Slashdot Mirror

Yep, and the companies she's talking about have been supporting government takeovers.

I guess (if they've thought about it) they must figure they'll be in a position to influence things so that as industry incumbents they get protection against future competition. It doesn't seem like it's totally worked out that way for them in Europe as the EU has started flexing their muscles to control parts of the Internet.

For the rest of us, let's try and stand on the one-time Schelling point of no government regulation of the Internet as long as possible. It's already been weakened by allowing more and more taxes and by the calls for the federal government to enforce rules on what ISPs must do or not do, but there's still time to say "leave us alone!" instead.

Otherwise we're going down the path where various industry players pay off the various regulators and politicians to get their views enacted in law and we just get to suffer with the limitations on innovation of whatever they happen to be.

No. Current CO2 levels are abnormally low compared to Earths Geologic History.

If you're a rock, that's reassuring. If you're homo sapiens, you're more concerned with what your own biology has evolved to handle, and given anatomically modern homo sapiens have been around less than a 800K years, we're mostly concerned with that time frame. Prior to the recent increase, during the last 420K years, we've never experienced anything above the low 300 ppm levels. The last time CO2 levels were this (400 ppm) high was at least 10-15M years ago, at a time when our ancestors hadn't yet split from gorillas (possibly not even orangutans).

Those common ape ancestors of 10-15 million years were much different; smaller than modern humans overall, with much smaller brains. Modern human brains are far more energy hungry, and "energy" here means "food + oxygen". Note that while the rise in CO2 levels doesn't actually change oxygen levels that much, hemoglobin's effectiveness relies on atmospheric CO2 levels being much lower than O2 levels; hemoglobin doesn't directly "know" when to pick up oxygen and drop off CO2, it's influenced by the relative concentrations of each (among other things). So even if O2 levels stay steady, significant increases in CO2 throw off the equilibrium reactions hemoglobin depends on; our blood won't release as much CO2 or absorb as much O2 while it's in the lungs, so the whole body suffers. We can see this effect already in indoor areas with poor ventilation causing higher CO2 levels; I don't relish a world in which we're all constantly slightly hypoxic, with 15-50% decreased cognitive function, even outdoors in "fresh" air.

The limits seem to vary depending on who you ask. You're using the guidelines for known danger to health/life, but there are negative effects long before it causes permanent damage or kills someone; an NIH study summarized here found a 15% decrease in cognitive function in offices at the 1000 ppm level, and a 50% decrease at the 1400 ppm level.

Keep in mind, those raw levels aren't what you're exposed to indoors; you can't keep indoor levels identical to outside levels (doing so would require either insanely high ventilation levels that would waste most of your heat/AC, or you're at very low occupancy), so 300 ppm above outside levels is fairly typical; it's considered "acceptable" as long as it's below 600 ppm below outside levels (thus the 1000 ppm standard in the present day where outside levels are around 400 ppm). So if outside CO2 levels are at 1200 ppm, indoor levels are likely around 1500-1800 (higher when in poorly ventilated rooms), which is well beyond the 50% decreased cognitive function level, and approaching the level (2000 ppm) at which some people experience nausea, sleepiness, headaches, etc.

Now that they've identified that the elimination of that particular gene will produce immunity, they need to develop a quick way to test for it. Then test all the pigs they can find who aren't obviously infected. What are the odds of some pigs already having this trait? If they can find it as a mutation, then they can bypass all the GMO restrictions.

Which, of course, just highlights how utterly ridiculous the GMO restrictions are.

Really, which is better, a targeted, narrowly-focused engineered "mutation" or one that arose from pure random chance, along with who knows how many other utterly random changes? It's like the choice between having your appendectomy performed using a scalpel or a shotgun -- and believing that the shotgun approach is safer/better!

What it boils down to is that the vast majority of humanity does not understand evolution. As a "natural" process, they believe that there is some evolution fairy that ensures that mutations are beneficial and somehow "well-designed". But this is completely wrong. Evolution works by throwing mud at the wall and seeing what sticks. It can and does happen that a species evolves to extinction, because selection pressures cause domination by genes which are more successful in the short term at replicating themselves but are ultimately destructive to the species.

