HNHacker News
TopNewBestAskShowJobs

Moshe_Silnorin

2,060 karma · joined March 5, 2015

submissionscomments
Moshe_Silnorin··on Open Philanthropy Project awards a grant of $30M to OpenAI
Technological advancement has been responsible for most of the increases in human welfare, most notably allowing us to (at least temporally) escape the Malthusian condition. It's not implausible that technical research likely to be neglected by markets and academia could be a better use of money than even the most efficient African charity. But I drink the AI-is-very-likely-to-change-everything-we-really-mean-it-this-time Koolaid, though.
Moshe_Silnorin··on Tech workers find communal living a solution for high rents
We had such dwellings once. The were declared illegal.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
That is...much better than I would have expected of Vox.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
They really are meaningful. How about we make a deal? You read Intelligence: All That Matters and I'll read The Mismeasure of Man again.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
IQ test are a proxy for g. They are meaningless when applied to computers.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
Not impossible but much, much, much more plausible with embryos. Modifying billions of neurons is a very, very tall order, and would likely be much less effective after the brain is grown. Embryo selection and modification are almost doable now, we just don't know which alleles to select for or edit in. A gift you can give your children but one you are unlikely to be able to get for yourself.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
Nope. Genetic engineering works far, far better. We've already plucked the low hanging fruit of adequate nutrition, early education, salt iodization, and lead reduction. Anyway, those reduce mental retardation, they don't improve the baseline too much. And the Flynn effect looks to be slowing or reversing. 50 IQ points is huge. It's the difference between a burger flipper and a physics professor. The change to the world when the average 5th grader is mastering chromodynamics is hard to even contemplate.
Moshe_Silnorin··on Teenager's sickle cell reversed using a treatment to change his DNA
I agree with this, but not a popular opinion. Hence the down votes you are getting. Korea and China don't have the same memetic immune response to such ideas, injecting one's children with growth hormone in hope of increasing future job prospects, for example, is extremely common in Korea. I suspect embryo selection and generic engineering for increased IQ will take off in East Asia first. Then we will be forced to allow it to compete. 50+ more IQ points looks very possible: https://arxiv.org/abs/1408.3421

However, any significant lag could spell the end of American economic dominance. For this reason, I intend to start investing more in Korean firms in the next few decades.

Moshe_Silnorin··on Should economists be more concerned about AI?
https://en.wikipedia.org/wiki/Moravec's_paradox
Moshe_Silnorin··on Will Democracy Survive Big Data and Artificial Intelligence?
Will my straw shack survive a hurricane? There are far bigger risks with artificial intelligence.
Moshe_Silnorin··on If Susan Can Learn Physics, So Can You (2013)
We have have a personal anecdote on one side, and modern psychometrics on the other:http://infoproc.blogspot.ca/2010/05/psychometric-thresholds-...

I'll go with psychometrics on this one.

Moshe_Silnorin··on Harvard and M.I.T. Scientists Win Gene-Editing Patent Fight
It may give our institutions time to figure our how to cope.
Moshe_Silnorin··on Harvard and M.I.T. Scientists Win Gene-Editing Patent Fight
It's polygenic, like height. People who I read in the space of cognitive genomics, such as Steve Hsu and Razib Khan, agree it's looking increasingly plausible. It would require modifying hundreds of alleles. Embryo selection is possible now, just not useful until we have a million genomes tagged with their donors' IQ. 1 of 10 embryo selection give you about 8 IQ points per generation. Hsu thinks Crispr will obsolete embryo selection before it's in wide use.

He also said the following:

>I think there is good evidence that existing genetic variants in the human population (i.e., alleles affecting intelligence that are found today in the collective world population, but not necessarily in a single person) can be combined to produce a phenotype which is far beyond anything yet seen in human history. This would not surprise an animal or plant breeder — experiments on corn, cows, chickens, drosophila, etc. have shifted population means by many standard deviations (e.g., +30 SD in the case of corn).

Moshe_Silnorin··on Harvard and M.I.T. Scientists Win Gene-Editing Patent Fight
The patent system remains absurd.

I have mixed feeling about biotechnology, though. On one hand I'm a big fan of human genetic engineering for increased IQ which is looking extremely possible in the near future (hopefully something that will be available here in Singapore before I have kids), on the other hand as these techniques get cheaper and easier to use we may get to the point where any madman with an IQ over 145 could kill hundreds of millions. Nick Bostrom asks this question in his book: if building a nuclear bomb turned out to be trivial, like putting sand in a microwave, would human civilization have survived? We will find out if synthetic biology keeps advancing. Or maybe this is my nightmare and the patent system will slow down progress enough to save humanity.

