stanfordkid

Spiky things hurt and round things don’t — would be interesting to simulate pleasure / pain of shapes against a virtual hand using a simulation then train a deep net to learn the pain response. Bet you would have similar findings.

It gets even crazier — there are people literally buying new paintings at higher prices to inflate the value of existing holdings. So let’s say you buy 3 Vassarely’s @ $200k each (or insert semi-hot artist here). You wait 4 years and have someone else buy a 4th Vassarely piece for 3x that price point. Now you use the inflated value for that single piece as a comp for the other 3 pieces that will then be donated. Art market reps basically find ways to facilitate these types of things.

This is why you always see wings of museums named after rich folks.

I agree it is too late to enact things like a lower minimum wage — the damage is already done.

The real problem came from cheap consumer credit... everything in our economy is built around the existence of such credit. Consumers are conditioned and targeted to consume and borrow. Full stop.

Same thing happened with education — 4 years of productive time were siphoned towards college degrees that didn’t mean much.

This is the gutting of the middle class... not fiscal conservatism.

The decreases in price for these consumer goods you mention came largely as a result of globalization and offshoring of production. Our economy went upmarket towards services which were largely meant to facilitate this consumption (eg advertising and mass media). The American public became lazy, conditioned to consume and entitled.

The modern American consumer is only in debt because debt is so easy to obtain, not because his wages are too low to survive.

Visit Latin America or Asia where there throngs of people much happier and healthier on far less. The welfare state makes the familial structure unnecessary in America — I won’t comment on this. But it is a third order effect even more deeply troubling than the economic ones I highlighted earlier.

Blindly investing money into “the poor” is a foolish endeavor that yields little strength to a country. Herd immunity and preventative health is a quantifiable thing and so I support universal healthcare. But we have to stop this disasterous assumption that the only reason people are poor is their circumstances.

The best way to help poor people is to lower the minimum wage so they can get jobs.

Subsidizing the poor increases consumption which is bad not only for the first order effect (throwing away money)... but also for the little thought of second order effect: it incentivizes people to build businesses that cater to consumption that was in the first place unsustainable. Smart people that could have been doing something else instead facilitate the creation of businesses that are only funded by inflation and cheap credit.

This is the story of the US economy in the past 30 years. The two largest disasters in recent history: the student loan bubble and the housing bubble, were both founded upon the notion that we should “give money away” without any real measurement of whether or not there would be a return on capital.

Our government should invest the majority of capital in basic research, healthcare, education and infrastructure... not the creation of a consumption driven bloated welfare state.

The cheat code:

— buy painting for $5 million

— have a museum value it at $15 million a few years later

— donate painting

— write off $15 million in taxes for a charitable contribution

Clearly the museums have an incentive to fudge the numbers for rich donors.

EDIT: Why height is relevant : "Why study height? Height is the “poster child” of complex genetic traits, meaning it is influenced by multiple genetic variants working together. It’s easy to measure, so makes a relatively simple model for understanding traits produced by not one gene, but many."

The complex system argument is a strawman.

It’s well understood that complex traits are polygenic (no single gene or sets of causal variants) but quite easy to predict such traits when looking at the full genome (I agree the resolution of prediction is still not there — but we are in the infancy of this field).

I recommend learning more about some of the polygenic risk scoring methodologies that are popping up in research: https://www.broadinstitute.org/news/giant-study-finds-rare-i...

A mega corporation with billions of dollars and thousands of workers is going to be held to different moral standard than a single person buying a camera lens. Clearly society doesn't judge 4 year olds to the same moral standard as 50 year old men. You'd laugh at a 50 year old man crying over spilling his juice ... but not a 4 year old. So why should we judge individuals to the same moral standards as mega-corporations?

The relative value of $500 for an individual is orders of magnitude higher than $13k for Amazon.

This guy needs to read the Tibetan book of the dead. People in the past have thought long and hard about consciousness... and actually experienced alternate states similar to death. We just choose to treat them as loonies in the west. Who knows what happens to your consciousness in that frozen state. You might just be paralyzed the whole time.

The analogy to jumping out of an airplane is apples and oranges, because you are guaranteed life or death within the current epoch in the case of the airplane jump — both states that we understand. But you have no such guarantee in the cryogenic state. You might wake up to a horrible world.

The entire technology revolution is based around the idea that "if we have more of something we find new applications"... You wouldn't have an iPhone if compute power was as rare as gold.

