wycs

AI is general use. This is a good bet.

AI actually makes money now. There will be no more AI winters as it can now support its own development without patronage.

The more full the refrigerator the longer it will maintain temperature.

Yes. Keeping food in a fridge for 48 hours. What a crime.

These people built a business that provided a lot of value and a bunch of bureaucrats killed it because they didn’t like the thought of something being in a fridge for less than 48 hours.

This is the message of these regulations: don’t try to do anything in the real world. Solve only ephemeral, virtual problems.

This guy is working on a travel information app now. He has gotten the message.

>Could be legit, especially as they're not creating their own token for this purpose.

Maybe I am confusing them for another project, but I am pretty sure they do have a token model.

I'm not fond of the economics of Golem. I suspect they will not be able to compete with centralized solutions, as they have to support all sorts of redundancy to make things trustless. They could make it more federated, but then how is that better than AWS? And why do you need a token at all?

Our revealed preferences often paint a bleak portrait.

Man does this motivate me to start saving more for retirement. No one and nothing will be able to care for you more than your past selves.

See gwern's write up: https://www.gwern.net/Embryo-selection

We can predict roughly 16% of the variance in IQ.

I don't think we can rule out an evolutionary explanation so quickly. We are starting to get firm genetic evidence that something like dysgenics is occurring in Iceland, for example:http://www.pnas.org/content/early/2017/01/10/1612113114

There is also a plausible mechanism of action, intelligent people tend to get an education proportionate to their intelligence. Educated people have ridiculously low birth rates. Intelligence is heritable.

Just like conservatives will have to overcome their global warming denial, those of the progressive tradition may well have to start accepting what cognitive genomics is making increasingly obvious: intelligence is hereditary and this has policy implications.

Should it prove to be a genetic decline, we have solutions, such as embryo selection, in the wings that can work now with little research. This will get us ~3-8 IQ points per generation:https://www.gwern.net/Embryo-selection

Iterated embryo selection and genetic engineering will be able to offer much more, several standard deviations in multiple traits simultaneously.

But we cannot repair problems we refuse to look in the face.

Rent is cheap.

I don’t think you finished the essay. It is very clear they don’t believe the policy proposed can be implemented and it is only slightly less clear that they think it is undesirable.

I'm not sure if it would be stable or not. After a couple generations I sort of feel it would be, as even if you hate your parents if you have kids you would still benefit from the policy. In fact, if economic growth continues you would benefit more than your parents did. And most people do love their parents.

You don't. You wait until they grow up and then you give the parents a commission on their income tax.

>So there is an unsolvable problem here, you have to compromise one way or the other.

This seems addressable by biotechnology, artificial wombs and so on. At a certain level of technological sophistication, all problems become engineering problems.

Children are public goods with massive positive externalities. I think it would be a shame if the population declined.

My favourite policy to increase birth rates is to give parents a percentage of their children's income tax revenue. Standard econ says we just need to internalize some of these positive externalities. This has the advantage of scaling with the wealth of the parents, as economic productivity has a significant heritable component.

Though this policy triggers an intense moral disgust response in everyone I've proposed it to, so likely not possible unless things get very desperate. But I think it has a chance of working, which is more than I can say of most natalist policies I have seen.

I thought it was a good movie but morally backwards. Once we have the ability to select against alleles that decrease happiness, intelligence, health, sanity, and longevity we are are morally obliged to do so. His mother’s choice was as monstrous as refusing to let your child learn to read.

Would you subject your child to a 1 in 1000 chance of being brutally disfigured? Well genetic defects that cause much the same thing are more common than that.

The use of genetic measures over measures of the phenotype in that movie was silly. Cognitive genomics will never be as accurate as an IQ test. It would be a mistake to, for example, replace the SAT with a spit swab, even in a society with genetic engineering.

I agree humans are capable of learning, but the growth mindset set claim people are capable of learning how to learn, for which there is little evidence.

The rate and extent of knowledge acquisition is limited by one’s general intelligence. You can not learn your way around a less effective learning apparatus.

I would also say that we have plucked all the environmental low hanging fruit. Lack of education may have explained much of the variance in outcomes 200 years ago. It does not today. We are educated to the gills. Dweck’s claims that we can squeeze more out of this desiccated rag do not ring true, even before you account for how poorly ideologically-pleasing findings have held up in the replication crisis.

I don’t think growth mindset interventions are going to do much of anything.

Programs like Head Start have been spectacular failures, which does not seem to be what growth mindset would predict.

Ultimately we are too focused on environment and have no focus on genomics.

Embryo selection will be far more effective than propagandizing our children with growth mindset memes, whose workbooks read to me like modern copybook headings.

