments and decisions requires a method for combining data. To compare the accuracy of clinical and mechanical (formal, statistical) data-combination techniques, we performed a meta-analysis on studies of human health and behavior. On average, mechanical-prediction techniques were about 10% more accurate than clinical predictions. Depending on the specific analysis, mechanical prediction substantially outperformed clinical prediction in 33%-47% of studies examined. Although clinical predictions were often as accurate as mechanical predictions, in only a few studies (6%-16%) were they substantially more accurate. Superiority for mechanical-prediction techniques was consistent, regardless of the judgment task, type of judges, judges' amounts of experience, or the types of data being combined. Clinical predictions performed relatively less well when predictors included clinical interview data. These data indicate that mechanical predictions of human behaviors are equal or superior to clinical prediction methods for a wide range of circumstances.

>>6295786
I'm interested. How are PCA used in machine learning? Just to find patterns/data reduction? I've had an interest in factor analysis since I started reading psychometrics. Brilliant idea of Galton/Spearman there.

>>6295827

You talk about 'simulated treatments'... as if it is a certainty that it will hold in reality.

Models can be wrong, and hence your inference, too.

The key is to gather a lot of data. And even then, person A is not always the same as person B. The ideal case would be that we could have each patient have hundreds of diseases, many times over and then formulate a 'very reliable' model of reality.

>>6296427

PCA is used to reduce data dimensionality
This can then be used to make better decisions on how to model your solution and hence a better performance.

For instance, if you have a number of portrait photos of say 10 persons, and want to figure out for a new portrait photo to who it belongs, then you can use PCA to reduce the images from a couple thousand dimensions to say 10-20. On this reduced dataset, you can use a simple algorithm like kNN and identify very accurately who the person in the photo is.

>>6296434
What data dimensions would you use from the image to begin with?

>>6294862
Don't know if you remember 1999, but when a new automation tech comes about, the first thing to be automated is CS jobs....

>>6296440
IT and CS are not the same thing.

Computer advances make CS skills more valuable. They may cause you to move on to greater projects, but not to be unemployed.

>>6296427
Let's say you use an Artificial Neural Network for face recognition. This network will try to create a model of the human face. However, this will just be a big mess of neurons and will be hard to teach.

The human face is not just a random blob of pixels, though. It has certain coherent features, like eyes, nose, mouth, ears and so on. You can use PCA to identify these features, and try to group neurons together that are responsible for recognizing these various parts of the face.

It's also used to throw out irrelevant dimensions, as >>6296434 said.

>>6294658
FUCK YOU YOU LITTLE GAY NIGGER YOU KNOW NOTHING ABOUT MEDICINE PLS STOP TROLLING MY SCI YOU DICKWAD

>>6296583
Future hobo detected.

>>6294862
this anon is correct.

>That robots, automation, and software can replace people might seem obvious to anyone who’s worked in automotive manufacturing or as a travel agent. But Brynjolfsson and McAfee’s claim is more troubling and controversial. They believe that rapid technological change has been destroying jobs faster than it is creating them, contributing to the stagnation of median income and the growth of inequality in the United States. And, they suspect, something similar is happening in other technologically advanced countries.

http://www.technologyreview.com/featuredstory/515926/how-technology-is-destroying-jobs/


>Report Suggests Nearly Half of U.S. Jobs Are Vulnerable to Computerization
>Oxford researchers say that 45 percent of America’s occupations will be automated within the next 20 years.

http://www.technologyreview.com/view/519241/report-suggests-nearly-half-of-us-jobs-are-vulnerable-to-computerization/

>An army of robot baristas could mean the end of Starbucks as we know it

http://qz.com/134661/briggo-coffee-army-of-robot-baristas-could-mean-the-end-of-starbucks-as-we-know-it/

>Even the finest restaurants are serving coffee made with capsules. Have we lost faith in the human touch?

http://www.aeonmagazine.com/being-human/julian-baggini-coffee-artisans/

>Hamburger-making machine churns out custom burgers at industrial speeds

http://www.gizmag.com/hamburger-machine/25159/

http://technoccult.net/archives/2013/10/08/report-47-of-u-s-jobs-at-risk-of-being-automated-out-of-existence/

you can also use left-hand eigenvectors to produce clusterings and weightings/centrality estimates on graphs and relational data (spectral analysis)

in some sense these are all kind of the same application. matrix factorization is very powerful though i think care should be taken to recognize its limitations and not read into it as too fundamental

>>6294862

Automation is now happening at an unprecedented pace. The graph to the left depicts the decoupling between median income and productivity - productivity has continued to skyrocket while little human workers are needed as wages stagnate.

