August 8, 2017

"Symbiotic relationships, both parasitic and mutualistic, are ubiquitous in nature."

"Understanding how these symbioses evolve, from bacteria and their phages to humans and our gut microflora, is crucial in understanding how life operates. Often, symbioses consist of a slowly evolving host species with each host only interacting with its own sub-population of symbionts. The Red Queen hypothesis describes coevolutionary relationships as constant arms races with each species rushing to evolve an advantage over the other, suggesting that faster evolution is favored. Here, we use a simple game theoretic model of host-symbiont coevolution that includes population structure to show that if the symbionts evolve much faster than the host, the equilibrium distribution is the same as it would be if it were a sequential game where the host moves first against its symbionts."

He's talking about Google, right? 

(I'm trying to understand James Damore — not just why he was fired, but why he was hired.)

139 comments:

rhhardin said...

He didn't understand the slowly-evolving females, nevertheless.

mockturtle said...

The author needs a course in low-fog index writing.

Michael K said...

You might take a look at his PhD and the department.

Like so many brilliant young men, he followed the information to a logical and unacceptable conclusion.

Charles Murray is hated and feared for similar reasons. Following the data is not permitted if it leads to a conclusion that is not allowed.

Etienne said...

"... The humorous, but not altogether joking, solution to the game [chicken] is to throw your steering wheel out of the car, showing your opponent that you have committed yourself to driving your car straight and giving him no other option but to swerve."

"In our model, the host “throws the steering wheel out of the car” by evolving very slowly relative to its symbionts, committing itself to a certain strategy."

This is an obvious description of Alphabets social engineering policies. I don't know why he couldn't see he was making a deal with the devil.

I think he threw his steering wheel out, but Alphabet knew he would be squashed, so kept the course.

He forgot the most important biological fact: that size matters.

sparrow said...

There's liitle reason to think that expertise in one narrow field translates to others, especially the more practical social arena. He erroneously believed that reason was permitted.

The Bergall said...

Can he repeat that again, he lost me at the first sentence............

Ralph L said...

He's certainly got grad school language down pat. It's too bad that Pat bitched so much.

eric said...

You can tell that so many know Google is in the wrong here, even though they agree with Google, because they keep lying about what was written.

The news media keeps saying the stupidest things about this. But, what else are they going to say? The truth makes Google look terrible.

Levi Starks said...

Badthink is a terrible thing.
Badspeak is even worse.

Mark said...

How is this related to internet search?

Seems like he felt he was so smart he could do anything, then tried and failed.

While weeping bitter tears at having failed for the first time in his life, he files pre-emptive NLRB complaint and then dumps his MRA Reddit rant into the internal corporate network knowing that this would violate the Code of Conduct and cause his dismissal.

What a fragile ego he has.

Biff said...

Google, and many other tech firms, have been recruiting individuals rather enthusiastically from the fields of systems biology, computational biology, and bioinformatics for many years. Success in these fields requires a relatively rare combination of mathematical/computational rigor and an interest in understanding and modeling complex, real-world systems. From a math perspective, many of the practical questions in social networking, search optimization, ad targeting, and artificial intelligence are similar to questions explored in these biological fields. For example, much of systems biology focuses on how genes, proteins, and other components of a cell interact with each other as networks, and the analytic approaches developed in systems biology have proven useful in a number of other fields, including social networking. Likewise, some of the techniques used in bioinformatics to recognize patterns in DNA sequences and the activities of genes and proteins in various diseases subsequently have been used to recognize patterns in markets and behaviors. Many of the "machine learning" methods that now are making headlines in fields ranging from accounting to cybersecurity to law had their first practical applications in genetic analysis 10-15 years ago.

For similar reasons, security agencies like NSA also have been popular destinations for individuals with systems biology, computational biology, and bioinformatics credentials. It's also striking how often companies that provide analytic tools to those biological fields have supplied tools to intelligence agencies.

As some mathematicians like to say, "In the end, everything always comes down to math."

Laslo Spatula said...

Places like Google need more women employees.

There are not enough women in the programming field.

Thus, hire women in the HR fields, where their education leads them.

So: more women in the workplace means more HR to talk about how more women are needed in the workplace.

The Moebius Strip of Workplace Diversity.

I am Laslo.

M Jordan said...

This Damond story makes me wonder if the PC monster that Obama released from Hell may have already jumped the shark (yes, I know that phrase has jumped itself). Trump's win revealed that it had already happened in the culture. Now Damond is exposing the rot in collapse. Read Brietbart's excellent journalism in which other people from inside are speaking out ... albeit under aliases. I'm guessing the culture turned somewhere around 2012.

M Jordan said...

"I'm trying to understand James Damore — not just why he was fired, but why he was hired."

The passage you just quoted is your answer. This fellow was able to speak the fancy talk of the academic left which is simply the ability to spin gold out of bull crap.

Witness said...

"The Moebius Strip of Workplace Diversity."

Sounds like we need a Moebius Strip Club.

Nonapod said...

Sounds like he produced a model that gave him a result he wanted. He "showed" that slower evolving hosts influence the evolution of their symbiont much more that vice versa, I guess. I never fully trust models.

MarkW said...

How is this related to internet search?

It really isn't. A lot of companies hire smart people with degrees completely unrelated to the work they'll be doing. Engineering grads becoming 'quants' in finance, for example. But also, what he did in this paper was to build and run experiments with a computer model (e.g. software). So switching to software as a profession wasn't that big of a jump.

Seems like he felt he was so smart he could do anything, then tried and failed.

It's not clear at all that he was a bad Google engineer -- job performance is obviously not why he was fired.

Also. The ideas in the paper really aren't that obscure or hard to understand. Typically in the case of hosts and parasites, the parasites are much faster evolving than the hosts because the hosts live much longer and reproduce much more slowly (insects attacking trees, bacteria attacking humans). So how do the slowly reproducing hosts manage to keep up and survive?

Henry said...

"I'm trying to understand James Damore — not just why he was fired, but why he was hired."

Because he was really smart and applied for a job.

Some software developers studied computer science. Many software developers studied engineering or hard science. No one learns engineering or hard science without also learning how to program.

