clfougner

7 năm trướcdiscuss

To anyone else wondering what's meant by "rejected energy":

https://energycultures.org/2014/07/rejected-energy-much-ener...

clfougner

8 năm trướcdiscuss

That isn’t the same thing at all. When you fly for a vacation, the flying is a means to an end (transportation), not the goal in itself. This discussion is about driving for fun, in which the driving is the goal in itself, not transporting yourself to a “fun” destination.

clfougner

9 năm trướcdiscuss

Darmyian founders wrote the following in another post[1]:

> Roughly 97% sensitivity (3% false negative) and 85% specificity (15% false positive)

So with this information, the calcuation of Bayesian probability is as follows:

    (0.97*0.016)/(0.97*0.016+0.15*0.984) = 0.095

So a 9.5% chance of actually getting Alzheimers within the following 15 years if Darmiyan's test is positive. I don't have much formal statistical training, so I'm all ears if I'm making a mistake in the calculation here.

They also write the following:

> False positive here is not real false positive, as the software is detecting abnormality in people who are still cognitively normal.

Which is a terrible excuse - abnormality is only clinically relevant if it leads to disease. By this logic I can create the world's greatest test for cancer simply by saying that every given individual has cancer (100% sensitivity), and if they don't have it yet, they do carry the genetic abnormalities that will lead to cancer eventually - no "real false positives", right? Simply saying that "sooner or later, our test will prove correct" is not good enough here.

Founders of Darmyian: I commend your efforts in this space as a tool for research, and can potentially be very valuable for clinical trials. However, offering it to consumers as a screening tool when there are no proven preventative measures is - in my opinion - completely unethical and comes across as an attempt at trying to profit off fear mongering.

    [1] https://news.ycombinator.com/item?id=15083617

clfougner

9 năm trướcdiscuss

"At Darmiyan we detect Alzheimer’s disease up to 15 years before symptoms" (original post)

"At present, there is no definitive evidence to support that any particular measure is effective in preventing AD"[1].

"Now we have analyzed more than 3000 brain scans and our software’s predictions are 90% accurate." (original post)

"In the United States, Alzheimer prevalence was estimated to be 1.6% in 2000 both overall and in the 65–74 age group"[2]

I'm assuming the 3000 brain scans you're referring to is from individuals which progressed to Alzheimers, or at least a dataset with such individuals highly represented (if this were 3000 random individuals, at a 1.6% prevalence, that amounts to 48 individuals who eventually got Alzheimers).

So according to my calculation of Bayesian probability, with a 90% sensitivity (as I'm interpreting your comment), and a 1.6% prevalence in the population, a randomly screened individual with a positive test will only actually have a 12.8% chance of getting Alzheimers. So you'll be diagnosing lots of people so that they can have an impending Alzheimer's diagnosis hanging over their head for the remainder of their life without actually being able to do anything about it, and of this cohort just over 1 in 10 people will actually end up getting Alzheimers.

Please tell me you're only planning on offering this for researchers, and not actually going to try to get individuals screened? Or am i missing something about your value proposition?

    [1]https://en.wikipedia.org/wiki/Alzheimer%27s_disease#Prevention
    [2]https://en.wikipedia.org/wiki/Alzheimer%27s_disease#Epidemiology

Edit: fixed sensitivity vs. specificity error

clfougner

9 năm trướcdiscuss

Subscribers pay 10$ a month, of which 70% goes to the artists/record labels. Spotify itself then has a a revenue of 3$ per user per month, or 36$ per user per year. 60 million subscribers gives an annually recurring revenue of 2.16 billion dollars. From the chart in the article, they grew from 50 to 60 million subscribers over 4 months, so that annualises to a 60% growth rate on 2.16 billion in revenue (ex royalties). Questionable how long they'll be able to maintain that growth rate, but at the moment it's looking great.

One factor that hasn't been accounted for is the hosting costs for streaming/data transfer. Ballpark: assume the average user listens to music 1 hour/day, 365 days a year, at a bitrate of 320 kbps: 320 * 60 * 60 * 365 = 420480000 kilobits = ~50 gb per user per year. AWS' pricing page [1] has the lowest listed rate of 0.05$/gb (in reality likely lower for a huge client like Spotify), which amounts to a data transfer cost of 2.5$ per user per year, on a revenue of 36$.

Further Spotify has a large number of free subscribers (around 50-100 million, don't think figures have been released for this recently), which produce far less revenue per user from ads, but do generate revenue nonetheless.

Seems to me that it should be entirely possible to become decently profitable, provided that customer acquisition costs don't grow out of hand. For context: Facebook had an ARPU in the US and Canada in Q2 2017 of 19$, whereas Spotify has an ARPU of 9$ per quarter (ex royalties). If Spotify can be half as profitable as Facebook (on a per user base) and can keep growth going at this rate, they should be able to do fine.

[1] https://aws.amazon.com/ec2/pricing/on-demand/

clfougner

9 năm trướcdiscuss

Route your page through Cloudflare (or presumably any other CDN) and you can get HTTPS set up on your Github Pages hosted static site quite trivially.

clfougner

9 năm trướcdiscuss

Route your page through Cloudflare (or presumably any other CDN) and you can get HTTPS set up on your Github Pages hosted static site quite trivially.

clfougner

9 năm trướcdiscuss

> No ones knows for sure because no one can objectively explain the nature of consciousness and the mind in the first place.

True, but I'm still waiting for a single proposal as to where a mental illness would subside other than physical structures/molecules, because the only other explanation I can think of are in the realm of the metaphysical

> The study has no control group and therefore is not objective. If you read the Methodological Issues section, you'll see that this is essentially a study of the neurochemical effects of antipsychotic drugs.