We are not simply overcoming the limitations of our muscles. We are overcoming the limitations of our minds, and this is happening on a scale that is completely unprecedented in human history.

We're seeing that artificial intelligence systems like IBM Watson have sufficient natural language understanding to beat the world Jeopardy champions simply by reading wikipedia. This same technology is now set to eat away at medical diagnosis by being able to better diagnose a patient than a panel of board-certified doctors.

At the same time, we're seeing self driving cars fast approaching, with 6 million driving jobs in the US under threat from this technology. Robotics is also encroaching in factories and buildings, and Amazon's recent purchase of Kiva robots for warehouse automation is a great example of this.

Self checkout counters, automatic instantaneous language translation, e-commerce, legal software, 3-D printing, and artificial helpers like SIRI are quickly becoming ubiquitous.

Soon, Amazon is planning on using drones to deliver packages, and Google is now buying up the most advanced robotics companies on the planet because they too see where things are headed.

Are we entering an era where there is little need for human workers? If so, what are the social and political ramifications of a populace that does not have a means of income?

http://www.cbsnews.com/8301-505124_162-57578162/robots-are-going-to-take-your-job/

http://www.cbsnews.com/8301-18560_162-57601121/are-robots-hurting-job-growth/

http://www.wired.com/gadgetlab/2012/12/ff-robots-will-take-our-jobs/all/

http://abcnews.go.com/Technology/google-buys-boston-dynamics-acquires-robots-play/story?id=21233076

>>6296597
Good. Once we've got AIs and artificial muscles working properly organic people should just lie down and die.

>>6296604
It will both be a utopia (for civilized countries) and the beginning of the extinction of organic humanity. There will be enough automation and robotic self-sufficiency for people to barely need to work, but the need for reproduction will be extremely minimal. Artificial companions or people that have their brains gradually replaced with hardware will steadily outnumber organic people. Things will constantly become more efficient and people will steadily gain more and more freedom as their limitations are removed.

>>6296604
>what are the social and political ramifications of a populace that does not have a means of income?
Well, there's two outcome predicted by the short story.
marshallbrain.com/manna1.htm

Hint: the united states will not become like the australia project.

>>6296604
and let me point out it's no coincidence that the guys in silicon valley are building completely parallel communities with their own campuses, recreation, housing and transport, etc. at the more extreme you have guys like peter thiele who want to live in international waters.

the economic movers in CS are all randian libertarians and will be totally happy with a world where 45% of the US is unemployed "because they're not smart enough" to be white collar technocrats. they're just going to move into their gated communities -- they already have -- and similarly it is no coincidence that military drone tech was the first to be developed and these are now being used for policing throughout the US

the leap from "i employ these people at minimum wage -> i have no responsibility for their wellbeing" to "these people are unemployed -> i have no responsibility for their wellbeing" is a really easy one to make

Have you ever noticed that the three professions that can best be replaced with information technology, just aren't? Medical doctors, instructors and lawyers.

Question: Which geopolitical racial group tends to dominate those professions? Hint: Rhymes with "Jew".

Once you realize who really rules your society, then you start to understand why certain professions are immune to replacement.

>>6296615
see the thing you futurist assholes need to remember is the ending of dune if you like sci-fi so much

all these ultra-privileged space emperors and mentats babel/on and on about their schemes and THE SWORD IS THE MIND for like 300 pages and then just get fucking trampled by fremen muslims and technophilosopher paul is forced to admit his kwisatz haderachness gives him exactly zero power over the future

important lesson

>>6296625
>lawyers

I would prefer that the legal system be made of meat. You can render meat inoperable far more readily than you can a network.