I'm not sure what group Damore was in, but most news articles call him a senior engineer. That means that when he interviewed he was probably asked to describe system architectures, given programming problems to solve, and required to demonstrate knowledge of algorithms, data structures, computational efficiency, etc.

People write books on the Google interview. And Google is open about it as well.

Ann Althouse said...

People are like bacteria... where have we seen this before?

Michael K said...

"What a fragile ego he has."

Unlike you, he has real accomplishments. I think he will do fine.

Unknown said...

yes, he should not have been hired because he is an idiot for buying into sophisticated writing. however, i assume that his colleagues are the same. so writing is irrelevant to writing code.

Clayton Hennesey said...

Google is in one business only, and that business is dependency, specifically, making as much of the human world as irreversibly dependent as possible upon Google.

Imagine if you could systematically socially engineer your food to seek you out, conform itself to your tastes, prepare itself in the manner most appealing to you, and then leap joyfully into your mouth.

A biological scientist specializing in interdependencies is a natural hire for an entity seeking to learn how to make its human symbionts even more dependent on it.

Try to do without Google. Once certainly cannot and also enjoy Althouse, whether in commenting or even reading her. Althouse is a Google property. She could theoretically migrate elsewhere, but why would she?

Ignorance is Bliss said...

Understanding how these symbioses evolve, from bacteria and their phages to humans and our gut microflora..."

He's talking about Google, right?


I don't know. My facebook feed is often indistinguishable from an intellectual fecal transplant.

Ann Althouse said...

It sounds like some of you are resisting reading technical writing just because it's hard to read. I was not intending to get that reaction. He's writing for people in the field, so you have to take that into account. Then just read slowly and reread until you understand.

You're giving me flashbacks to my days as a law professor, dealing with students who didn't want to believe they had to actually read and understand judicial opinions.

But of course, you're not training in the field of microbiology. Nevertheless, I wish you would put the effort into grasping the field of expertise this man had and trying to think about how Google may have wanted to repurpose it into dealing with people.

We know from the man's memo that he had some blind spots when it came to dealing with people. But Google must have wanted a person with that mix of aptitude and expertise. They expelled him when his particular deficiencies were exposed to the whole world, but that doesn't mean they didn't like what he was when they chose him.

I am asking you to think about the sort of mind he had, as a person who studied the interactions not of people but of microorganisms.

Dust Bunny Queen said...

The ideas in the paper really aren't that obscure or hard to understand. Typically in the case of hosts and parasites, the parasites are much faster evolving than the hosts because the hosts live much longer and reproduce much more slowly (insects attacking trees, bacteria attacking humans). So how do the slowly reproducing hosts manage to keep up and survive?

I didn't find the ideas that difficult. Perhaps he might have been a bit clearer in his writing. However, he is writing for a specific audience.

In a successful host/parasite situation, there will eventually be an equilibrium, despite the unequal pace of evolution. Equilibrium where both host and parasite combine to have a synergy or symbiotic existence.

I suppose that he IS referring to Google as well as many other societal situations. In this case, in my mind, the entity/host is the organization and the parasites are the SJW who infect the organization. One (SJWs) are overwhelming the host. IF the organization is too slow to evolve or unable to counter the infection, it is in danger of death. It sounds like he is merely stating this biological principle and applying it to what is happening in our society.

The same analogy can be made about the Muslim invasion of Western Society.

wildswan said...

There is a very good information article on Damore that just went up to day at https://everipedia.org/wiki/james-damore-1/#ixzz4pB6L3iOC

Allison said...

Google does a lot of things, not just internet search. Even inside search, we're not just looking up web pages in some gigantic card catalog that already had LC or Dewey numbers on the volumes.
I haven't seen any info on which department he worked in, so this is speculating. But here are places he could have been hired:

Google is involved in AI to write algorithms to detect malicious activities on the web. One example: spammers. Google folks write algorithms to detect spam,and spammers try to defeat their spam filters. Some folks think maybe a biological model, like an immune system response, can be coded to evolve new detection and new defenses. Some of those algorithms learn. So there's research done at Google on how to create these immune system style algorithms.

That doesn't just apply to spam. It applies to fake web sites, fake chatbots, the annoying things that ruin your blog comments, and even Captcha defeating.

Google uses AI for its autonomous cars. It has other biological analogies and interested related to parsing human speech into directions. These are just a tiny sample of reasons our hire a guy who won cs contests and did systems biology..

But mostly Google hires smart people who can pass their clever tests.

mandrewa said...

I have always known that a key part of Google's corporate strategy has been to kill off future competitors.

Thus for example Google employs a lot of people that try to come up with neat and useful computer programs. A tiny percentage of the programs that these people come up with are something that Google might make real money on and that Google realizes that this is the case. The interesting question is what Google does with the vast number of programs that it invents that either have no apparent monetary value or might have value for a small company but given the limited number of different directions any organization can go in aren't worth it for Google to pursue. So what Google does with these programs that they can't monetize is to make them free. Now that sounds kind of good, but what it also means is that they are suppressing all the new companies that might otherwise have a got a start trying to sell one of these programs.

Maybe a better example is how Google treats its employees. Google expects all of its employees to give everything they have, especially their ideas, to Google. They actually set aside a significant proportion of their employee's time where they are expected to do nothing but expound on their understandings and insights. The problem with this is that, and of course Google knows this, historically a major portion of the growth and innovation of the tech industry has come from employees of tech companies leaving and starting their own companies.

The way it worked prior to Google was that people were hired to do a job. And they were evaluated based on how well they did their specific task. If they had ideas that had nothing to do with the task, well then pragmatically speaking, those were that person's ideas and they could leave and build a new company around their ideas. Google does everything it can to prevent this from happening. They claim ownership of everything that their employees think of, and they try really hard to extract and document all of what their people are thinking.

The fact that they hired this theoretical biologist suggests that Google may be looking for ideas from nature about competing with and destroying other companies. That is not a bad source of inspiration. People don't really get it but nature is incredibly competitive. For example most plants are engaged in chemical warfare where they secrete chemicals that poison or slow down the growth of neighboring plants.

