Hacker News new | comments | show | ask | jobs | submit login
The duties of John von Neumann's assistant (1993) (plover.com)
249 points by AndrewDucker 5 months ago | hide | past | web | favorite | 66 comments

The original this is taken from is better [0].

0. https://www.maa.org/sites/default/files/pdf/upload_library/2...

Wow, the ending is brutal:

> At the end of May I left Szeged. I was grateful for all he had done, but it was only much later that I appreciated reasonably well his contribution to my education. On the overnight train from Budapest to Venice, a young Hungarian man accompanied by a young woman kindly asked me in poor German to leave them the compartment for the night. They had just been married that day and were beginning their honeymoon. On arrival at Venice the next morning, he opened the corridor window and got off the train, while his bride, who spoke only Hungarian, lowered their suitcase through the window. Then a terrible thing happened. The train pulled out, with her still on it. The poor girl was cut off from everything that she knew, everything she had been living with. I knew what she felt. Hadn't I too just been cut off from my previous life?

It __is__ better. It helps him explain his case as to why he chose to accept the Cutting Travelling Fellowship rather than work for Von Neumman, and he also mentions that the job was later split into 4 parts.

The original link on the other hand just dismisses this as "he fled to Hungary". True, but an incomplete truth.

Thanks for sharing this. "Once, after I suggested a problem, he[Frigyes Riesz] he repeated it slowly twice to make certain he understood. Then, absolute silence. After half an hour, he exclaimed "Ja uj. I have it. Yes, I have it.", and he leaned to me, his eyes full of joy and mischief. But that was the end of it. He never made me privy to his solution." - We can learn a lot by working with individuals like these, who sometimes reveal nothing about their method in solving a problem, evoking (mild) stress/frustration upon others until they too solve the problem or at least spend some good time thinking about it.

Mmm. In particular it notes that (a) in previous years the assistant had been only expected to do the first two parts of the job description and (b) after Lorch turned it down they ended up dividing the work between four assistants. So it's more of an example of an over-ambitious job description than the reality of being an assistant to von Neumann...

Nice post. About the article: it must have taken some balls (for lack of a better expression) to turn down the position as von Neumann's assistant.

This is beautiful. Really interesting reading. Thanks.

What they don't tell you is that at the Institute for Advanced Study the poor assistant had to use the same filing cabinet for his list of duties, to do list, and all of Von Neumann's correspondence and papers.

While up at Harvard the assistants were given separate cabinets to make their lives easier. But woe betide the poor assistant who tried to put a file in the wrong cabinet!

Please feel free to include in your comedy CV that you managed to sneak this one past a computer science professor until the third read-through. Kudos. :)

This is an excellent and under-appreciated joke. Bravo.

What's the joke?

von Neumann architecture vs. Harvard architecture (of computers)

It was traditional belief that Einstein's assistant did nothing. The only requirement for him was to be a Jewish exile from Nazi Germany.

Einstein's assistant Kemeny ended up designing the BASIC language and being one of the founding fathers of computer science. His assistant Rosen ended up being president of Ben Gurion Univesity and co-founding the Israeli Academy of Sciences. Based on their accomplishments, seems to me like their religion was not the only requirement made of them.

If you read it closely, it seems like fluency in advanced mathematics and significantly above-average IQ were implicit criteria. At least that was my takeaway.

I think "Jewish" in this case refers to whatever racial criteria Germany was using to choose people to exterminate, not religion.

As a young post doc at the Institute for Advance Study, my Dad knew von Neuman, who according to my Dad, tended to work late at night and work with loud music playing - very much like young hackers today!

I remember reading a similar anecdote in the book Turing's Cathedral, where it suggested that von Neumann had a particular preference for loud marches. It also makes me think about what Stanislaw Ulam wrote of von Neumann[1]:

"If one has to divide mathematicians, as Poincaré proposed, into two types—those with visual and those with auditory intuition—Johnny perhaps belonged to the latter."

[1] http://www.ams.org/journals/bull/1958-64-03/S0002-9904-1958-...


Not completely similar but somewhat interesting, here's an interview where Feynman admits to not being "great" (for lack of a better word) at music.


“Lot’s of people love music and I never could carry a tune.”

I've read a funny anecdote (no idea whether it is accurate) that von Neumann's loud gramophone music was a big source of frustration for his office neighbor, Albert Einstein.

In some ultra-low power distance societies, places like Sweden, the idea that anyone, no matter how important, is ever entitled to an assistant or secretary is almost anathema.

Imagine if von Neumann had been required to do all of that in addition to making his own airline and hotel reservations, as is the case for many university professors these days.

Could you confirm that first statement?

That is, when I visited Sweden for a consulting job about 15 years ago, there was definitely a secretary for the group I visited. She helped organize the paperwork for the visit, booked my hotel and taxi, etc.

The word for secretary in Swedish is "sekreterare", and a search of the Uppsala University domain find things like https://translate.google.com/translate?sl=sv&tl=en&js=y&prev... :

> Being the Secretary-General's secretary means helping coordinate, plan and prepare meetings, travel and reception of foreign and domestic guests as well as coordinate, prepare and plan other commitments. For example, involve helping to prepare information, book and plan trips and meetings, and to be responsible for the calendar as well as ex. certain world surveillance.

> You will also be responsible for calls, mailing and agenda appointments, and telephone answering services as well as daily office tasks such as copying, billing, updating of records, ordering office supplies, and providing easier computer support, etc.

