xherberta

Crime as in algo crime; there are not humans pulling a lever labeled “trash this relationship.”

Observe how this amounts to crime. The platform has motive: -newly single users attract more views -newly single users are worth more to advertisers because they’re likely to spend on appearance, travel, new hobbies, and big ticket items formerly shared with a partner. The platform also has opportunity because it has data on exactly what content has highest probability of nudging a particular user to break up. (Perhaps the “you are being abused by a narcissist” stuff others have mentioned works nicely.) (See also Shoshanna Zuboff)

What if a learning dataset simply has not enough info for a correct answer to have a greater weight than all the “con” noise?

Indeed. I wonder what happens as available training data shifts from purely human-generated (now) to largely AI-generated (soon). Is this an information analogue to the “gray goo” doomsday that an uncontrolled self-replicating nano device could cause?

can it answer “I don’t know this” Such a fabulous question. This statement likely appears infrequently in the training data.

Curiosity killed the cat. We’ve developed disgust and taboos and physiological responses like vomiting to safeguard our interactions with living things that are small enough to eat.

Perhaps an advanced form of life/intelligence that has survived contact with various others would be guided by some form of wisdom/disgust/caution/discretion. It would not be surprising if they noticed little clues like our rapid deforestation, rising temperatures, shrinking ice caps, nuclear detonations… which would make us seem slovenly or likely to have “poor friendship skills.”

Here's a discussion on pain points in the ICU: https://news.ycombinator.com/item?id=33661482

think I’m burning out (home, kids, marriage, etc)

Do you think you could manage the job if there were no expectations on you outside the job? Could your partner get on board with you needing some time with very few home responsibilities (as it's best for the family in the long run)?

Can you find access to nature? For me, a couple of unstructured hours getting bored in a truly natural setting feels almost like a vacation.

Karezza could help create a generous, non-transactional vibe at home and help you maintain creative energy.

Best wishes as you navigate this.

Art Against Despair is a book put together by The School of Life. It's arranged with one page of text for each image, so it's an easy serving size when you don't have much time or attention. These are not nice pictures of flowers but, for example, why it's worth spending some time gazing into one of Rothko's black canvases, and for that matter, the darkness inside one's own being.

One that Loeb has mentioned: the status quo 10-year plan involves searching exoplanets for oxygen, which could arise without life, and which organisms would not necessarily need or produce. Instead, Loeb thinks we ought to look for industrial pollutants. A positive result on that would be extremely interesting.

If the US hadn't dropped the ball on enforcing existing antitrust laws in the 80s and 90s, we wouldn't have to ask. Perhaps many more regular people would own productive enterprises, and wouldn't be looking to government to address inequality.

Now that both corporations and government are so big and intrusive that they're somewhere between a nuisance and prohibitive toward small-scale enterprise, we should at least think about ways to sic them on each other.

Wow! I did somehow miss the 15 years aspect. Mysterious. I'm glad you're feeling better.

I am skeptical of the headlines touting medical mistakes as the No. 3 cause of death in the US. But to be fair, someone could put up a similar site listing thousands of people who have been harmed under conventional medical care.

Your site doesn't provide enough info to compare the rate of harm. For mild ailments, which is what I assumed OP was speaking of, homeopathic treatments might have a lower rate of harm.

Iatrogenesis is a possibility with any health care. And, to give your link credit, homeopathy certainly sounds like a dangerous route for serious conditions.

"The great secret, known to internists and learned early in marriage by internists wives, but still hidden from the general public, is that most things get better by themselves. Most things, in fact, are better by morning."

- Lewis Thomas, The Lives of a Cell (Among other achievements, he was dean of Yale Medical School and president of Sloan-Kettering. This collection of essays was published in 1974 and is absolutely prescient and fascinating - he offers profound thoughts on the nature of science, computing, and medicine that have stood the test of time astonishingly well.)

I agree.

Kahneman and Harari also discussed how the shift to agriculture, while a great step forward for humanity as a collective group, was a major bummer for the individual. (Much more interesting and healthier to be a hunter-gatherer than a serf.)

On people becoming superfluous to the market and the state as machine intelligence progresses:

"... [in the future, work no longer exists for] most of humanity... That mass of people cannot work, but they can still kill people..."

"and how will they find some sense of meaning in life when they are basically meaningless, worthless?

My best guess at present is a combination of drugs and computer games as a solution for most"

On "one of the big problems with technology":

"It develops much faster than human society and human morality, and this creates a lot of tension. But, again, we can try and learn something from our previous experience with the Industrial Revolution of the 19th century, that actually, you saw very rapid changes in society, not as fast as the changes in technology, but still, amazingly fast.

The most obvious example is the collapse of the family and of the intimate community, and their replacement by the state and the market."

Even after multiple generations of mutations,

Bacteria can do thousands or millions of generations more rapidly than we can tweak our nanotech.

I don't know. With the third-party payer system, cost is hardly a factor in medical decisions.

Point taken. Thanks for the call out.

Why talk in terms of one big company? Maybe small investors with conflicting ideas and interests would like to own and run small companies. Except, a couple of big companies are making that very difficult.