Well in any case, while Google has done some good things, in many ways I think this is actually an incredibly destructive company that is eating up our future for their narrow self-interest. I think it should be split into nine or so competing companies. And that these companies should share ownership of all of Google's current intellectual property. This should be done on anti-trust grounds, because, yes, Google really is a monopoly in a number of key areas.

exhelodrvr1 said...

Presumably Google is putting quite a bit of effort into developing AI/genetic algorithm-based applications for their company. That would seem to be a natural fit for Damore; I suspect he was involved in those aspects. And that mindset would lend itself to formulating the observations he made in his memo.

Mark said...

"Unlike you, he has real accomplishments."

You mean the Presidential Physical Fitness Awards for 2005, 2006, 2007 he lists on his resume?

The next time I go to someone in a technical position, I'll evaluate them based on physical fitness accomplishments not professional ones.

Pinandpuller said...

How many more made up BS job titles can they come up with to bring women into Google?

VP of Diversity? Why did she need a couple of weeks to get up to speed?

Henry said...

I am asking you to think about the sort of mind he had, as a person who studied the interactions not of people but of microorganisms.

Some smart people are drawn to a field of study simply because it gets them to exercise their brains.

If there is a key connecting Damore and Google (and as I implied in my Google Interview comment upstream, I'm not sure there is), I don't think it's microbiology. I think it's game theory and mathematical modeling.

I'm guessing that Damore was a student when he co-authored two papers with Associate Professor Jeff Gore. In other words, these were student papers written by a guy who then took his game theory and programming chops to get a job at Google.

Here is Jeff Gore's home page at MIT. Read his publications. They are about game dynamics, feedback loops, evolutionary scaling, etc. -- all things a mathematical programmer would be thrilled to work on.

Henry said...

For pete's sake, Damore is not a theoretical biologist. His work on microbiology was as co-author to a much-more published professor in the MIT Physics department.

MarkW said...

"But of course, you're not training in the field of microbiology. Nevertheless, I wish you would put the effort into grasping the field of expertise this man had and trying to think about how Google may have wanted to repurpose it into dealing with people. "

I doubt that Google did want to use this specific expertise, just his brains once suitably retrained for software engineering. But, if they did, it would most likely have had to do with computational game theory sorts of things, not 'treating people like bacteria'. Look at some of his references, for example:

Axelrod R, Hamilton WD. The evolution of cooperation. Science. 1981;211:1390–1396. [PubMed]

Bob Axelrod won a MacArthur 'Genius' award for that work:

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



Art in LA said...

Alphabet (Google's parent) has a life sciences company called Verily. Maybe he worked there?

http://www.mercurynews.com/2017/07/18/heres-the-real-buzz-google-releases-20-million-mosquitos-in-fresno/

And, BTW, don't we all have blind spots when dealing with people? @Althouse, maybe you can provide an example or two of blind spot-free people.

Unknown said...

he is talking about google being its own bacterial culture; the conclusions from this are quite clear in context of saying that PC and biological differences are harming the company and the valley

possibly a misogynist but not sexist

obvious

Allison said...


Here is a typical interview question at Google. A person who enjoys and quickly solves a question like this,and enjoys the process of receiving rapid fire questions like this, has a certain kind of mind. You can search for "Google interview questions" and find many more. They are very similar to the questions on midterms and finals in comp sci and discrete math ugrad and grad courses at places like MIT, Stanford, UC Berkeley, CMU.
Here's an example:
A building has 100 floors. One of the floors is the highest floor an egg can be dropped from without breaking.
If an egg is dropped from above that floor, it will break. If it is dropped from that floor or below, it will be completely undamaged and you can drop the egg again.
Given two eggs, find the highest floor an egg can be dropped from without breaking, with as few drops as possible.

Henry said...

According to wildswan's link, Damore's PhD is in Systems Biology:

Systems biology is the computational and mathematical modeling of complex biological systems. An emerging engineering approach applied to biological scientific research, systems biology is a biology-based inter-disciplinary field of study that focuses on complex interactions within biological systems...

Source: https://everipedia.org/wiki/Systems_biology/#ixzz4pBJJN0WB

Skills like this are obviously of great value to a company like Google.

HoodlumDoodlum said...

Why the "bad science" tag? Where is the bad science?

We know from the man's memo that he had some blind spots when it came to dealing with people

Please be specific--what are the blind spots?

Biff said...

Another thing that comes up in various computational biology problems: how and when do you optimize things for an individual case versus how and when do you optimize things for a population of cases? Related: what happens when an individual is substantially different from the overall population? What happens when there are sub-populations which have measurable differences from each other?

Usually, these questions are raised in terms of medical treatments (i.e. the best thing for an individual may bankrupt the system if applied over a population, while the best thing for a population may have little or no direct relevance to a particular individual), but they can be asked in other contexts, such as social dynamics, employment patterns, diversity, and so on.

I think it's possible to argue that the exact kind of thinking for which Google hired Mr. Damore to perform is the exact kind of thinking that led to his dismissal.

Michael K said...

The next time I go to someone in a technical position, I'll evaluate them based on physical fitness accomplishments not professional ones.

I would not be surprised. You seem just that dense.

Birkel said...

UnknownIga51 is back.
Maybe this worker can just adopt a new old handle and keep plugging away as if nobody wud notice.

Michael K said...

We know from the man's memo that he had some blind spots when it came to dealing with people

Please be specific--what are the blind spots?


There is such a thing as not understanding people who are not as smart but think they are.

Fernandinande said...

Google is an advertising company, and

"'Red Queen marketing' is defined as the business practice of launching new products in order to replace past failed launches while the overall sales of a brand may remain static or growth is less than fully incremental (Donald Kay Riker, 2009)."

mandrewa said...

Yes, Henry, you are right. He is not a biologist nor is his coauthor. But the paper they wrote is a part of theoretical biology. You do not have to be trained as a biologist to do work in the field.

Now I did wonder as I was reading that paper whether some parts of their understanding of biology were not quite right. Actually, to put it another way, the paper is really about a model of biology. As with any model a huge number of things have been left out.

So maybe Google hired him because he was a computational model builder, which is skill that can be applied in many directions, and not because of his biological insight. Or maybe he was hired because of the combination of both skills. Who knows.