The whole article is written with the tone of someone who doesn't understand the concept of "unknown unknowns".

MCTS estimates a score for each action by stochastically sampling the future states. It has nothing to do with a problem such as this with continuous inputs and outputs.

Can't use my Nest lock to let guests into my house. I'm pretty sure their infrastructure is hosted in Google Cloud. So yeah... definitely some stuff lost.

Disagree. Google just had way better results than every other search engine. Secondly, it had a much more streamlined and simple interface (they had a simple search box with a "Search" Button -- no page directory, no email, no banner ads -- during that time period search engines had giant directories of pages by category when you first opened them up) . People would often refer novice users to Google because there were less points of confusion. By the time the company IPO'd they were already on an exponential tear.

Those existing buildings have people inside. Pulling them out to demolish the buildings and add new, taller, skyscrapers has significant costs I think you aren't factoring in. Like all those businesses inside have to relocate... which costs money.

Well the simplest explanation is the classic nature vs. nurture, paired with selective breeding. If you are raised by hippie shroom eating chimp parents, you'll probably be attracted to a different sort of mate than if you are raised in a warrior clan. This sort of taming can happen within 3-4 generations and have permanent effects.

Dmitry Belyayev was able to produce a tame variety of the red fox within 4 generations of breeding: https://en.wikipedia.org/wiki/Dmitry_Belyayev_(zoologist)#Be...

Shouldn't college admissions officers already be able to take this into account based on data in the application? I am pretty neutral on it, but think the idea is pretty pointless and just leads to people trying to game the system, rather than having decision making criterion distributed.

I agree with your statement. The core machinery for transcription and translation is still the same -- I will be impressed when a synthetically designed ribosome or DNA/RNA polymerase is created.

The redundancy is probably there to so that the DNA->RNA copying is more robust against chance mutations in DNA that occurred during cell division.

This is intriguing -- I would guess that a successful strategy would be to short stocks with high growth that have a CEO with "evolutionarily advantageous" traits (height, attractiveness, depth of voice) ... and go long stocks with high growth that have CEO's lacking these characteristics:

The company with the "suave" CEO is likely building revenue as a result of sales and people charisma rather than core innovation.

It's intriguing you bring up Jung. But what you are saying (that the removal of privacy helps integrate the shadow and thus "fixes jungian" duality in humans) -- is not at all what actually happens.

Our true shadow is not what Facebook exposes. Rather it collects data and exposes a subset which is in fact the anti-shadow.

YOU choose what data to expose on facebook. The act of that data being exposed has effects that are tangible.

If you've read Foucault he describes a similar effect during the Victorian era on human sexuality -- the idea that by exposing something and quantifying it -- you in fact put a sort of control on it. When it was hidden, no one knew what was normal -- so they used their intuitions and personal judgement.

It doesn't share everything -- because it can't possibly KNOW you (or your shadow) -- people are incredibly smart and complex -- as you point out. But what it does do is expose a superficial subset of you that in turn changes your behavior and the perception of your behavior.

Jungs whole point is that the shadow is an intrinsic aspect of any social society -- something MUST be repressed in order for us to not be completely individualistic psychopaths (like the snakes we evolved from) ... simply exposing this duality doesn't just "fix it" ... instead society needs to function with the knowledge that these things exist. Jung's point is that the individual should not go into psychosis trying to balance dualistic morals... and society shouldn't be so harsh in judging morals (NOT that everything in the shadow should be made public, or even that the shadow is a bug rather than a feature)

It's really not. You gamble with your margins -- it's a continuous variable. You could be profitable by a cent or millions of dollars, or not at all. Each of these decisions yields a very different growth and risk trajectory.

The label of "profitable" itself is pretty meaningless, given that you are at battle over many many years against competitors. One quarter of profitability might yield doom 3 quarters down the line due to lack of investment.

I disagree with this viewpoint. At the end of the day wealth needs to be distributed to people in proportion with their ability to allocate it effectively towards the growth in the knowledge and capabilities of a society. In other words -- we want smart people that want to improve society to have lots of money.

Dynastic wealth needs to be paired with a code of ethics and training ... nothing inherently wrong with it.

Junk DNA is a phrase biologists use for describing the long held (and recently disproven) idea that non-coding DNA ("dna not coding for proteins") is unimportant relative to coding regions (genes).