>>6294864
Sorry but it's true.

>>6296604

Stuff it. The "legal software" angle has affected the need for a lawyer by nearly ZERO percent. All trials are conducted by lawyers, often Jewish lawyers, before judges who were also lawyers. They all make sure that they are physically needed in the legal process.

>>6296625
Medical robots require more flexible and dexterous hardware, and true AI to deal with emergent situations. Same with instructors, and this is simply a matter of time before the software to handle these situations IS developed.

Robotic lawyers would be unpopular because people don't want perfectly logical defenses.

>>6296629
Dune is a story anon.

>>6296583
This. And since when did /sci/ became /med/?

>>6296625
>lawyers can be best replaced by technology
man i have heard this stupid idea from lawyers themselves and i have no idea where it started legal arguments are entirely social/subjective/interpretative and computers are terrible at that. law isn't functional

i've studied both CS and law so i'm like, one of the only people actually qualified to talk about this

>>6296629
>ultra-privileged

yeah but that's in a sci-fi book

irl the privileged will always win because they are privileged for a reason and don't have brain problems relating to their genitals

>>6296635
>Dune is a story anon.
so dune is a story but the transhumanist/singularitarian chestnut of "organic humanity becoming extinct" is

what

what would you call that it's speculative fiction it's literally storytelling tautologically

>>6296588
This.

>>6296643
>privilege is a meritocracy
lol

>>6296635

The instructors angle is BUNK. You can RECORD a professor's lectures for an entire year and therefore just REPLACE HIM for the subsequent years, merely playing back his lectures on screens and monitors. NOTICE WELL that that sort of thing isn't happening, or it happens in cases where you can't get cerification for what you learned. What a coincidence, bub.

The education profession is LOCKED DOWN. It has a PRIESTHOOD. And so do lawyers and doctors. And there's ONE racial group that figures prominently in those, and that's JEWS.

Once you learn to see the Jew, you finally see reality. You understand why things happen, and for whose benefit. Hint: Not for your benefit.

>>6296643
>Dune is a story anon.
>yeah but that's in a sci-fi book
like are futurists so far gone that when they start talking about the impending roboapocalypse slash merger of cyborg life they think that's like, serious hard science and we can't learn anything from alternative narratives like dune because it has sandworms which are definitely more goofy than a singularity, sure definitely

>>6296641

No. Wrong. Lawyers are on BOTH SIDES of the conflict in question. So naturally they work to make it seem like they're necessary. When as lawyer wants work, he gathers a client and sues, and that prods the other side to GET A LAWYER, and then it's well and truly ON.

>>6296662
anti-singulitarians are nerds who want to be better than other nerds so they adopt a contrarian view

>>6294936
I thought pampered doctors were mocked not just on 4chan but in the world, as a dinosaur technique.

>>6296654

it is

you just don't want to face the fact that you don't have enough merit

>>6296673
oh my god is that what you actually believe

honest to god you're telling me singularitarians are the "nerd" mainstream and anyone else is a contrarian

oh my god fuck sci-fi forever this genre should be banned

>>6296673
in fact i have a comic for this occasion

>>6296684
i'm in a stem phd program at an ivy not even joking

>>6296689
you mean you have a hack edit of good comic that originally had a more nuanced message
>nice try

>>6296686
Why do you write like a Tumblr person?

the nocaps

the linebreaks

the overwrought flouncing

>>6296690
Yeah, it's never the people at the bottom of the ladder who question their place. It's always the people for whom success is just out of reach. You probably got passed up for something and don't want to face the fact that you're defective.

>>6296693
sir this restroom is not open to the public

>>6296694
so like
a. gonzo stream of consciousness is a respectable art form and personally i think it helps me to be more truthful
and
b. it tends to piss off idiots which i like
and
c.
>bottom of the ladder success defective
lol psychoanalysis what tipped you off hari seldon was it my use of viewtiful joe

>>6296686
oh my god

oh my god

i paid money to post here oh my god

>>6296700
why would you have a 4chan pass

unless you were running a kind of cool eliza chatbot-ML experiment with 4chan as the training pool i've thought about doing that

>>6296701
>a. goony shitposting is 2deep4u
>b. u mad
>c. why doesn't anyone want to hire me, is it because they'll have to make a new bathroom for my whalekin altgendersona

What's wrong with my hand /sci/?