Henry said...

There is such a thing as not understanding people who are not as smart but think they are.

There is also such a thing as not understanding people who can fire you.

Roughcoat said...

Evidently "n.n." is John Damore.

themightypuck said...

Maybe they hired him to better understand sex differences in order to improve ad targeting.

lgv said...

Did everyone read his memo, or "screed" if you prefer?

I did. I was underwhelmed with it. I found contradictions and flaws in his thinking. While I think Google and its ilk, with their social justice, diversity, and PC non-sense to be a bastion of totalitarian group think, I think I would have fired him, too. I'm not sure he is the right poster child for the anti-intolerance corporate religion.

HoodlumDoodlum said...

Ann Althouse said...He's talking about Google, right?

No, he's not. If you want to apply what he's talking about, via metaphor, to Google, you can do so. You should do us the favor of identifying the host and symbiont in your metaphorical application, though, as well as stating whether you thing the symbiotic relationship is parasitic or mutualisitc.

The paper describes their model:

Here, we model a host-symbiont relationship as a game that each host plays in pairwise interactions with each of its symbionts. We combine the common asymmetry in evolutionary rate and population structure to allow for rapid adaptation of each symbiont population to its individual host. Thus, when the symbionts evolve much faster, only those adapted to their individual host will survive; the population of symbionts in a host will evolve to play the best response to their host's strategy. In turn, only the hosts that fare best when the symbionts respond optimally to the host's strategy will flourish.


How would you apply to Google? To Google employees? To Google as host (evolving slowly) and Google users (evolving more quickly)? To Google as a host and SJW employees vs non-SJW employees as symbiotes?

I don't see any evidence that he was "talking about Google" in this paper. If you'd like to apply it to Google please be specific about how.

chuck said...

> But mostly Google hires smart people who can pass their clever tests.

ISTR that they found that the clever tests were useless and now use other criteria.

HoodlumDoodlum said...

Igv said... I think I would have fired him, too. I'm not sure he is the right poster child for the anti-intolerance corporate religion.

Why would you have fired him? Who is the right poster child for anti-intolerance corporate religion? Would you have fired him if his memo was less flawed and had fewer contradictions?

Ralph L said...

One thing that confused me is that they mention mutualistic relationships but then talk only about parasitic ones.

I wonder if they excluded the equilibrium where everyone's dead.

My brother met his wife when both were copy editors for the Amer. Soc. for Microbiology's publications. They had to translate and correct (mostly) foreigners' English into this stuff.

Dust Bunny Queen said...

@ Hoodlum

see my 11:03 comment

His paper was not about Google specifically, however it does appear that he was applying the biological phenomenon of host and parasite evolving into the equilibrium of a host with symbiotic relationship with the parasites. Game theory as a psychological thing is interesting enough, later it became applied to biological systems.

Game theory being shown in real life in a real game....World of Warcraft. The Corrupted Blood event (I was there it was awesome!!!) is a study that serious entities such as the CDC and others have studied as it involved real people (role playing in a game) reacting in real life to a simulated medical catastrophe.

Fernandinande said...

Biff said...
...As some mathematicians like to say, "In the end, everything always comes down to math."


+10. And he developed simulations - computer programs - to model the rather complicated game-theoretic ideas.

Michael K said...

"There is also such a thing as not understanding people who can fire you."

They are pretty much the same as those who aren't as smart as they think they are.

I think he is going to do fine, probably with his own company.

They forced out the Uber founder and Eich but, last I heard, they had not taken away their money.

Bay Area Guy said...

I would like to have a parasitic, symbiotic, mutualistic relationship with Kate Upton, but she is married to some washed-up pitcher on the Tigers.

wildswan said...

I think Biff explains why Damore was hired by Google. And I have some thoughts on Damore's opinions based on the everedia site information. I know Damore did not cite Murray but he knows the theory and is using a variation on it, I am sorry to say. I am sorry because he is being piled on and I hate that.

1. Damore is a chess and math whiz. From his circle and his experiences therefore he knows exactly how many women, Asians, blacks are really good at what he is really good at. His picture suggests that he also knows exactly how many women value and are good at what he not good at.

2. His PhD was in the field of morphology, “Morphogenesis and morphometrics.” This an intensely interesting field which is stuck - it has no successful leading ideas from the field of mathematics. If you go into it, you fail. You don't fail to get your PhD but you fail in the sense that you aren't able to explain the phenomena that interest you mathematically. It has been like that since Theophrastus. And when you fail like that - not in your career but in a a personal sense, you are more open to ideas which are the opposite of your original inspiration but which are more successful. The field opposite to morphology since the time of Darwin or rather since Darcy Wentworth Thompson wrote On Growth and Form, has been standard evolutionary theory which explains form by a sort of random but plastic flow analyzed by statistics. Observed patterns of distribution analyzed by statistics are regarded as basic evolutionary information and this formed the basis of eugenics as it existed until the Human Genome Project upended all previous work. The link between social and cultural distributions and genetic distributions is simply unknown at this time. But I think Damore would disagree with that - I think he thinks it is scientifically known and that Google is harming its own evolution by trying to fight against science.

3. You could think as I do that most women do not like math and are not good at it without getting into evolution and Charles Murray. But in our day scientific credentials seem to make people foolish because such credentials make them respected. Damore was good at math and weak on genetics. He saw Murray had his sums right and that Murray's conclusions agreed with his own experience on differences in mathematical interest among and between groups. he didn't ask sophisticated questions about underlying forms of data collection in relation to evolution - he wasn't interested in that kind of math. And not being interested in the scientific basis of what he actually was talking about he should not have sounded off on how Google was biologically unsound in its HR policies - and this should stop. It would have been enough for him to say that try as you might you will never get as large group of women to take a real interest in math as you will easily get of men. And it clogs the effort to make money to pretend otherwise.







HoodlumDoodlum said...

Dust Bunny Queen said...His paper was not about Google specifically, however it does appear that he was applying the biological phenomenon of host and parasite evolving into the equilibrium of a host with symbiotic relationship with the parasites. Game theory as a psychological thing is interesting enough, later it became applied to biological systems.