The article isn't a study/original research, it's a review based off of existing research - generally - in which brains from individuals with schizophrenia are compared to the brains of individuals without schizophrenia (controls). Antipsychotic drugs as a confounding factor is definitely an important factor in these studies though. There are however a number of factors that are are not affected by antipsychotics, notably genetics[1, 2]. Certain genetic variants strongly increase the risk of mental illnesses - by what mechanism does this work other than by translating to proteins that exert an objective, physical effect? (Note that studies on this are primarily GWAS based on large populations, not simple hereditary/familial studies in which one could argue that environmental factors were confounding)

  [1] https://en.wikipedia.org/wiki/Causes_of_schizophrenia#Genetics
  [2] https://en.wikipedia.org/wiki/Bipolar_disorder#Genetic

clfougner

9 năm trướcdiscuss

You're creating the same false equivalence again by stating that an objective post mortem diagnosis not being practically feasible is the same as an objective cause(s) not existing. In my previous post I linked to multiple pages listing objective causes/mechanisms for two mental illnesses. Diagnosing a mental illness post mortem isn't as simple as finding rhinovirus in a person's nasal cavity, but just because the current knowledge of the disease isn't sufficient to create a satisfactory model that can reliably make a diagnosis post mortem doesn't mean that it can't be done. And if you don't think mental illnesses subside in molecules and physical structures in the brain, where exactly do you think they come from?

Oh, and you kind of can diagnose certain mental illnesses post mortem: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3181616/

clfougner

9 năm trướcdiscuss

> There are objective tests for all of these pathologies, though those tests are often inconclusive in the face of the symptoms.

No, objective tests are in these cases used to exclude other causes to the symptoms, not to diagnose the mentioned illnesses. These are diagnoses of exclusion[1].

> There are physical symptoms of mental illness, but as yet there is no proof of physical causes, which is why the chairman of the Department of Psychiatry at Duke University School of Medicine and the DSM-4 stated "psychiatric diagnosis still relies exclusively on fallible subjective judgments rather than objective biological tests". You'll find no objective biological tests in the DSM.

I'm not disputing the fact that it's not possible to do pathophysiological tests for mental illnesses in current clinical practice, but you're creating a false equivalence by implying that being unable to test for a pathophysiological factor is the same as that factor not existing. There are plenty of pathophysiological changes that occur in mental illnesses [2-4], but for obvious reasons it's not feasible to haphazardly take biopsies of the brain to test for them, especially when the patient history and subjective examinations suffice to make a diagnosis in most cases.

Conversely, my point still stands that the vast majority of somatic illnesses are diagnosed off subjective symptoms - doctors don't bother doing objective lab tests for a cold or a sprained ankle when the diagnosis is glaringly obvious based off subjective symptoms.

  [1] https://en.wikipedia.org/wiki/Diagnosis_of_exclusion
  [2] https://en.wikipedia.org/wiki/Mechanisms_of_schizophrenia#Pathophysiology
  [3] https://en.wikipedia.org/wiki/Biology_of_depression
  [4] https://en.wikipedia.org/wiki/Causes_of_schizophrenia

clfougner

9 năm trướcdiscuss

Afraid that's not true. Around 60-80% of diagnoses in primary care settings are made purely on the basis of the patient history (followed by a brief clinical examination which usually aims at eliciting subjective responses)[1]. In principle it's true that you could find a pathological explanation for all somatic ("physical") diseases which in practice are diagnosed off patient histories (with the exception of diseases for which the pathogenesis is unknown, f.ex. fibromyalgia, IBS, CFS, etc). However - in principle - the same goes for mental illnesses, although the pathogenesis of mental illnesses are generally poorly understood. If you don't think there's a physical basis for mental illnesses though, skim through the figures in this article[2] for a prime example of the ways modern medicine is disproving that. Diagnoses of both mental and somatic illnesses are heavily based off subjective factors, and both mental and somatic illnesses have very real pathophysiological etiologies.

[1] Source: medical school. Not finding a study with exact numbers from a quick search. [2] https://www.nature.com/nm/journal/v23/n1/full/nm.4246.html

clfougner

10 năm trướcdiscuss

Basically, because AT&T is paying more for Time Warner than the market had valued it at. As an example:

Company A has 50 billion dollars cash on hand, and their core business is valued by the market at 100 billion, so their market cap is 150 billion. Company B's core business is valued by the market at 25 billion and has no cash on hand, so their market cap is 25 billion.

Company A then purchases company B's core business for 50 billion dollars. Company A then has no cash on hand, their original 100 billion core business, and a new business that the market valued at 25 billion, so their new market cap is 125 billion (a 16.6% decrease).

Company B (or their stockholders, strictly speaking) on the other hand has 50 billion dollars in cash, and no core business, so company B now has a "market cap" of 50 billion dollars (100% increase).

More often than not, the market (i.e. those buying and selling the stocks) will assume that the company being acquired was correctly priced before the purchase and that the acquiring company overpaid.

clfougner

10 năm trướcdiscuss

Thanks for the great video! Between the three you recommended I ended up going with Django, largely because I've been meaning to get a grasp on Python for a while (I do a fair amount of bioinformatics work in R, and Python should be a useful complement for statistics tasks). Also, the Django project website has great tutorials for beginners to follow!

Submissions

What it would take to reach the stars

Behind the scenes at Faraday Future

The U.S.'s Misplaced Admiration for Europe

Ask HN: How does a beginner get started with web development in 2016?

comments