My mom says I'll have arthritis because my middle knuckles are so big, but I don't think so.

Is there a comp sci major or doctor here who can diagnose me?

>>6296708

Your middle finger's fucked, bro.

And no, go to an actual doctor, you retard.

>>6296706
>whalekin altgendersona
actually laughed we're cool

>>6296715
No, we're not.

>>6296708
It looks like it got ran over

go to a doctor

>>6296708
It looks like you've already got some kind of inflammatory joint problem. Maybe it's just not very painful yet.

I was like that. They didn't start hurting until my late 20s. I had digestive problems I didn't do anything about, because I thought they weren't important. When I buckled down, figured out what was bugging my guts, and eliminated it from my diet, my finger joints stopped hurting, straightened out, and slimmed down, along with a lot of other changes in my body.

Inflammatory problems can come from a lot of causes. Some can be fixed by medication, others by lifestyle adjustments. Talk to a doctor, and do your own homework.

There are a lot of jobs that could be automated, but they aren't because then people lose jobs.

>>6296706

why are the most privileged white kids the only ones who give a shit about "privilege"

you don't see chinks or bindis whining about how they got a crappy roll, it's always white young people and very occasionally a milk-with-a-splash-of-coffee brown person

>>6294759
>No, the burden of proof is on the claim maker. I don't have to prove you wrong, you have to prove you're right.

When did science become so fucking lazy? If you have some good arguments against something, even off the top of your head, which you obviously think you do because you're so dismissive, then just blurt it out rather than act like a lazy-brained idiot.

>>6296708
Take it to the vet, you ass

>>6296720
decided by fiat we're cool nothing you can do about it you could meet me at a conference (i get around) and you would never know i was this anon

>>6296733
literally jerry seinfield does this mode of reasoning have a name?

also what's amazing is that i was literally talking about the space-privileged in a ridiculous hyperbole society but it STILL spawned kneejerk meritocratic defense of the status quo by association

dune best book only good sci-fi book completely 100% accurate test of futurist character

>>6296757
Go back to your fucking forum, you toothless clown

>>6296764
and if you love meritocracy so much need i remind you that the entire reason the fremen revolution succeeds is that the space wizards are not as smart as they think they are and the fremen are not as dumb as the space wizards assume them to be

>>6296772
this is my forum i've been posting here like all week

The singularity is communism for people who know C++

As such, it will happen, but it will be monstrous and unendurable.

>>6296777

The singularity already happened, and it's already monstrous and unendurable.

And it's unsustainable, since it was fueled by dirt-cheap petroleum, and that's over, FOREVER.

>>6296782
So you're saying we're going to have to enslave the nogs all over again?

>>6296777
>singularity
>communism
so far as i can tell literally 100% of singularitarians are libertarian objective rationalists so

>>6296690
>i'm in a stem phd program at an ivy

So you are the 'privileged', then? What are you doing to offset that?

Note: Cutting off your penis, while convenient, will not actually offset that.

>>6296788
>tfw an actual privchecking tumblrtard on a hard science board

you're in for quite a ride, my friend

>>6296789
yes obviously. "solutions" i have seen
- make a lot of money donate to optimal charity. common utilitarian solution
- advance technology solve real problems of lower economic class. not many people manage to do this since it's top-down
- advance technology hope it precipitates godbrain singularity worth infinite utility points. funny
- work in education, political advocacy, etc. because activism starts in your community. this is why i work in education

the unfortunate reality is that everything is political and political problems (all problems) are really complicated because they reflect the perspectives of billions of people so if you thought i was going to produce some kind of "i do this and i am redeemed" fair trade thing then slavoj zizek wants 2 have a word with you

>>6296708

First off, need a better picture -- one that doesn't have an interfering shape in the background (because the knob coming out of the machine in the background merges with your middle finger's middle knuckle, making it look diseased).