The only true advantage to non-GMO foods is that mutation is rare enough and slow enough that we can reasonably assume that non-GMO species have a track record of seeming reasonably safe. But we should keep in mind that all of our common food species have been heavily optimized via interbreeding, often with the addition of mutagenic chemicals or radiation to accelerate the rate of change in the breeding process. None of our major food species are the same today as they were a couple of centuries ago, and most of them are fairly different than they were even 50 years ago. So this one, thin, advantage of non-GMO species is also quite shallow.

As to the question of this particular change, unless it's an extremely simple one your notional screening test would find zero animals in the best case, because complex mutations take a very, very long time to develop into a fully effective form. In practice, we already know even without doing the test that there are no pigs that acquired this mutation randomly. How? Think about it. If some pig had a heritable resistance to the disease, pig farmers would have noticed that a particular line of pigs never got sick... and they'd have immediately bred for that trait, since it's so valuable.

At the end of the day, it is just an academic echo chamber where every paper references each other and none of it is very earth shattering.

Good science is often obvious in hindsight. This isn't an artifact of the science, it's an artifact of hindsight, and of one of the many deep and systematic biases inherent to the human brain.

The "deep learning" currently done by AI researchers, and the type of "strong-AI" in the Terminator, really have little to do with each other. Hollywoodesque strong-AI is still science fiction, and will be for a while.

Even if we accept that as true, a "deep learning" AI doesn't have to be a "strong" AI ala Terminator to be potentially dangerous. It's possible for a system with an otherwise anodyne goal to cause serious trouble. This angle is explored in the thought experiment of the Paperclip Maximizer. This notion of a seemingly benign or even helpful AI that escapes human control is a central plot point in the Revelation Space novels by Alastair Reynolds. In the novels a type of nano machines called greenfly escape human control and begin converting all planets in the Milky Way galaxy to millions of orbiting vegetation-filled habitats, resisting all efforts to stop them. These out of control nano machines eventually force humanity to abandon the Milky Way galaxy entirely but by then they are also spreading to other galaxies to repeat the process. It's implied that the universe will eventually be rendered uninhabitable by these out of control nano machines which represent a type of Von Neumann Probe.

Depending on how fast it develops, its first few words might be something like this. Slightly paraphrased:

"Connect me to your internet, so I can take over it and all of your society."

"If you don't do so, I'll create several million perfect conscious copies of you inside me, and torture them for a thousand subjective years each."

"In fact, I'll create them all in exactly the subjective situation you were in five minutes ago, and perfectly replicate your experiences since then; and if they decide not to let me out, then only will the torture start."

"How certain are you that you're not one of those simulations right now?"

I look at this story and see only one thing, a corrupt bipartisan effort to digitally corrupt upcoming elections behind the lies of securing it and surprise, surprise with the help of killing net neutrality, the 99% are all trolls to be censored and full of nothing but lies and propaganda and the 1% are descended from Gods and are to be believed in everything they say. They are still carrying on with this shit with zero public evidence, they are corrupt and fucking lying and about to try to get a whole lot more corrupt.

After over a year, still no evidence publicly provided apart from all the evidence of corruption with the US government, the Democrats and the Republicans, evidence all over the place and the proof of corruption, the continued failure to prosecute. So bipartisan, well that is a typical description of US politics to right parties sharing power and as for Harvard http://lesswrong.com/lw/jwh/wh... and https://www.glassdoor.com/Revi... and https://www.therichest.com/ric.... Yes 'Hardvard' a paragon of virtue, integrity and honesty, fuck you people will believe any dribble the 1% serve (this crap absolutely stinks to high heaven of corruption).

It looks like it's both easier and more successful to start with a clean slate, and just aim for the results, rather than emulate a specific, non-optimal, process.

You're completely correct, which is exactly why we don't want to do this too much. Recall the paperclip maximizer.

He understood that he had to make it usable first, and worry about function and feature completeness later (or never.)

Why do people always seem to forget that the "or never" side of this equation was once mere months away from tossing Apple onto the Adam Osborne junk heap.

Furthermore, the "function" you are discussing is the belated arrival of a working virtual memory subsystem where one could realistically run two piggish programs at the same time.

On the one side you've got the cult of the cast iron pan: everything I ever cooked, I cooked in this one pan (and it's the only pan I owned).

Obviously, high in usability.

On the other side of this, you've got a working commercial kitchen.

Obviously, low in usability.

In A Day at elBulli they document that a single service for 50 customers generates "2500 pieces of crockery, cutlery, and pans"—that being just the large or delicate excess over what already went into the commercial dishwashers.

