wycs

This will be my attitude until we are all replaced by pirated brain emulations of Linus Torvalds.

If he wanted to prove his thesis, he should have adopted.

>I wonder what are the implications for healthy brains. If humans are born with 80 billion neurons and you can increase that number

Or do some somatic gene therapy before injection. Seems like about the only way to increase IQ after birth, short of an extremely advanced neural prothesis.

All these things get something in return. For all thats been spent on mining, we are getting 7 transactions per second in exchange. It is absurd.

No. I think the claim "one can raise IQ several standard deviations by editing <200 alleles" is likely true:https://arxiv.org/pdf/1408.3421.pdf

There is no human with all the positive variants; it is the the distribution of these variants that explains 5.2%. GM would allow you to create such a human.

That paper does not imply what you think it implies: that editing all these variants in an individual embryo would have a small effect.

An individual with the top 200 IQ-enhancing variants edited in (and an otherwise normal genome) would be far, far smarter than than than you seem to think, even though the variants edited only explain 5.2% of the differences within the population.

I am aware of this. Genetic engineering implies far more power than Gwern mentions for one round of embryo selection. 5 standard deviations may sound a lot for you, but in animal breeding we have changed the distribution of various traits by far more than 5 standard deviations. Into the dozens.

Most likely because it seems to be a phenotype that does not reproduce itself. None of Von Neumann’s children had kids if I recall correctly. This environment does not select for IQ past a certain point; this does not imply we cannot engineer people to be more intelligent any more than it implies we cannot engineer people to be dwarfs.

Though thousands of alleles contribute to intelligence, we would only need to edit a few hundred to get extremely large increases in IQ, especially if you perform a round of embryo selection (or even iterated embryo selection) before editing.

Editing hundreds of alleles may sound incredible to you now, but I don’t think it will sound so incredible for long.

Even ignoring GM, embryo selection can get us pretty far with current technology: https://www.gwern.net/Embryo-selection

This is not true. Intelligence is polygenic. The average human is no Von Neumann. At the very least genetic engineering could raise average IQ to Von Neumann’s level. We know that phenotype is possible with no ill effects. It will likely take us much further.

In the short term I think the risks are overstated. Vastly increased health and intelligence looks like it will be possible to select and/or edit for, as well as the elimination of Mendelian disease. The scale of this good is so large it overwhelms most scruples.

Perhaps this is true. I do not know. But human enhancement through genetic engineering is coming. Changes in traits, such as IQ, by 5+ standard deviations looks possible. We would be fools to outlaw it here, for we would become as children compared to future citizens of nations that do not share our taboos.

Few people have not heard of crypto. If it is a ponzi there are no suckers left.

China is full of extremely bright, hardworking people. It is not a nation of brain-washed automatons. I think AI alignment is some of the most important work in the world, and I think spreading and translating AI alignment research in China, such as Stuart Russell's work, may well be more important than direct AI-safety research. Because of their huge population and much higher average quantitative ability among Han Chinese, China has roughly 20 times the number of people capable of AI research, at least if you trust their PISA scores.

Should we get human-level AI, there is a good chance it will be made in China. As I don't think the Chinese are a nation of heartless automatons, I think spreading alignment ideas in China is not at all a hopeless task.

This is no longer the case; human computer hybrids are no longer competitive in chess.

From Wikipedia:

Computer chess engine Zor won the freestyle 2017 Ultimate Challenge tournament. The best human plus computer came in 3rd place. In 2017, chess computer engines are superior to human plus computers.

Web development feels AI complete to me. In the sense that it would take human level AI to do. However, chess felt AI complete to Kasparov.

You are right. I don't think I read his comment very carefully before replying.

>He says general AI is on the same spectrum as the AI technologies we have now, but is qualitatively different.

No it is not. The basic premise of fixed-winged aircraft was the same from Wright brothers to modern jets. Yet the wright brothers flyer was useless and a modern jet is not.

We have agents that can act in environments. His claim is that getting these agents to human-level intelligence is a matter of compute and architectural advancements that are not qualitatively different that what we have now. This just does not strike me as an absurd claim. We have systems that can learn reasonably robustly. We should accord significant probability to the claim that higher-level reasoning and perception can be learned with these same tools given enough computing power.

He claims we cannot "rule out" near-term AGI. Let's define "rule out" as having a probability of 1% or lower. I think he's given pretty good reasons to up our probability to between 2-10%. For myself, 10-20% seems a reasonable range.

I think the Chinese are harder working and seemingly smarter if your trust their PISA scores. I don't think there is any measurable trait of the quality of human capital in which Chinese workers do worse than Americans.

Though I agree with the normative claim behind what you are saying, it is not descriptive. The economy should exist for the people; it does not now. And making it do so will be very difficult, short of the invention of Culture style Minds.

Imagine I said the following: "The ecosystem exists for the welfare of the animals." This would strike you as absurd.

The economy exits because of people; it does not exist for people. In the limit of perfect competition, we should expect a race to the bottom, provided imperfect coordination.

>...Then he basically asserts that we can extrapolate the near-term availability of tons more compute power from Moore’s Law, which is where he lost me.

That's not what he's asserting. Even with Moore's law dead, OpenAI claims there is significant room with ASICs, analog computing, and simply throwing more money at the problem. There is a ton of low-hanging fruit in Non-Von Neumann architectures. We should expect it to be plucked, as we have huge use case which is potentially limitlessly profitable.

Much of the value is standby time. Even if workers are doing very little most of the day, the fact that you can spin up a worker instantly (between 9 and 9) in China can make their companies extremely efficient.

I don't know. Sometimes it is hard to argue with results. The pace of Chinese technological capitalism is staggering. Perhaps it is just a race to the bottom with working hours.

Then use your exit rights and don’t participate.

>But like taxes, death is inevitable.

Taxes yes, but I am not so sure about death.

>The sooner the Baby Boomer stop being consumers, the better, in regards to the rest of society.

To make your comment explicit, you are cheering on the death and decrepitude of millions of people. I hope your children are kinder to you.

You say that like vastly increasing the productivity of our largest cities is neutral at best. This would create an extraordinary surplus, and other cities that can build out would compete for citizens, just like Texas is doing right now.

The solution is to get rid of zoning laws. Baring some breakthrough in public choice theory allowing us to break out of this mess, I predict we are stuck with this until the robots take over.

Yes. It is all very obvious. For example, suppose a marginalized group uses the hashtag #killallwhitemen to express the disempowerment they feel from the historical weight of white privilege. This, I think we can all agree, is an expression of righteous outrage through hyperbole, not intolerance. In fact, "white men" in this context does not refer to any person or group of people, rather a self-reenforcing system of oppression and privilege. In this way, we see "#killallwhitemen" is tool of enforcing tolerance in an intolerant society.

Now suppose someone were to make an argument like this:

"We observe sexual dimorphism to some degree in most animals. Thus our prior should be that it is unlikely that women and men are 100% psychologically identical even when controlling for environment. Because of this it would be epistemically incorrect to assume differences in representation in engineering firms can be explained entirely by discrimination."

This argument is clearly reactionary and must be censored, for we know 100% that any difference in preferences, abilities and outcomes between men and women must be explained by environmental effects. To think otherwise would be exclusionary and thus an example of intolerance we must be intolerant of, even in our own thoughts.

The person making this argument should be shunned. Even raising the argument should result in immediate deplatforming and perhaps loss of employability.

Even this notion of a prior in which you allow yourself to assign a probability to reactionary thoughts must fought, for even a very low prior can be overcome with enough sensory experience.