The US hasn't enforced its anti-trust legislation in over 30 years. Some link Reagan's "assault" on anti-trust to the breakage of the American job market. There's been zero net job creation in the US between 1999 and 2009, a time which also saw significant consolidation.

Small businesses drive net job growth, and they are also more innovative, producing more patents per employee than large firms.

Not to mention, smaller businesses allow more people to own capital. I'm sick of economic discussions that focus on jobs and wages - ownership is what gets people out of serfdom.

But, little guys are feeling the squeeze. "The Goliaths of today are so big and so adept at protecting their turf that they leave few niches open to exploit."

So, this isn't only about privacy, or the false perception that "big is evil." It's about economy as ecosystem. "Biodiversity" in the economy creates robustness against hard times and distributes resources in ways that allow many participants to thrive. Monocultures are extractive, and ultimately fragile.

Edit: source - http://bit.ly/2rhPiXp

I guess as long as there has been cheese, mice have been eating it.

I think you may be underestimating the real-world messiness of this water sanitation example, and other types of problems where location-specific factors WILL make or break the project.

There's an enormous gap between knowing the specs of a pump and knowing which company will make one that won't break and will provide decent service for it. A water sanitation plant that would work in one part the US (or one part of a state) might fail in another due to any among hundreds of factors.

Any time you're doing something that involves silt, algae, microorganisms, weather, human factors... etc... those factors are likely to make a fool of you and your textbook. The "scoop from a grizzled veteran" suddenly becomes essential.

There's a well-known example of a sludge field that worked on a small scale but did not work when executed on a large scale. Total disaster. Hugely expensive project, requiring hugely expensive cleanup.

Sure, there are plenty of realms within tech where "grizzled veterans" aren't needed and where solutions can transfer and scale nicely. You can't safely extrapolate that to real-world problems.

In many endeavors, local knowledge can't safely be ignored, and hubris leads to tragedy.

Nina Munk's reporting on the Millenium Villages Project is illustrative of the difficulties one can be blindsided by when trying to get something done with insufficient regard for local conditions: http://www.econtalk.org/archives/2014/01/nina_munk_on_po.htm...

I agree. The incentives are out of whack. Being a self-pay patient over the last several years has been enlightening. The health care people I've dealt with are generally ready to go well out of their way to help make care more affordable. Once I was given an unasked-for 55% discount on an ER visit. (They're happy they don't have to deal with an insurance company!)

I've heard that people with high-deductible plans are now finding it advantageous to just say they are self-pay to get the discounts.

For comparison shopping, healthcarebluebook.com can give an average price for a certain procedure in your area.

Here are a couple of related resources:

An eye-opening conversation with a palliative care physician: http://www.econtalk.org/archives/2014/11/becky_liddicoat.htm...

"Five Wishes," a tool to guide you through conversations with loved ones about end-of-life care (not to be taken as a replacement for a living will, as I don't think it will hold up legally in all states) https://agingwithdignity.org/five-wishes/

And if you really want to get motivated to have these conversations with your loved ones, check out Otis Brawley's book "How We Do Harm," which has a chapter that is a HORROR story of what happens when a patient's family asks that "everything" be done to keep him alive. The doctors felt that they were torturing the patient, who had no reasonable prospect for recovery. Yet the family insisted that "everything" be done, including extremely painful procedures and withholding pain medication because the patient's respiration was too fragile.

Don't be this family! Don't be this patient!

There's only one place I'm aware of on earth where organs can be sold: Iran. Here's an interview with an economist who helped set up that market, which does seem to function fairly efficiently and has some safeguards in place to prevent coersion.

http://www.econtalk.org/archives/2015/09/tina_rosenberg.html

Edit: specifically, this is about kidney donation from living donors, not after-death donation.

Right. Monetizing death. Ugh. I'm also not certain that you can donate organs if you've been on pain (and other) medications.

Unfortunately "Magician" doesn't sound like a person you would want messing around in your code base, though.

This is a good point, and to her credit, Cathy O'Neil is careful to say that humans are worse. She's arguing for better algos, ones that are carefully thought out to avoid the worst of these pitfalls.

Even a minimum wage job should cover the cost of day care for their youngest child, who is too young for full-time schooling.

In many markets, no it won't, if you can even find a daycare with an opening.

Anyway, this plan sounds like a net negative for the family unless the job can cover daycare PLUS commuting costs, work wardrobe, increased food costs (less time to cook from scratch), and the hassle involved with finding alternate care when any of the four kids are sick.

Some people consider the "humanity value" of having a parent at home to care for children to be worth the financial downside. That's their choice. By the way, most developed nations agree...

From the New Republic: *Among developed nations, the U.S. is now the only one that doesn’t guarantee some kind of paid employee leave available for new parents." http://bit.ly/2pZ03Ay

This is a VERY important point.

I don't use FB, but a data-promiscuous person who knows a ton about me recently unfriended my closest contacts. Ignorance is bliss?

As Cathy O'Neil points out, black-box algos are deciding things that have a huge impact on individuals' lives: which teachers get fired, which prisoners don't get parole, and which loan applicants get approved and denied. In the latter two cases, the algos are tougher on minorities.

It's troubling when we don't know why an algo is racist. We need ways of checking that these influential algos aren't reinforcing trends we would rather diminish.