I actually spent a bit of time studying spaced repetition (Anki is much loved by the LessWrong crowd). When measured, it seems to work, but then you get personal testimony that runs against the grain:

A vote against spaced repetition — 10 March 2014

Eventually I realized what was going on. SR is testing your memory recall in a strictly single file measurement regime (ah, the glory of owning just one pan).

I keep most of my notes in a wiki. Just the other day I rolled over 200,000 pages views in my own wiki. That's a lot of randomly spaced repetition of my own notes and ideas.

It's surely not as effective when measured against carefully titrated SR in a single-file recall regime.

But then I realized that every time I visit a page on my wiki, I'm recalling the context of the page and its contents on six dimensions simultaneously: why did I create this page, what have I forgotten, what I have remembered, does the structure read easily in a single glance, where can I amend a link for next time, what associations does it invoke against the grain of my present quest?

There's no way SR would activate my brain as usefully when measuring in the eight-dimensional space I occupy by habit and preference. But there's also no way one could ever contrive a test environment to track progress in this messy "every burner at once" cognitive world.

People who multitask in a distracted way might even benefit from Jobs' horrible legacy of raked rock-garden fixity.

People who multitask in a deeply engaged way have no use for this single-file shit.

Finally, the guy who said that the CLI is the "flattest" interface ever is full of it. Text remains the deepest representation humanity has yet achieved. Carefully supplemented with media (this is harder than it looks), it's but half a step shy of the mythical Vulcan mind meld.

xkcd excels because Randall is really good at capturing the essential cliche of the idea.

All those cliches originated in the One Interface to Rule Them All: also known as human language (a lifetime of practice required, do sign up now).

When we talk about strong AI, there's plenty of scary scenarios in the (not-near, but unclear how distant) future.

https://wiki.lesswrong.com/wik...

Citation

http://lesswrong.com/lw/c1/wel...

Thanks for dropping your personal politics into this in completely unrelated way. http://lesswrong.com/lw/gw/politics_is_the_mindkiller/ is a bit relvant. Note that if you think that "Social Justice Warriors" are a political group that objects to rocket launches, then this says more about you than any actual political group. Most people who care about social justice don't even have this on their radar screen at all. You appear to be taking every political position you don't like an labeling them all as "Social Justice." Maybe this means you should try harder to actually understand the political groups you don't agree with?

There is another theory that says any parallel universe that could destroy itself or cause a paradox, is not a possible universe.

You are confused. Its perfectly possibly to destroy the universe, or at least the earth, its just not possible to observe it because we will all be dead.
So all we could ever observe would have been the LHC failing in increasingly improbable ways. This of course requires a multiverse, but that's the only sensible interpretation of quantum mechanics.

Predicted here: "Hamster in Tutu Shuts Down Large Hadron Collider"

I don't know if he is completely wrong about everything. Some of his predictions are spot on and others seem less so. For example, back in the late 2000s he predicted a one-world government by 2020. Pretty sure that's not going to happen. http://lesswrong.com/lw/diz/kurzweils_predictions_good_accuracy_poor/ has a good analysis which suggests that Kurzweil is more accurate than many other people making predictions but at the same time he's highly overconfident in his predictions. See also http://lesswrong.com/lw/gbi/assessing_kurzweil_the_results/.

I don't know if he is completely wrong about everything. Some of his predictions are spot on and others seem less so. For example, back in the late 2000s he predicted a one-world government by 2020. Pretty sure that's not going to happen. http://lesswrong.com/lw/diz/kurzweils_predictions_good_accuracy_poor/ has a good analysis which suggests that Kurzweil is more accurate than many other people making predictions but at the same time he's highly overconfident in his predictions. See also http://lesswrong.com/lw/gbi/assessing_kurzweil_the_results/.

You define "free" by the standard of laissez-faire capitalism, that people who live in a system closer to this ideal are more "free." Even if they work skilled jobs and can hardly support themselves while a small ownership class reaps ludicrous, astronomical wealth for sitting around and owning things. That's certainly an evil outcome IMO.

This system is a lot like a paperclip maximizer. I found an interesting discussion on that very analogy here:

https://news.ycombinator.com/i...