1-10 embryo selection gives you 3 to 8 points per generation. This is basically possible right now. If we had a million genomes tagged with the donor’s IQ this would for sure be possible right now.

Iterated embryo selection and direct genetic engineering allow get you to many standard deviations.

Growth mindset is sneezing into a hurricane. And I am not even sure the sneeze is real.

It is startling how ineffective basically all educational interventions are. I remember reading a quote to this effect in the introduction to the Decline and Fall of the Roman Empire:

“Teaching is a skill we have not yet mastered, save for to those happy people for whom it is almost superfluous.”

We have known intelligence is heritable for a long time. Anyone who has lived for a great length of time in a small town has seen this in their neighbours.

We have had twin studies basically confirming it for ages; and soon cognitive genomics will confirm it beyond the smallest shadow of a doubt.

It is a testament to what epistemic hacks humans are that we allowed these extreme priors to be overwhelmed by empty political fashion.

>Vote for people who will.

It really, truly is not that simple:https://en.wikipedia.org/wiki/Public_choice

You would need Wechat for America. I actually think Apple is well placed to do this, but it would go against their hardware-first culture.

If they released iMessage and ApplePay on Android, added a Wechat style P2P payment system, and started cannibalizing the AppStore/Playstore through a WeChat style API I think they could pull it off.

I hope to God I have the right to suicide by the time that happens to me.

It is about the tragedy of the commons and what it implies when self-modification becomes possible, such as genetic engineering in humans and recursive self-improvement in potential future AIs.

That essay is a hell of a piece or writing. It didn't so much change my object-level policy opinions for the here and now, but it ripped the heart out of my Libertarianism. I've always found markets and selection beautiful and fascinating and powerful, which is all true. But I also thought they had a teleology that was humanistic. This is false.

To the extent this seems true now, it is only because humans are the current hardware markets are running on. Fundamentally, they are not our side in the long term.

I have found few pieces of writing more powerful:

http://slatestarcodex.com/2014/07/30/meditations-on-moloch/

Nick Bostrom explores similar themes in The Future of Human Evolution: https://nickbostrom.com/fut/evolution.html

It should be noted that some people don't really want glory or fame, and are perfectly happy just becoming filthy rich. I think of fame, in this age of online mobs, as a curse, a sort of disfigurement that lives in the minds of other people.

The scope of engineering talent in China is hard to contemplate. Because of the, on average, higher quantitative talent of Hans Chinese over gentile whites and much larger population, China has, on average, at least 10 times the number of people capable of high-level engineering. This is not a false alarm like India. India's talent pool is proving much shallower than people expected - though not to the people politically incorrect enough to keep an eye on PISA scores.

This just is not the case with China. Chinese human capital is more plentiful and of higher quality than that of any other country on earth. An additional multiplier is an almost fanatical work culture that is comparable to that of Japan.

Combine this with local governments that are as competent as the Bay Area government is incompetent, and there is finally a threat to America's domination of tech. My money is on China, or it would be if it was possible to get exposure.

I sort of wonder what the comparative advantage of the west will be when this all shakes out.

This old comment from nopinsight goes through the scores:

Key competitive advantages of China are their strength in quantitative skills as well as huge population and the hard working and competitive culture of its populace. An objective measure is PISA results [1]. When comparing with even the best performing US state, Massachusetts, China has many more top performers in Math, as a proportion of population [2].

(In 2015 only four provinces of China participated, but their combined population was 230 million vs Massachusetts's 6.8 million. The math result of Shanghai (24 million pop.) alone would show an even larger gap.)

Since PISA results are scaled such that OECD country's mean is 500 and standard deviation is 100. China's 531 math score implies country mean at 0.3 SD above PISA mean, and US' math score at 470 implies 0.3 SD below mean. If people capable of doing AI research or proper AI implementation need to have math skills at, say, 2 SD above PISA mean, then there will be a tremendous difference in proportion between two populations with 0.6 SD difference. My back-of-the-envelope calculation, assuming above figures, is the difference in proportion will be about five times. But China has more than 4 times the population of the US, so the difference in potential numbers of AI-capable natives could be over 20 times. (Since other provinces may drag down China's mean, it could be a bit less. We'll see soon since China as a whole will participate in 2018.)

2%. Seems crazy low to some, but I put very conservative numbers in my MonteCarlo simulations.

To match the performance of a market economy, wouldn't that require modelling agents that themselves have access to the same computing techniques you have?

The most buzzwordy thing I heard from a blockchain guy was from Vinay Gupta, of Ethereum fame. He said quantum computers can solve the economic calculation problem, so we can get Quantum/Blockchain/AI socialist central planning.

I always feel like people are trying to launder blockchain, which appears to me to be mostly useless, with AI, which is highly useful. I fail to see how the two technologies are related in any way.