Yes, I'm aware, and I saw how you guessed the paper's model could be applied to the situation at Google. Is that the same application the Professor had in mind? I kinda doubt it, but until she specifies how she thinks they relate I don't know.

Everybody's getting so vague all of a sudden. I'm sure the guy has all sorts of "blind spots." Which ones are you talking about? This post, linking to this paper, has a "bad science" tag. What's about about the science discussed here? Could this paper's topic be applied to Google in some way? Sure, in lots; some more compelling than others. Which way did you have in mind?

If there's a point to me made, why not make it? Waving at something and saying "oh, this is like that other thing" isn't saying much.

HoodlumDoodlum said...

Bay Area Guy said...I would like to have a parasitic, symbiotic, mutualistic relationship with Kate Upton, but she is married to some washed-up pitcher on the Tigers.

In what way would the relationship be mutualistic? Parasitic I'd buy, but she's doing pretty well without you now, yeah?

Michael K said...

The Google HR Diversity Nazi lady was a big Hillary activist last year. That kind of thinking is welcome at Google.

She has now made her Twitter account "private." Wonder what's in it.?

Fernandinande said...

Mark said...
Seems like he felt he was so smart he could do anything, then tried and failed.


No, it doesn't seem like that at all.

Dust Bunny Queen said...
The ideas in the paper really aren't that obscure or hard to understand.


Apparently people who don't know much about biology, which is most people, and have never pondered over relationships between hosts and parasites, find it hard to understand. Buncha furrin concepts and big words.

Plus: lichen!

I suppose that he IS referring to Google as well as many other societal situations.

Pretty sure the paper was written (pub 2011) well before he had anything to do with google.

One (SJWs) are overwhelming the host.

You're not the only person guilty of that thoughtcrime.

Breezy said...

"I'm trying to understand James Damore — not just why he was fired, but why he was hired."

OMG! He's a white male! WTH were they thinking?!

Yancey Ward said...

"Sounds like we need a Moebius Strip Club."

Is this the one where you end up stuffing the bill into your own G-string?

J2 said...

I don't understand it; yet I love it.

51st grade level.

Bay Area Guy said...

In what way would the relationship be mutualistic? Parasitic I'd buy, but she's doing pretty well without you now, yeah?

I posit that she would do even better with me.

themightypuck said...

I think there is an issue in tech where some of the most productive people are not neurotypical. They don't function that well in a large corporate structure but in a highly competitive environment corporations want these people. At a certain size, it makes more sense to get rid of the non conforming but it isn't without costs.

Sebastian said...
This comment has been removed by the author.
Sebastian said...

Damone is a kid. Naive in the extreme. Not a "biologist," properly speaking, though genres have blurred.

And what did he think he was doing with game theory? Everyone knows that's an exercise in cosmopolitan white male supremacy, if you know what I mean. Von Neumann, Roth, Axelrod--nuff said.

"mostly Google hires smart people who can pass their clever tests." Correct, mostly--but who are the people they hire who don't pass their tests? and doesn't everyone know that minorities do poorly on tests due to performance anxiety, so surely they get a pass, no? let's see all the googlers test results, particularly for the VP for diversity.

Anyway, the Google question is whether humanity is host or parasite.

buwaya said...

In a previous job (working with manufacturing quality-control systems, over 30 years ago) we hired several "non-traditional" sorts for particular skills. One gentleman was a UC Berkeley astrophysicist who specialized in orbital dynamics. He was working on curve-fitting algorithms to inspect parts with complex geometries based on the minimum number of measurements.

J2 said...

He probably understands "Westworld".

Michael K said...

" At a certain size, it makes more sense to get rid of the non conforming but it isn't without costs."

Yeah, that was the reasoning of Xerox when they close Xerox PARC.

"Get back to work on copiers and forget that computer stuff !"

DanTheMan said...

>>How many more made up BS job titles can they come up with to bring women into Google?

How many men took the day off because they were afraid of something written in a memo?

Fernandinande said...

I am asking you to think about the sort of mind he had, as a person who studied the interactions not of people but of microorganisms.

I'm wondering about the mind of a person who would wonder such a thing. Darwin's major interest was earthworms. Dawkins, IIRC, majored in ants. Why didn't they study psychology or - shudder - sociology, or something trivial and easy like law? Why doesn't everybody?

People are involved in symbiotic relationships, mutually beneficial as well as parasitic, both with other people and with other organisms.

It's easier to uncover the principles of "life" using simple organisms than using more complicated organisms like people, who have more confounding variables as well as the moral problems with experimenting on them. Evolution in microorganisms is observable because it's so fast, but many of the principles apply to all organisms.

Brando said...

Not sure I see why you ask how he was hired--presumably he met their needs for an engineer. How he writes internal memos may have nothing to do with that.

Anyway, it seems the big "controversy" here is that he is suggesting biological traits may be behind why women are underrepresented in tech. Whether that's true or not I don't know but it's hardly "offensive" to raise that question (particularly in the context of a long memo going over many other alternative ways to increase diversity). Remember over a decade ago when Larry Summers got into big trouble at Harvard for making similar musings? My question now is has today's society gotten worse (from the standpoint of more intolerant of such nonconformist views)--is the public reaction far more hostile to this engineer than it was to Summers?

That's tell us how much society is in decline.

Brando said...

"OMG! He's a white male! WTH were they thinking?!"

It may take years, but at some point all white males will just refer to themselves as Hispanic to avoid the public shaming. Though as some "white Hispanics" have learned, that won't even protect them.

mockturtle said...

It sounds like some of you are resisting reading technical writing just because it's hard to read. I was not intending to get that reaction. He's writing for people in the field, so you have to take that into account. Then just read slowly and reread until you understand.

No, Ann! I have written quite a few technical articles--some published in well-known journals. As I mentioned before, the company I worked for insisted that we ALL take a course in effective writing using lower fog index terminology wherever possible. Too many syllables and too many unnecessary words only muddy the water [which, in all too many cases, is the intent!]. Shakespeare was a very low fog-index writer, BTW, and a very good one.

DanTheMan said...

The great lefty syllogism:

Free speech I don't like isn't just words, it's violence.
I need to be protected from violence.
Therefore, I need to be protected from free speech I don't like.