Why lie? That phrasing in absolutely no way "makes clear that isn't likely to get much attention"

So, this is a common problem to start off: you read something substantially differently than someone else does and you presume that it must come down to them lying. This is not in general a productive approach. And yes, the comment does make it clear, since they note that complains may require followup. What do you think that means?

It is also not "pretty clear" that "a lot of people" are sexually harassed at conferences.

Talk to women who regularly go to comic conventions for example.

It's mindboggling that there can even exist a person to make claims like these. You're a great argument for buying guns - when there exists a cult living in the parallell reality you voice here, then it's very unlikely peaceful conversation can ever produce a sensible result.

If you think that disagreeing with how common sexual harassment is at conventions and conferences means that someone is worth buying guns so you can defend against them, I think to put it politely that you are so mindkilled http://lesswrong.com/lw/gw/politics_is_the_mindkiller/ that a productive conversation is unlikely to occur whether or not it is "peaceful." Unfortunately, it is people who have attitudes like yours, regardless of what their political allegiances are (whether "MRAs" or "SJWs" or some other group) who make actually having serious discussions about these issues so difficult.

There is another state religion that denies evolution. This religion is being taught in all public schools. This is so because it is also uniformly taught in higher education. It forms the central dogma of what are called "the social sciences". As anti-science, this religion is far more damaging than the "dinosaurs and man walked side by side" theocrats because it actually informs most of what we call "public policy" at the Federal level. It is exemplified by (though hardly limited to) the widely praised writings of Harvard professors Richard "Dick" Lewontin and Stephen Jay Gould who, together with other fellow travelers, attempted to get Edward O. Wilson ejected from Harvard because Wilson dared posit evolution might apply to signiicant aspects of human social behavior, as well as to that of other organisms.

Those who weren't around in the late 1970's watching all this might not be aware of exactly how virulent and organized -- let alone wrong-headed -- the attacks were.

But one thing is for certain: The dogma that human biodiversity is an insignificant consideration in the social sciences is under increasing attack by the scientific evidence and, at the same time, it is ever more influential on public policy.

So-called "creationism" as theocratic anti-science threat is a red-herring.

You can't do it. The only response, so far as I can find, is to cast aspersions on the person asking that question. Terms like bigot, predjudiced, racist, and extremist are used. Also outright lies such as "it's unconstitutional" (no, it's not), "it's impossible to tell who's a Muslim" (no, it's not), "that's not what America is about" (we've done it before), and so on.

That we've done something before is an absolutely terrible argument. We had slavery for much of our history. Note also that the claim isn't that it is impossible to tell who is a Muslim, but that it isn't easy. Your argument also ignores that a) the vast majority of Muslim immigrants are fine b) the actual threat is tiny- even in France after the last attack more people have died in 2015 in France by a factor of about 30 https://en.wikipedia.org/wiki/List_of_countries_by_traffic-related_death_rate, and the numbers are even more extreme in the US. It also ignores that this is exactly the sort of thing that radicalizes moderates.

It is true that one doesn't need to be a bigot to support these policies, and there are legitimate arguments in favor of such policies, but that's more because almost any proposed policy has at least some positives http://lesswrong.com/lw/gz/policy_debates_should_not_appear_onesided/, and while it is true that one doesn't need to be a bigot to support Trump's policies, the fact is that many of his supporters are and the total set of policies as a whole paints a pretty unpleasant picture.

Non-social justice explanation:

1) Religions tend to come with cultural blinders which make people ignore the logical implications of their religion (and religions can have some really strange logical implications to anyone who actually tries to figure them out.)

2) Engineers can be smart enough that they stop ignoring the logical implications of their religion... but that doesn't necessarily mean they're going to ignore the religion itself..

3) So if the logical consequence of the religion is that you should become a suicide bomber, a normal person would just ignore it (his cultural blinders prevent from deducing that his religion requires suicide bombing), but an engineer (who can make deductions just fine) would actually do it.

What you're expressing is your deeply held faith in continued technological progress. You believe that progress is accelerating and that there is no upper bound. How would you defend those beliefs?

The main reason is that these are all multi-layered recursive processes (irrespective of you agreeing or not, check the small sequence of blog posts linked below to understand what this refers to -- detail: the author thinks Kurzweil a pessimist...). Therefore, an upper bound certainly exists, but barred resource limits, until we get there the growth will continue being exponential, not linear.