Second -- arthritis is caused by 30+ years of tense muscular architecture in the relevant joints. Have tense knees, get arthritis in the knees. Have tense arms, get arthritis in the arms, etc..

What you wanna do is stretch and breath exercises.

Learn how to yawn proper.

>>6296791
i don't use tumblr you're just so reactionary that using privilege to describe a space emperor in a sci fi book is enough to shunt this conversation into a simple template that you feel you know your role in

keep telling stories to explain life i'm going to a meeting bye

>>6296795
>this is why i work in education

no child should be allowed unattended anywhere near you, you fucking fruit

hope this helps

>>6296791
also
>hard science board
i'm crying tears

>>6296799
Don't you want your kids to learn about Truly Equal Space Communism ?
???

>>6296777
that can't be true: knowing C++ takes work, and communism is libertarianism for people who hate work

>>6296625
But the professions that are the easiest to replace are manual labor, not white collar jobs.
Factory workers, cleaners, builders started getting replaced by machines 50 years ago.

Doctor is basically a human lookup table. You go see a doc, tell him what bothers you, he thinks about it for a second, looks at you, and gives you a prescription.

Fact is that 90+ of these visits can be automated with AI. Just by telling AI what your symptoms are, AI can quickly come up with a proper diagnosis which will be more precise than what a human can come up with.

Think of the human doctors as librarians and AI as Google search engine. Does anyone even use librarians anymore? Sure.. but a tiny, tiny, tiny fraction of searchers.

RIP Doctors. AI will save a lot more lives than human docs and for much, much less!

>>6297071
You couldn't be more wrong. There are a lot of things that can be really hard to translate into a computer program.
Experience, intuition, practice. You can ask a doctor to write down everything he knows in a pragmatic way, but you won't get half of his expertise. How would you tell a computer that someone looks pale? Or maybe their skin has an orange or greenish tint? What if that's their natural skin color? Are you going to ask the patient to look into a mirror and objectively describe his skin color?

Humans are extremely good at spotting something that's off. This is why we have the uncanny valley in robotics. The tiniest change in the skin's texture and reflective properties can hint an experienced doctor that the affected body part might get inflamed, a slightly different texture can hint to renal failure, another can lead to trombosis, and so on. How will you translate this into code?

> doctors can be replaced by robots

HAHAHAHAHAHAHAHAHAHAHAH

WHO ELSE ARE THEY GOING TI THROW UNDER THE BUS WHEN SHIT GOES WRONG? Corporations that make the robots? Nurses with their $20 annual $3,000,000 med insuance? Nigga please.

>>6297108
>There are a lot of things that can be really hard to translate into a computer program.
>Experience, intuition, practice.

Bullshit. Have you even read posts in this thread? Your nonsense has been debunked numerous times. AI makes better diagnoses and makes fewer mistakes. That will translate into 100 thousand lives saved each year just in the US!

>>6296797
how has no one called this faggot SRS yet, they are literally spouting SRS ideals phrased with SRS idioms

is your fucking equipment even plugged in

>In 2009, the Mayo Clinic—a national integrated group practice—used an ANN program to help physicians rule out the need for invasive procedures by accurately diagnosing patients previously thought to have endocarditis—a deadly heart infection. By decreasing expenses and lowering risk (at a confidence level of 99 percent), the clinic’s program clearly demonstrated its worth to patients and doctors alike. The same year, General Electric developed a software program that could suggest treatment options for patients in real time by recognizing patterns of information in data. Software engineering firms like Artificial Intelligence In Medicine, Inc. continue to develop informatics products for the healthcare industry. And in 2007, the Healthways Center for Health Research in Nashville, Tennessee—a company that helps manage the health of millions—was recognized for implementing AI-based computer models that help target members for the center’s healthcare programs. Healthways’ long-term value has since been attributed to reducing healthcare costs.