Moshe_Silnorin··on Mathematical genius is fragile
https://en.wikipedia.org/wiki/Raven's_Progressive_Matrices
Moshe_Silnorin··on Mathematical genius is fragile
Why?
Moshe_Silnorin··on Mathematical genius is fragile
Math is a high IQ game. Anyone who tells you otherwise is kidding themselves.
Moshe_Silnorin··on [dead]
He does at public universities.
Moshe_Silnorin··on Deep learning algorithm diagnoses skin cancer as well as seasoned dermatologists
Yes, you are safe because of licensure and regulatory capture. Hardly seems fair to the factory workers.
Moshe_Silnorin··on Basic income is the worst response to automation
The upper classes do have higher average IQ, an effect which is more pronounced the more meritocratic the society. Yes, there are lots of stupid rich people and tons of smart impoverished people, but we are talking about averages here. As most people desire wealth and intelligence is your capacity to realize your desires, intelligent people acquire more wealth than less-intelligent people. As intelligence is very heritable, this has unfortunate class implications. I’m not endorsing this description as normative, I just think it is accurate. This state of affaires is allowed to be both true and objectionable. Natural selection is a horrible process that runs on death and suffering, it would be very odd if it shaped our variations in a manner that matches our ideological preferences.

> Programming is nothing but an area of work which is more suited to people with a certain analytical skill sets.

If you think you can turn 100 IQ people into productive programmers, you could make billions by training your own staff and undercutting everyone else. Give it a try.

Moshe_Silnorin··on Basic income is the worst response to automation
It seems to have stopped in developed countries: https://www.gwern.net/docs/iq/2016-dutton.pdf
Moshe_Silnorin··on Basic income is the worst response to automation
It seems clear to me that there is an IQ cutoff below which one cannot be an effective programmer. And this cutoff is actually quite high, comparatively. I'd be very surprised if more than 5 percent of the population, with the current distribution of intelligence, could become effective programmers, even if highly motivated. All psychometric evidence points to IQ being largely genetic, and the environmental components irreversible after after a very young age. It's odd, but my very "conservative" view on intelligence and heredity, that of agreeing with the scientific consensus, leads me to support some forms of wealth distribution.
Moshe_Silnorin··on ‘Routine’ Jobs Are Disappearing
Universalize that statement and apply it to X demographic not choosing a STEM degree and you'll find yourself in a struggle session.
Moshe_Silnorin··on Unexpected Risks Found in Editing Genes to Prevent Inherited Disorders
>Not if the people with the degenerate version also have another mutation that interacts.

Is this common? An uncommon detrimental variant that is slightly-mitigated by another uncommon detrimental variant, such that the removal of one makes the other worse isn't the type of thing I would expect to come about through natural selection.

Moshe_Silnorin··on Unexpected Risks Found in Editing Genes to Prevent Inherited Disorders
If we replace degenerate alleles with the most-common version of that allele, isn't that pretty safe? It would have been tested on billions of people naturally.
Moshe_Silnorin··on Deep Learning 2016: The Year in Review
The human brain has 100 trillion synapses. So it would be 17 iterations of Moore's law for computing power to increase 100000x. That's what I was thinking. Wrong to assume NNs are no more efficient than brains. You know way more about this than me, so I'm just going to replace my opinion on this with your own. If you think we can chop off 4 orders without the help of Moore, I believe you, though this is much less comforting than my prediction!

>What do bees do that is remotely as demanding as being able to translate between English, French, German, Japanese etc at very high quality?

This seems more impressive than anything any one NN can do, but less impressive than what all NNs can do: https://en.wikipedia.org/wiki/Bee_learning_and_communication

Moshe_Silnorin··on So far, equity crowdfunding hasn't matched the hype
That's fraud not crowdfunding. If it's a sincere attempt to start a business, crowdfunding seems like a fine idea. Though I agree, it is a lemon market and I wouldn't participate myself. Many people will lose money doing this, but the cost of freedom is some people will make poor investments.
Moshe_Silnorin··on So far, equity crowdfunding hasn't matched the hype
People have the right to make poor investments.
Moshe_Silnorin··on Life in Life (2012) [video]
Watching this is the closest thing I've had to a religious experience.
Moshe_Silnorin··on Deep Learning 2016: The Year in Review
Naive trend-extrapolation (assuming 1 synapse == 1 parameter) implies 17 iterations of Moore's law before we get human level AI, or some sort of ASIC, I suppose. This is what made me think human-level AI is a ways off.

Brain dead trend-extrapolation is a pretty good prediction tool, but I suppose my assumption that modern NNs aren't more efficient was sort of an arbitrary assumption that makes the exercise useless. Is there any work on this?

I'm not sure what evidence the performance of these systems show us, as bees are pretty smart for their size. These system are about as optimized for lip-reading and translation as bees are for their various tasks, so it wouldn't be surprise me if it turned out that they are similarly powerful per synapse to NNs.

← PreviousPage 3 of 12Next →