Angel-Dyne said...

Mark: How is this related to internet search?

Lol.

Seems like he felt he was so smart he could do anything, then tried and failed.

While weeping bitter tears at having failed for the first time in his life, he files pre-emptive NLRB complaint and then dumps his MRA Reddit rant into the internal corporate network knowing that this would violate the Code of Conduct and cause his dismissal.

What a fragile ego he has.


Mark, do you have anything interesting to say on the larger issues at play here, or do you just want to show off what a good little cocksucker you are willing to be for Stasi-flavored corporate cultures?

Yancey Ward said...

I doubt he actually had a blind spot if he did file an NLRB complaint as a preemptive strike. Indeed, I think his plan all along was to get Google to prove his point by firing him. I read his entire memo, and Google is in trouble here. I don't remember who wrote it above, but it was correct- the lying about the contents of the memo is almost epic in its scope and depth.

Titus said...

Hello? He has Princeton, MIT and Harvard on his resume. Of course he was hired by google.

Dust Bunny Queen said...

Anyway, the Google question is whether humanity is host or parasite.

Even fleas have parasites. Parasites all the way down '-)

Art in LA said...

Corporations are dictatorships, so something like this guy's firing is totally expected. Company culture is "communism with a choice." If you don't like (or fit in with) the corporate culture, move on, or get fired like this guy.

HoodlumDoodlum said...

Bay Area Guy said...I posit that she would do even better with me.

Only one way to find out! Good luck.

buwaya said...

"If you don't like (or fit in with) the corporate culture, move on, or get fired like this guy."

This removes the option of reforming the culture of the enterprise, which has happened.
CEO's are often very big about this sort of thing.
A long shot though.

That was the whole point of a vast range of management books, from Peter Drucker onward.

Ralph L said...

I'm surprised they used "gut" instead of gastro-intestinal track.

The whole point of capitalism is (voluntary) mutual symbiotic relationships, which doesn't seem to be what the paper is about.

mockturtle said...

An all-powerful State is much more to be feared than a powerful company or even a powerful corporate oligarchy.

Ralph L said...

This removes the option of reforming the culture of the enterprise, which has happened.
We know what happened to Communism and Woolworths and Sears and most of Detroit. We'll soon see what happens to the blue hells of CA, CT, IL.

Pinandpuller said...

Don't excrete where you phagiate.

Yancey Ward said...

Allison, if I can redrop the eggs, I really only need one egg and can start at ground level and keep going up one floor at a time until I find the floor at which it finally breaks- the number of drops needed equals X, the floor at which it finally breaks. So one uses the second egg first to find a higher point to start with the other one. This reminds me of the pricing game on The Price is right where a person has 30 seconds to guess the price of the item to the dollar.

I would take the first egg and drop it from the mid-floor of the building (50th floor)- if it breaks, then I take the second egg and start at floor #1 going up one at a time until it breaks- X+1 drops total where X is the lowest floor at which the egg breaks. If the first egg doesn't break, then I take it to the 3/4 height of the building (75th) and repeat. I keep going up half the remaining distance until the first egg breaks, at which point I take the second egg and drop it from the floor just above the highest floor at which the first egg didn't break and keep going up one floor at a time. The narrowing algorithm, of course, may eventually turn out to determine X because 1/2 the remaining difference eventually becomes a single floor, in which case I don't even need the second egg at all.

Art in LA said...

@buwaya, yeah, I agree, a long shot. It is the rare company that reforms from the bottom up. Apple is a good example of top-down re-engineering of the culture, when Steve Jobs returned to the fold. Google needs to refine their "do no evil" mission statement ... "evil" seems pretty broad. I could argue that advertising is evil, and demographic segmentation and profiling is wrong, right?

D. B. Light said...

All he's saying is that complementary differentiation of roles by sex proved to be advantageous in evolutionary terms and as a result became encoded into our genetic structures.

Oso Negro said...

A TSA agent insisted on feeling my penis yesterday. When I objected, a second agent was called as a backup. I guess most Americans with penises calmly let the government feel them. While Titus might have enjoyed it, I was near insurrection. If it happened in the workplace, it would be considered sexual assault. And I was on my way to work. So women can complain and stay home because a man doubts their capability, but I am supposed to keep calm and carry on after a government mandated groping. Does anyone have statistics on how many weapons the TSA finds concealed behind penises? I would like to report I was packing a massive steel rod. But my penis is average in every way.

mockturtle said...

Oso, how did they know your penis isn't an explosive device! [I'd be willing to bet it is, on occasion].

Yancey Ward said...

Mockturtle wrote:

"An all-powerful State is much more to be feared than a powerful company or even a powerful corporate oligarchy."

You obviously don't watch many modern motion pictures or SyFy series.

Michael K said...

That was the whole point of a vast range of management books, from Peter Drucker onward.

Peter Drucker was fortunate to die before this crap became the normal.

I read everything he wrote, including his autobiography.

I once tried to hire him for a one day seminar on management for the medical association when I was president of it. That was in the 80s when managed care was the big issue.

His fee was $10,000 for one hour or one day.

He wrote a very good book on non-profits. His model was the Girl Scouts and how they handled the transition from stay at home moms to working moms.

mockturtle said...

Yancey Ward surmises: You obviously don't watch many modern motion pictures or SyFy series.

You're right. I don't.

DougWeber said...

Interesting if applied to Google and its users. In that case the quickly evolving partner is Google, since it has code change cycles on the order of months and users on the order of years. Thus once would expect that in circumstances where either Google or users could be made to be responsible for some portion of the interaction, that Google would be the partner who would change to accept the expense.

Caligula said...

"The great lefty syllogism: is really simpler than that, not exactly a syllogism at all. It is:

Your speech is violence, and
my violence is speech;
therefore, "free speech" means I have the right to smash your mouth.

______________________
As for Damore, I suspect he drank the Google kool-aid: that is, when they said "Google is a totally data-driven company" he took them at their word and assumed they'd respond to a data-based,logical with data and logic.

Instead, they responded with neither with data nor logic but merely with authority.

Mark said...