>It was only a matter of time before AI would evolve into high-skilled fields like medicine. In fact, medical AI has been on the drawing board for decades, but usability issues and incompatible terminology and language prevented early software programs from gaining the traction needed to significantly affect the industry. What makes AI and medicine such a good fit is the highly structured reasoning methods they share. But some professionals still don’t trust the idea of computers making health decisions for people, and the prospect of swapping out consciences for motherboards may not sit well ethically. But there’s really nothing to fear. With AI counterparts at their side, doctors may actually serve their patients better.

http://singularityhub.com/2010/05/10/the-ai-doctor-is-ready-to-see-you/

Doctors are so fucked. CompSci will neuter them quickly.

>>6297108
>How would you tell a computer that someone looks pale?
computer vision and bayesian inference cross referenced with previous visits.

>Humans are extremely good at spotting something that's off
yes because we've had our entire lives to observe patterns of skin colors and textures. a sufficiently advanced AI would have no problem doing this.

it's funny, you're taking pattern recognition, the problem which basically all of machine learning is built on, and trying to use it as an example of something that humans are more capable of than machines. which they are, but only for the time being.

You do realize that even if AI makes 100% perfect diagnoses (not going to happen ever) and is proven to be almost always better than doctors, most people will still want a regular doctor, right?
The FIRST time an AI misdiagnoses someone, the newspapers will be all over that shit, the people's confidence will be lost and the whole thing will collapse.

These automation and technology reliance threads are hilarious. The fact of the matter is that all of these automation jobs require massive amounts of energy. Since oil companies own everything and will never allow solar energy to reign supreme, it comes down to how much innovation we can accomplish and how long we can sustain it until the oil runs out. I mean honestly do you fuckers think that humanity will choose innovation and technology over comfort and job loss? When it comes down to it we WILL run out of oil and we WILL have to choose what we do and don't want automated. That is asuming we even make it to a cyberpunk esque future and another war doesn't set us back another 50 years.

>>6297129
>most people will still want a regular doctor, right?

LOL… do you even know how technology adoption works? They will do that for the first few years and then once everyone acclimates to the fact that machines are much better at diagnosis and treatment, insurance companies will have a field day with it and start eliminating the need for 80-90% of medical positions… and the people will be fine with it because they're getting better treatment and it's CHEAPER!

>>6297131
>These automation and technology reliance threads are hilarious.

You sound like Nero fiddling while the Rome is burning. Laugh all you want but the jobs (and your job) will disappear before you even turn 40.

>>6297129
that's like saying nobody would have wanted to drive cars because horses were familiar to them. neither are perfect, but one is VASTLY more convenient than the other.

>>6297131
>oil companies own everything and will never allow solar energy to reign supreme, it comes down to how much innovation we can accomplish and how long we can sustain it until the oil runs out
Yeah, that's right. The world's going to end before it changes enough that you'd have to adapt.

Good luck with that theory as a guide to planning your future.

>>6297131
Oil companies will just use their massive jewgold to buy out whatever alternative resource emerges once oil runs out and then sell it at artificially high prices.

>>6297131
>hilarious

Holy fuck, I never thought that automation threads on /sci/ would be worse than the automation threads on /g/ and /pol/ combined.

Look up that famous clinical heart disease trial on 500 south African males, where they tried to predict their likeliness to have heart disease based on 12 lifestyle factors by Logistic Regression. What you need to understand in machine-learning, is that to make good predictions you need to train your data set on a large sample of KNOWN cases.

That means that the INPUTS to our ANN or Logistic Regression or whatever, is 12 life-style factors and the results of 500 guys being opened up and doctors making 500 assessments of whether on not the guy had heart disease or not. The doctors are the ones who provide the diagnoses that constitute the training set that goes in to the algorithm.

>>6295054
Sorry but a med student isn't going to convince anyone whether or not there will be robotics system and artificial intelligences that are fully capable of performing surgery. That's because med students don't know fuck all about any of that.

>>6297178
>The gasoline for this newfangled "auto-mobile" was brought by HORSE CARRIAGE.

>>6295054
>But what does a robot do when the patient goes into cardiac arrest in the middle of a procedure?