That prediction about it taking centuries then? Make it "n * log( guess in years )" years and, all things remaining the same, you'll probably be closer to the actual time it'll take, if it'll happen at all. If no "because bounds", then no, but that's another matter entirely.

Part 1: The First World Takeover
Part 2: Life's Story Continues
Part 3: Surprised by Brains
Part 4: Cascades, Cycles, Insight...
Part 5: ...Recursion, Magic

PS.: If the above feels like TL;DR, the important ones are parts 4 and 5. Parts 1 to 3 help 4 and 5 feel more intuitive, but they can be easily skipped.

What you're expressing is your deeply held faith in continued technological progress. You believe that progress is accelerating and that there is no upper bound. How would you defend those beliefs?

The main reason is that these are all multi-layered recursive processes (irrespective of you agreeing or not, check the small sequence of blog posts linked below to understand what this refers to -- detail: the author thinks Kurzweil a pessimist...). Therefore, an upper bound certainly exists, but barred resource limits, until we get there the growth will continue being exponential, not linear.

That prediction about it taking centuries then? Make it "n * log( guess in years )" years and, all things remaining the same, you'll probably be closer to the actual time it'll take, if it'll happen at all. If no "because bounds", then no, but that's another matter entirely.

Part 1: The First World Takeover
Part 2: Life's Story Continues
Part 3: Surprised by Brains
Part 4: Cascades, Cycles, Insight...
Part 5: ...Recursion, Magic

PS.: If the above feels like TL;DR, the important ones are parts 4 and 5. Parts 1 to 3 help 4 and 5 feel more intuitive, but they can be easily skipped.

What you're expressing is your deeply held faith in continued technological progress. You believe that progress is accelerating and that there is no upper bound. How would you defend those beliefs?

The main reason is that these are all multi-layered recursive processes (irrespective of you agreeing or not, check the small sequence of blog posts linked below to understand what this refers to -- detail: the author thinks Kurzweil a pessimist...). Therefore, an upper bound certainly exists, but barred resource limits, until we get there the growth will continue being exponential, not linear.

That prediction about it taking centuries then? Make it "n * log( guess in years )" years and, all things remaining the same, you'll probably be closer to the actual time it'll take, if it'll happen at all. If no "because bounds", then no, but that's another matter entirely.

Part 1: The First World Takeover
Part 2: Life's Story Continues
Part 3: Surprised by Brains
Part 4: Cascades, Cycles, Insight...
Part 5: ...Recursion, Magic

PS.: If the above feels like TL;DR, the important ones are parts 4 and 5. Parts 1 to 3 help 4 and 5 feel more intuitive, but they can be easily skipped.

What you're expressing is your deeply held faith in continued technological progress. You believe that progress is accelerating and that there is no upper bound. How would you defend those beliefs?

The main reason is that these are all multi-layered recursive processes (irrespective of you agreeing or not, check the small sequence of blog posts linked below to understand what this refers to -- detail: the author thinks Kurzweil a pessimist...). Therefore, an upper bound certainly exists, but barred resource limits, until we get there the growth will continue being exponential, not linear.

That prediction about it taking centuries then? Make it "n * log( guess in years )" years and, all things remaining the same, you'll probably be closer to the actual time it'll take, if it'll happen at all. If no "because bounds", then no, but that's another matter entirely.

Part 1: The First World Takeover
Part 2: Life's Story Continues
Part 3: Surprised by Brains
Part 4: Cascades, Cycles, Insight...
Part 5: ...Recursion, Magic

PS.: If the above feels like TL;DR, the important ones are parts 4 and 5. Parts 1 to 3 help 4 and 5 feel more intuitive, but they can be easily skipped.

What you're expressing is your deeply held faith in continued technological progress. You believe that progress is accelerating and that there is no upper bound. How would you defend those beliefs?

The main reason is that these are all multi-layered recursive processes (irrespective of you agreeing or not, check the small sequence of blog posts linked below to understand what this refers to -- detail: the author thinks Kurzweil a pessimist...). Therefore, an upper bound certainly exists, but barred resource limits, until we get there the growth will continue being exponential, not linear.

That prediction about it taking centuries then? Make it "n * log( guess in years )" years and, all things remaining the same, you'll probably be closer to the actual time it'll take, if it'll happen at all. If no "because bounds", then no, but that's another matter entirely.