Most non-coding DNA does have a regulatory purpose. Things like copy number variations (where information is "duplicated") between promoters and enhancers can can change the rate at which particular genes get transcribed. This is my point: that viewing the genome as "information" in the simple sense of each ATCG nucleotide corresponding to bits is absurd. The genome is a geometric AND chemical entity and encodes information using both paradigms.

My point is not to say that you are wrong, but rather that you do not know. It is FAR from well established that the genome is inefficient at storing information.

That's a pretty cursory understanding of how DNA works. The "Junk DNA" hypothesis has long been invalidated.

The non-coding genome is a large part of what causes us to look different from a mouse or a fruit-fly. Gene regulation (what gene to transcribe, and when) is largely controlled by non-coding DNA: https://en.wikipedia.org/wiki/Regulation_of_gene_expression

There are energy costs associated with useless information being stored and transcribed in the genome... so it is pretty reasonable to believe that information theory does play some (however minor) role in how data is stored.

Explain.

Totally. Just the HUD systems on the F-35 cost tens of millions of dollars.

My thought is that there is just ZERO innovation in this area at a big picture level. Sure the person designing the radar has a PhD from Caltech -- but it's mostly incremental optimization of individual components.

Someone is going to figure out how to be the Elon Musk of fighter jets and we better hope it's going to be someone from the USA.

But again -- it doesn't matter how good your tech is if the next person can do it 85% as good for 20% the price.

My last major point is this: Fighter jets are limited in their maneuverability when they have humans on-board -- they can't pull more G's than the body can withstand. AI fighters don't have this limitation.

I do worry that there is opportunity for massive "disruption" for fighter jets in precisely the Clayton Christensen / Innovator's dilemma sense.

The innovation shouldn't be around stealth and maneuverability, but rather production cost/rate and AI automation. It is really hard to train a fighter pilot, this means you have very few of them, and the one with the best jet wins.

In the new world of AI the jets themselves are trivial, it's the AI that matters and the VOLUME you can produce them at. In this case someone like China is way better positioned -- just steal the IP for the physical airframe (or develop one that is just good enough) then innovate on the assembly line and AI aspects.

The AI fighter jet problem feels much easier than the self driving car problem.

I think that's a bit reductionist. Highly recommend reading the Muqaddimah (one of the first sociology books) which talks of a natural cadence to power: http://www.politics.ankara.edu.tr/dergi/pdf/39/1/13_barbara_...

Nietzsche also delves into some of these ideas when he describes the rise and fall of civilizations as a by product of decadence. This decadence is not just the simplistic notion of "consumption", but rather a weak moral system that arises as a result of decadence. It is not consumption, but rather the morality of consumption: https://www.reddit.com/r/askphilosophy/comments/5avje6/what_...

Looting machines (like companies) required skill, fearlessness, and capital to build.

> I feel like when ever you need to put that much money into ?an idea, it’s likely not an effecient use of money. For instance, if you invest in 200 startups at $5m each, you’re more likely to get better returns, than betting almost $1b on a single company. There’s higher risk with the one company (probably).

There are fewer entities that can place bets like that, which means that it is unexplored and likely has more opportunities. Seed funds with $20mm are a dime a dozen.

With this being said the failures are much more catastrophic when such investments are done blindly, without fundamental market / technology forces (e.g take a look at Theranos or Better Place). I believe this is what you are alluding to.

Good investments in this $100mm + category are precisely justified in terms of demand and precisely justified in proof of technology execution -- they are largely for scaling technology to fit a known demand.

A fundamental paradigm shift happened in the past decade which Peter Thiel describes as "moving from the world of bits to the world of atoms" ... this necessitates larger capex to win. With software the scaling costs are constant, with atoms -- they become linear.

This paradigm shift was driven by mobile technology and AI ... so the world is completely different from the days when spreading 10 $1mm bets was a solid strategy. One of the fundamental problems with applying past intelligence is that it was built on top of a different world.

Physicists don't have any imagination to break past certain assumptions (e.g wave function collapse is random)

I liked Stephen Wolframs ideas around cellular automata and network based structures -- it seems that if we could look at physics as more of an information transfer problem -- paired with network optimization and information storage optimization, on a network: this is where real progress will be made. There have been theories around emergent spacetime from thermodynamic processes that were abandoned and IMO need to be re-investigated.

I highly doubt that wave function collapse is just completely random. We just don't see the structure that is creating the "random" distributions we observe.