"Mark, do you have anything interesting to say on the larger issues at play here, or do you just want to show off what a good little cocksucker you are willing to be for Stasi-flavored corporate cultures?

8/8/17, 12:53 PM"

Given the level of your critique and the fact it is nothing but a personal attack, I will respond in kind.

Go fuck yourself.


Sebastian said...

"he took them at their word and assumed they'd respond to a data-based,logical with data and logic" It's not entirely clear he did so naively, and I'll entertain the hypothesis that he baited them, saying nice liberal things, footnoting "controversial assertions," and so on, in order to make them prove his echo-chamber point. More likely, he meant it--that prog powers would actually consider demoralizing diversity and be driven by data and retreat from counterproductive PC silliness.

But just as in organizations no one ever cares about process, no one ever cares about data, except insofar as they serve their cause. They are always invoked dishonestly or disingenuously. Althouse might need a data bullshit tag. The science is settled: it's a mistake to buy the BS. Not that this datum will make any damn difference.

Michael K said...

Mark explains his personal philosophy.

mockturtle said...

Mark suggests: Given the level of your critique and the fact it is nothing but a personal attack, I will respond in kind.

Go fuck yourself.


He doesn't have to, Mark! That's what sexbots are for!

chuck said...

> Darwin's major interest was earthworms.

Nope, he had many interests, but his early publications were on corals and barnacles. He spent eight years working on his barnacle books. The earthworms came late in life, the publication a year before he died.

Michael K said...

Yes, Darwin's interest was barnacles and he went on the Beagle to study them. They stopped at the Galapagos Islands and he got interested in evolution there.

Allison said...

Yancey, I like what you noticed, and I like your solution.

So let's play a little, because that's what happens next at the interview-they want to see what you noticed. You did not state it, but you realized you can solve the problem always in n+1 drops using just one egg, where n is the highest non-breaking floor.

So the point of the second egg is to make it so you aren't just testing in linear order. You can halve the search space.

Can you prove this is the best possible answer? That is, what the worst case number of steps?

And now we could push more: what's the smallest number of drop if you had more eggs? You'd halve at each step, for k-1 eggs and in the kth egg search linearly bottom up. What's k? What's the least k if the building is 2^7 stories? 2^n for any n?


Ralph L said...

Allison, I would drop from every third floor until one breaks. Then you use the second egg to see which of the middle two is the highest non-breaking floor.

Mr. Fabulous said...

(World Famous Lurker Says....)

Allison, I think you and Yancy have part of it, but if we use a non-linear solution, we could do something such as this: The first drop is at the middle - 50. Whether the egg breaks or not, you've eliminated half of the floors. The second drop is in the middle of the remaining 50, giving you a remainder of 25. The 3rd drop in the middle of the 25 leaves you with 12 or 13. The fourth drop 7, the fifth drop 4, the sixth drop 2, and the 7th drop is determinative.

So, with this method, 7 drops max to determine the correct floor. At least that's what I've come up with after a little thought, without consulting other resources. Please feel free to tear this apart.

mockturtle said...

Allison, I think you and Yancy have part of it, but if we use a non-linear solution

The solution is, in fact, nonlinear.

mockturtle said...

No two eggs are the same, either. The curvature of the egg will affect the angle of impact.

Mr. Fabulous said...

(World Famous Lurker says....)

On the ride home, I realized that my response above gives the answer for using just one egg per drop. Since the question stated the use of 2 eggs, we can determine the highest floor an egg will not break in no more than 4 drops of 2 eggs each.

Michael K said...

Since the question stated the use of 2 eggs, we can determine the highest floor an egg will not break in no more than 4 drops of 2 eggs each.

Leibniz would say you are trying to invent calculus;

The Toothless Revolutionary said...

(I'm trying to understand James Damore — not just why he was fired, but why he was hired.)

No you're not.

I'm sure the biology he wrote in the excerpted passage above is not as controversial in biology as it might be among whomever's trying to make a political case about this whole thing one way or another.

Owen said...

Mr. Fabulous: in your scheme, suppose the egg breaks on the first drop (50th floor). You propose to drop the second egg from the 25th floor. Suppose it also breaks. Now you have no more eggs and you only know that the answer is somewhere between floors 1 and 24.

mockturtle said...

A theoretical model is only as sound as the variables factored in. This is the problem with the climate change projections.

Mr. Fabulous said...

(World Famous Lurker says....)

Owen, respectfully, if you only have one drop of the 2 eggs, you cannot determine the answer. I think Allison didn't state the question clearly: "Given two eggs, find the highest floor an egg can be dropped from without breaking, with as few drops as possible." My interpretation of the problem is that the part about "... as few drops as possible" implies successive drops of pairs of eggs until you have your answer. Perhaps Allison can clarify if I've interpreted the problem parameters correctly.

Ralph L said...

Original puzzle:
A building has 100 floors. One of the floors is the highest floor an egg can be dropped from without breaking.
If an egg is dropped from above that floor, it will break. If it is dropped from that floor or below, it will be completely undamaged and you can drop the egg again.
Given two eggs, find the highest floor an egg can be dropped from without breaking, with as few drops as possible.

I still like my answer at 5:21

Ralph L said...
This comment has been removed by the author.
Henry said...

Zero drops is the least number of drops. And the answer is the first floor, because eggs.

Henry said...

Let's have breakfast.

sena nabila said...

nice blog
Walatra G Sea Jelly

chuck said...

@Ralph

Given that you have eactly two eggs and get no more, then you are in the right direction, but 3 isn't optimum. Once you have an interval bounded on the bottom by good and on the top by bad, you can proceed from the bottom of the interval up. You don't want equal sized intervals either. The intervals can also be rearranged without changing the answer, so there isn't a unique strategy. I get a minimum of 14 tries, ending up with both eggs broken but the floor known.

Ralph L said...

I conked out on the sofa trying to come up with a way that works for both high and low floors. I'm going to blame that on too much dinner and not too many years.

Jonathan Graehl said...

why was he hired? dumb question. when your company wants to perform well you hire high performers. he's a high performer (although not socially - there's video of him w/ stefan molyneux)

Qwinn said...

Ralph: Your method could take as many as 35 drops...33 to get up to floor 99, then 2 more on testing 97 and 98.