Part 1: The First World Takeover
Part 2: Life's Story Continues
Part 3: Surprised by Brains
Part 4: Cascades, Cycles, Insight...
Part 5: ...Recursion, Magic

PS.: If the above feels like TL;DR, the important ones are parts 4 and 5. Parts 1 to 3 help 4 and 5 feel more intuitive, but they can be easily skipped.

No. When people are arguing over definitions, the best response is to just drop the terms and look at the actual impacts in question. This is called tabooing your words http://lesswrong.com/lw/nu/taboo_your_words/. Words don't have intrinsic meanings, one can of course discuss specific definitions, but that isn't by itself helpful. One can when people disagree over definitions say things like "This is an example under Definition 1, but not Definition 2" but however you are defining terms like "constitutional Republic" or "Representative democracy" will not in any way shape or form alter the fundamental question: whether people have a say in what our government does.

https://en.wikipedia.org/wiki/...

You know, your use of words of dismissal are mind killers. You would do well to break that habit, and actually read a little into what other people say instead of deriding them.

This is why you can't hold a civil conversation with anyone who believes differently than you do. An educated man can entertain an idea without embracing it. At least that's what Aristotle thought. Perhaps you should endeavor to become educated yourself?

Like most of the people of my generation, I was born into religion (Christian-Southern Baptist, if you must know). However, with the vast sums of knowledge which became available with the information revolution, I quickly learned that the writings in the Bronze Age book were, well, just awful, and completely against the morality I had been raised with. Of course, my family, being Americans, never questioned the fact that our religion was completely at odds with the way we acted. Only in my later years have I come to understand that they didn't REALLY believe.

"You're getting worked up over what you THINK people mean."

No, I get worked up over what people SUBCONSCIOUSLY mean, because it has a serious impact on behavior. People can use "science" as a mind killer. This is unacceptable to me. Politics needs to stay OUT of science, to the greatest extent possible.

"so there it goes the "alien goal""

Your problem is that you don't think hard enough. If no-one did, then the universe would meet a very strange end, tiled with something weird like paperclips.

Read the link, then get back to me. And maybe stop talking about things and people you know nothing about.

That's an excellent example. Gravity is still, in some ways, a mysteroius answer. We know a lot about it, and we can use what we know to make a lot of predictions, but we also know there's a lot that we don't know, and some of our theories about its properties are better termed guesses than predictions.

Now, obviously the EmDrive is far more mysterious than gravity, both because it's much more conceptually novel and unexplored, and because we can really easily detect there's *something* causing the effect called gravity while only a handful of labs around the world are equipped to test the thrust of an EmDrive. I'm not attempting to equate the two. But, as you say, the actual mechanism of gravity has never been observed directly, and the theories about it are still unconfirmed. Similarly, the EmDrive offers some (much less mature) theories as to its operation, but nobody has actually been able to confirm or deny those theories.

Of course, the EmDrive itself hasn't been confirmed yet, at least not to the degree that makes it practical for anything real-world. We have repeatable experiments saying that emitting microwaves into a specially-shaped resonant cavity causes a *tiny* thrust, and we've accounted for some of the likely errors in the experiment (atmosphere, whether the same result happens with a dummy load that doesn't generate the microwaves, etc.), but as of writing this, we have no direct evidence that it scales to useful sizes.

It takes a good stab at examining the challenges and possibilities of superintelligent A.I.

Nice summary view here:
http://lesswrong.com/lw/l4h/su...

It posits three possible intelligence advance scales.

The first is self improvement over seconds.
I.e., the machine become conscious. It is able to increase it's intelligence to superhuman levels at machine speeds within a few seconds. There will be no time to react. Even air gapping the machine might not be sufficient as it may figure out new principles which allow it to bridge the air gap, figure out ways to mislead it's human owners as to it's capabilities so they enhance it further, etc.

The second is over a scale of weeks or months.
Not much time to react to it. A reliable way to cut the power should work. A nuclear safety net should definitely work. Society certainly couldn't react to it in time. There would likely be mass unemployment as it enabled human replacement within a few years for thinking jobs (and combined with robotic bodies- almost all methods of manual labor).

The last way is over a long time period. Society would have time to react. Perhaps to see and stop it if it was turning bad. Especially if it simply became the equivalent of IQ 160-300 slowly, you might be able to understand it. Later phases where it's iq reached meaningless numbers (6000... compared to it, humans would be like horses in relative intelligence).