You don't get how technology works. Technology works by first going after the low-hanging fruit.… that's the 90% of cases that GPs see every day. People who need heart medications, GI diseases, etc. You know… non-emergency stuff that can be easily mitigated with medications. That's what AI will attack first.

Once AI decimates the need for GPs, robotics will be advanced enough to eliminate the need for the rest of he medical staff.

Technology works in stags but the first stage is going to wreck everything right away. That's how disruptive business work and no industry is immune form it.

>>6297191
what I like are the nerds who are smart enough to whine about 'disruptive' technology but not enough to realize they're basically plagiarising dumbed-down Uncle Ted

>>6297351
>plagiarising dumbed-down Uncle Ted

that guy was scary-smart and way ahead of his time. but we went nuts and started killing people. he would have been a lot more effective if he used a podium and a microphone instead of bоmbs.

>>6297362
>a podium and a microphone instead of bоmbs.

>not a turntable and a microphone to drop lyrical bombs.

>>6297386
>not a tumblr to reblog hilarious memes created by privileged brooklynite hipster shitbags

>>6297126
>singularityhub.com
lol

>>6297436
>muh inefficient meat

either you get replaced by a robot or you work for less than a robot does

or you sperg out and revolt and the people that use robots shoot you full of holes as soon as you step out of your cave

suck it down

>>6297436
>tiny brain that can type internet acronyms

>>6297459
>not abbreviating everything in a character-to-information dense lexicon
it's like you don't even want elizier yudkowsky to deduce you from 1st principles

>>6297469
seriously how do you deal with the fact that computers are cheaper than people

where does that fact fit into your kneejerk dismissal of anything your owners tell you is dumb

lmao i bet you call bitcoins "buttcoins" you eunuch'd ponce

>>6297469
even though I hate that autistic fаggоt, I'm pretty sure his name is Eliezer.

and attacking where something has been posted, and not attacking the point that was made, is the lowest possible form of argumentation and a sure sign of a mental midget and a low-IQ} dumbfuck.

pic related but pointless in your case.

>>6297486
i deal with it by actually working in computer science research and having some sense of the state of the art and its limitations

and i hate sci-fi and don't read it so it doesn't affect my judgement

that's all really

>>6297489
>and i hate sci-fi and don't read it so it doesn't affect my judgement

lmao get a load of this posturing dork

>>6297489
like lol @ this idea that cs research is synonymous with loving bitcoin and observing argumentative rigor on 4chan

>>6297491
oh man is this thread becoming about me again it was already pretty funny the first time

i don't really want to blog tho so w/e you can have the last word be sure to knock it out of the park

>>6297492
Why do you post like some goony faggot retard? Do you not realize how it makes you sound?

>>6297495
haha look im really not involved in this and your not getting to me

im so above this

thats why i keep coming back and poasting even though i say im leaving

>>6297486
>seriously how do you deal with the fact that computers are cheaper than people

Because bindis and chinks are cheaper than computers and you don't have to pay for their upkeep.

>>6294794
Congratulations! That is the most retarded and childish argumentative maneuver I've seen all week!

>>6297589
It doesn't matter whether or not they're all incompetent. The point is money. You can either pay some guy to do shit you can write a program for, or you can make a computer do it and your only costs are an admin and electricity. The people who do the former are going to lose, because their operating costs are higher compared to the benefit they provide.

It's the same reason we don't have many fighter pilots anymore, just drone-fleet operators.

>>6296458

It's actually possible to view PCA as a special case of a neural network where the output layer is the same as the input layer. If we add more than one hidden layer and use a non-linear activation function then we achieve a deep, non-linear variant of PCA called a stacked denoising auto-encoder.

http://deeplearning.net/tutorial/dA.html

In many ways the neural network is much more powerful, since it can be optimized to learn both an unsupervised objective (like reconstructing an incomplete version of the input) and to learn a supervised objective (like recognizing faces) whereas PCA is unsupervised, so it may learn irrelevant features.

Please, someone call the police, all capital letters and periods have been kidnapped !
The return key is a prime suspect.