I say, do every 10 floors. When one breaks, then you test one at a time from the last one that didn't break. Max number of drops: 19.

Yancey Ward said...

On the egg problem, you have to determine the actual highest floor from which an egg can be dropped without breaking- failing isn't an option. As I pointed out, with one egg, it can be determined in exactly X drops where X is the lowest floor from which it is dropped and breaks by simply starting at ground level (floor #1) and going up one floor at a time. That one egg must be used in this manner to ensure you do, in fact, solve the problem. What the second egg gives you is the ability to, as someone stated above, the power to narrow the search space from 100 floors to 50 in a worst case scenario, and narrow it further in better cases. I don't see a better solution than taking the first egg and dropping it from the 50th floor- it either breaks or it doesn't. If it breaks, you are still left with starting at ground level with the second egg and going up one a time to find X, and the total number of drops will be X+1 where X+1<50. If the first egg survives the drop from the 50th floor, then you again halve the remaining search space by dropping from the 75th floor. If the egg breaks, you take the second egg to the 51st floor and proceed upwards one floor at a time until you find the correct floor in X+2<25. You can, of course, continue this iterative process to the 87th floor, the 93rd floor, the 96th floor, the 98th and the 99th.

So, in the very best case scenario, you can determine it in 7 drops with only the first egg. As far as I can see, though, with two eggs, it should never take more than 50 drops in the worst case scenario where the first egg breaks at the 50th floor, and the highest floor from which you can drop an egg without breaking it is the 49th floor. In the next better case, the first egg breaks at the 75th floor, so it should never take more than 26, and so forth.

Now, Allison asked if I could prove that was the best use of two eggs. Not sure what would constitute a proof of this- it almost seems like an axiom that one egg can only narrow half a given search space in a single iteration. Similarly, if I had three eggs, I could drop the first two simultaneously from the 33rd and 66th floors, thus ensuring my search never requires more than 35 drops total. However, if the egg from the 33rd floor breaks, I am wasting the second egg on 66, which then makes me wonder what the optimal floor for the first egg is again- is it still 33, or should I drop it from some other floor? That would require me writing things down systematically, and it is bed time for me.

Qwinn said...

Actually, max 18. If you drop from the 90th and don't break, you can test 95 next rather than stepping up one at a time. Worst case is breaking on 90th floor (9 drops), then dropping at 81, 82, up to 89 (another 9 drops).

Qwinn said...

Yep, I tried shrinking and growing that interval of 10, and it never seems to get better than max 18 (though there are other intervals that also yield 18). I'm gonna make 18 my final answer.

Yancey Ward said...

Qwinn,

Your example has convinced me I am thinking about this wrong. My method has a worst case scenario of 50 drops, with a best case of 2. Of course, my worst case is a low probability event- the floor is the 49th floor specifically as is the best case when it is the first floor. What is the sum of all possible outcomes in your example and mine? For example- if the highest floor an egg survives is 68, then your method takes 7+8=15, where as mine still takes 2+18=20. However, if the highest floor is 93, then your method takes 10+3=13, whereas mine takes 4. I now understand better why Allison asked me if I could prove it. I can't yet. My guess without doing explicit summation of all the possible outcomes is you method defeats mine.

Yancey Ward said...

Just to get a feel for what I was trying to understand, if the highest floor is the number outside the parentheses, then the number of drops required is the number within for the two methods. I think Qwinn beats me here just based on the following, though I haven't summed all the possible scenarios- mine performs far worse in the 1-50 range, and better beyond 75.

Qwinn method 1(2) 2(3) 3(4) 4(5) 5(6) 6(7) 7(8) 8(9) 9(10) 10(3) 11(4) 12(5) 13(6) 14(7) 15(8) 16(9) 17(10) 18(11) 19(11) 20(4) 21(5) 22(6) 23(7).....

Yancey method 1(2) 2(3) 3(4) 4(5) 5(6) 6(7) 7(8) 8(9) 9(10) 10(11) 11(12)......49(50) but then 50(3), 51(4).....74(26), but then 75(4) ......86(12), but then 87(5)....92(6)

chuck said...

> Yep, I tried shrinking and growing that interval of 10,

Do it this way. Just to be clear, the drop on floor on is from the floor to the floor, so floor one is known good. Now add 14 to that and drop from floor 15. if it fails, work up with the other egg from 2 to 14 to find the last good floor. That is 14 tries total. OK, suppose floor 15 is good, we have used one turn, so jump up 13 floors to floor 28. If floor 28 fails, work up from 16 to 27, that is at most twelve more steps + the two already taken, so again 14. So on and so forth. In the end, we are looking for the smallest positive integer n such that 1 + n*(n+1)/2 >= 100. Plugging in 14, 1 + 14*15/2 = 106.

Yancey Ward said...

chuck,

Thanks, I think I understand that, and can get a good night's sleep.

Qwinn said...

If you're arguing that 14 is a better interval than 10, I disagree. I showed how my worst case (egg breaks at 89) would take 18 tries. Your method, say the break floor is 84. You did 15, 29, 43, 57, 71, 85 (break). That's 6 drops. Now you need to drop from 72 to 84, that's another 13, total 19 drops.

chuck said...

@qwinn

The maximum number of drops needed is 14. Did you notice that the intervals decrease in size as you go up? Rephrase the question to "how far up can you get with two eggs with at most n drops" and it will make sense. For fun with recursion, do it with three eggs.

Yancey Ward said...

chuck is correct- you have to pick a first interval that is large enough so that the sum of all the potential intervals, n + (n-1) + (n-2)....reaches at least 100, the height of the building. 14 is the smallest number that satisfies that condition. As long as you do that, you can ensure that the number of drops is 14 or less regardless of where the actual highest floor is.

Qwinn said...

Nope, I didn't get the decreasing interval. Clever. Seems to work. I concede.

mockturtle said...

Nice work, guys! But don't try this with real eggs.

Art in LA said...

Using the "interval" egg dropping method, can't you jump two floors from the last known "good" floor to determine the answer? Saves some drops, right?

Art in LA said...

Oops, my bad ... you need to go "brute force", one floor at a time, to figure it out. Sorry!