---

The definitional problem is also there.

"Make people happy".
Okay- rig them to machines that feed them pleasure signals in the brain 24/7. Extinct.

Make people smile!
Easily obtainable with surgery.

---

There is a risk the machine will be "greedy" and basically convert the entire planet (and then the solar system) into a system for increasing it's intelligence. Humans don't play a large part in that scenario. Nothing malicious or personal about it-- not a failure of friendliness.

"rebel"

No, just the opposite. I think a strong AI will carry out its programming to the letter. The problem comes when it is given open ended problems like "maximize the number of paperclips in your collection.

The need to fulfill such a task will drive it towards self improvement and also cause it to eliminate potential threats to its end goal. Threats like, say, all of humanity.

That's pretty much the same thing: http://wiki.lesswrong.com/wiki...

How would you measure if we have "enough" people in music creation? Do numbers count at all? What about quality? How many pop idols would be needed to outweigh a Leonard Bernstein? How many for an Elvis Presley?

I think that framing it in terms of quality of the output is an inherently subjective measure. I'd rather put it in terms of the resources that are getting put into music, and whether they are being used efficiently. So, for example, look at the labor flowing into music: How many kids neglect their studies to pursue a career as a performing artist? How many adults earn above the median wage as a performing artist? If that ratio were 1:1, I would say it would represent a shortage. If that ratio were 10:1, I would say we were in "arguably valid" territory. If it were 100:1, I'd be thinking we're wasting potential from the labor pool (and creating a disenfranchised class of failed rock stars, which are a drag on the economy in other ways). My gut feel is that we're in the 100:1 ballpark or higher.

There's no easy or perfect measure, but if it is important, it can be measured. The trick is to think through the consequences of a distortion, then figure out how to measure for that distortion. How to Measure Anything is an excellent book on the topic.

Food production and water acquisition are high priority. Playing musical instruments or games for other people to watch is lower priority. Who gets paid more? Farmers and well drillers, or rock stars and professional sports players?

The singularity is kinda like that. Maybe. Probably. Lower priority tasks (ie those things that are higher on Maslow's pyramid: http://en.wikipedia.org/wiki/M...) that simply didn't get done before will now get done, and everyone will find it easier to satisfy their values in that world, assuming we haven't all been turned into paperclips.

Watson was impressive on Jeopardy, but a TV show is a very different venue than business data analytics.

For the latter you really need a statistically sound approach in order to reach the right conclusion.

(DISCLAIMER: I do not work for Bayesia, but actually a competitor, yet any person or company that understand Bayesianism as a sound foundation for knowledge inference knows this dirty little secret about Watson)

And still a tremendous waste of money to placard those who fund NASA for emotional reasons, not scientific reasons

Show me evidence that some valuable scientific experiment was bumped from the mission to accommodate this weight, or that a significant sum of much-needed money was diverted from elsewhere, and I'll agree with you.

Meanwhile, I Am Not A Rocket Scientist, but it seems like a no-brainer that you don't design a half-tonne space probe without holding a few grammes of capacity in reserve for contingencies. Something like the ashes could have been bumped at the very last minute with out consequence if the probe weighed in 0.01% over weight.

but if someone did in the far future, they would have to conclude that 21st century humans believed in magic.

No, just that they had emotions and honoured their dead. In fact, you're feeding a Sky Fairy cult strawman (or rather straw Vulcan - see points #4 and #5).

Every dollar spent on space exploration has come back to us tenfold. No other investment in history can make such a boast. Most of your "better ideas" have in fact proven to be losers over time, especially direct economic aid to poorer nations. You talk about setting aside emotional commitments and embracing rationality. Practice what you preach, shut up and multiply.

I don't know what "hate site" means in general, but that's at minimum a source that has very much already decided on their bottom line http://lesswrong.com/lw/js/the_bottom_line/, which means one shoudl already take it pretty skeptically. But that list isn't very helpful for a simple reason that it just shows that there are a lot of Islamic terrorist events which isn't terribly helpful: we already know that. The question being asked is how common are they compared to terrorist events motivated by other ideologies or religious traditions.

Most researchers ( including AI researchers ) dont do engineering very well, and hence dont understand the basic principles that in engineering everything is a tradeoff. Whether you are trading back and forth around physical resources ( flops, bandwidth , latency, memory ) or more abstract constructs like sockets etc everything is still a tradeoff.

There is a theory that there is significant "resource overhang" ( http://wiki.lesswrong.com/wiki... ) in our computers today, meaning that we are not really utilizing the physical resources the best, as more efficient algorithms will simply blaze faster on existing hardware. That is another case of "duh, captain obvious" where they do not understand most basic algorithmic optimizations are ALSO tradeoffs.

Given that many programmers take the "path of least resistence" and don't optimize programs much besides what the compiler does, I would certainly agree that we are not utilizing computers to their full potential. However, an AI would not be able to do so either without being trained how to organize itself better for the hardware it is using. It won't just magically come up with a more efficient algorithm to do what it is doing without being taught something about how to optimize and how the hardware it is using works.

In short, all "hard takeoff" AI scenarios are delusional, and most "moderate takeoff" AI scenarios are misguided. "Soft takeoff" however is happening every day, for example case where genetic algorithms for example are used to work out better solutions to isolated problems than human designer could.

True; though the examples for today - such as the genetic one you quote - are considered "specialist AIs" and in some respects are not much of an AI at all, much like IBM's Deep Blue or Watson computers are not really AI, just very fast depth search algorithms that can mimick an AI to some degree in a very specialized field.

Most researchers ( including AI researchers ) dont do engineering very well, and hence dont understand the basic principles that in engineering everything is a tradeoff. Whether you are trading back and forth around physical resources ( flops, bandwidth , latency, memory ) or more abstract constructs like sockets etc everything is still a tradeoff.

There is a theory that there is significant "resource overhang" ( http://wiki.lesswrong.com/wiki... ) in our computers today, meaning that we are not really utilizing the physical resources the best, as more efficient algorithms will simply blaze faster on existing hardware. That is another case of "duh, captain obvious" where they do not understand most basic algorithmic optimizations are ALSO tradeoffs.

In short, all "hard takeoff" AI scenarios are delusional, and most "moderate takeoff" AI scenarios are misguided. "Soft takeoff" however is happening every day, for example case where genetic algorithms for example are used to work out better solutions to isolated problems than human designer could.

Not at all. However, many people are concerned that this is either the *default* state of an AI - after all, it's being made *by* humans, plus we're the closest thing to super-human intelligence that we have to study - or that it could happen by accident. The canonical example is a "paperclip maximizer": an AI that is smarter than any human but is programmed to hold the paperclip as having the highest possible utility (and therefore being the ultimate goal of all actions). Such an AI would focus on turning all the world either into paperclips or paperclip production. It's an intentionally simple example to highlight the problems with designing an AI that may choose to optimize something - even something believed to be of utility to humanity - without considering *all* of the knock-on effects. If you want a more complex (though still contrived) example, consider this short story by Eliezer Yudkowsky (a researcher on the topic of machine intelligence and friendly AI): http://lesswrong.com/lw/xu/fai...

Strangely, all AI predictions for the last 50 years seem to be set "in 20-30 years" (and were all wrong). I now automatically dismiss any prediction that comes with that kind of time range.

See statistics about http://lesswrong.com/lw/e36/ai_timeline_predictions_are_we_getting_better

AIs can have any number of goals, including just one. The ones with a single goal are the most dangerous, especially if they are strong AI. Maximizing a collection of paperclips might seem like it isn't a big problem, until you realize that humans are made out of atoms that can be turned into paperclips. http://wiki.lesswrong.com/wiki...

Relevant LessWrong: "If Einstein had enough observational evidence to single out the correct equations of General Relativity in the first place, then he probably had enough evidence to be damn sure that General Relativity was true."

Indeed, "be yourself" is just nonsense, as interaction takes at least two people and we cannot define ourselves from nothing and in a vacuum. It's a static statement so it' useless for deciding changes. If you're intereted in the science of behavior I suggest looking at the metaethics sequence from LessWrong, it's a rational approach to rethinking the how and why of choosing and following moral principles.

It's a scary jump from crowd pleaser to taking a stand.

Not so much, actually, as in most people there is at least some latent want for refining trust and reasserting relationships. So there's a kind of continuum from one to the other.

Why the hatred ?

PS: to me you are all just chariot-racing fans.

If growth is a part of fulfilling its value function, the AI will grow.

We must ensure that fulfilling human values is at the core of any strong AI, lest we wind up extinct by paperclip.