There was a secretary for the group, not for any one professor.

If you'd visited in, say, 1925 the professors in the group might've had one secretary each.

That has more to do with those secretaries being replaced by computers than anything else. When everyone can type, and mistakes are easily corrected, you don't need people whose only job is to type anymore.

I'm Norwegian, not Swedish, but AIUI the societal trends were similar.

There's been a trend towards typing ability, yes, but hardly decisive. The old secretaries would do the bulk of e.g. organising conferences or the professor's travel, that's needs typing but not only typing.

The trend (in .no/.se etc) towards equality, against certain symbols of prestige, mattered a great deal. A long time ago having a secretary was a symbol of power, appropriate for Professor Dr. Lundman. At some point forgoing one became a symbol of modernity, and private secretaries became much less common, and the universities reorganised to provide specialised services such as ticket purchasers and maybe grant application writers instead of private secretaries.

>When everyone can type

I've spent some time as a troubleshooter of sorts in academia, and I can say that no, not everyone can type. Some are barely literate, especially when they have to use English rather than their native tongue.

Besides, secretaries do more than type. Online booking services have to a large extent replaced travel agents, but they are really just dumb shells on top of the travel agent's consoles, and consumes the user's time and attention.

As long as we don't have general AI computerising all but the most trivial tasks are going to be terrible to work with and waste a lot of time, especially when you have some kind of edge case and have to fight some business rules set up by some out of touch bean counter somewhere.

Certainly there's a trend away from private secretaries.

But the original statement was specifically about "ultra-low power distance societies, places like Sweden", and the process you mention is also taking place in the US, which (I believe) is not in that category.

Moreover, the statement added "the idea that anyone, no matter how important, is ever entitled to an assistant or secretary is almost anathema [in Sweden]." Yet I easily found a job position for a personal secretary.

That's why I would like a confirmation that M_Bakhtiari's observation is based on something stronger than hearsay or an assumption of how Sweden compares to not-as-low-power distance societies like the US.

I suppose it's taking place everywhere where assistant labour is expensive. Faster in the places where having a private secretary is old-fashioned, like Sweden.

> Could you confirm that first statement?

Yes, and as you can see from one of the comments below, the same is largely true of neighbouring Norway.

> That is, when I visited Sweden for a consulting job about 15 years ago, there was definitely a secretary for the group I visited. She helped organize the paperwork for the visit, booked my hotel and taxi, etc.

Notice that I talked about cultural views on secretaries, not the existence of secretaries within the country's borders. It's a downwards trend, I expect they're even more thinly spread now than 15 years ago.

It's not unexpected that the secretary-general of any organisation, even in Sweden, has a secretary. And of course theirs will be the last to go, with modern universities favouring administration over faculty, as it's the former that brings the cash, the latter only cost money (with that attitude, why not fire the faculty altogether?).

He's seeking research or data supporting your claims and conclusions, not for you to repeat your claim again.

For example, it would be useful to provide supporting evidence of secretaries declining in numbers/ratios related to cultural preference, not technology advances.

>He's seeking research or data supporting your claims and conclusions, not for you to repeat your claim again.

Culture isn't really very quantifiable, if anecdotal evidence coming from living and working in that particular society is unsatisfactory, feel free to disregard the argument.

>For example, it would be useful to provide supporting evidence of secretaries declining in numbers/ratios related to cultural preference, not technology advances.

But that's not what I claimed. I think that's what's happening, but I never claimed to know how strong the cultural component is compared to the technological and economic ones.

I wasn't seeking research data.

I wanted to know the source of your conclusions. Are you Swedish? Have you lived in Sweden for decades? Are you a sociologist of workplace cultures?

Or have you heard it third hand, or have you based it on your predictions of what Sweden is like.

Here you imply that you have lived and worked in Sweden, but you aren't clear about how long or how well you understand the history of the transition? (Eg, was it a big topic like the "Ni" question was in the 1960s?)

But really, how do you know if the transition in Sweden is any different than the transition in, say, the US? (As you write "organisation" instead of "organization", I assume that you are not from the US, and may not have the experience in that culture to make a solid comparison. Feel free to use another low-power-distance but not ultra-low-power-distance culture instead.)

BTW, you originally wrote: "the idea that anyone, no matter how important, is ever entitled to an assistant or secretary is almost anathema [in Sweden]." You then wrote "It's not unexpected that the secretary-general of any organisation, even in Sweden, has a secretary."

I interpret your original statement to mean that it is unexpected that the director of a university has a secretary.

If not, where would you place the boundary for where it is expected, and what makes Sweden's "ultra-low power distance" a significant contributor to that?

Because it seems to my non-Swedish eyes that in my limited experience in Sweden it didn't seem much different than my experience in the US.

For what it's worth, I plugged "sekreterare" into the job search at what I think is the Swedish employment agency. One of the results was https://www.arbetsformedlingen.se/Tjanster/Arbetssokande/Pla... . Google Translate says:

> As a secretary, you will assist a friendly, impulsive and simple person with; • write different investigations. • Create documents, such as in excelfiles. The workplace is located in Hornstull and you start your assignment with an introduction. The role can eventually be expanded and become more than the 60% as it is today.

That doesn't sound like a career which is "almost anathema".

I would expect that "maid" is something which is almost anathema, at least in the US. It looks like the Swedish word for that is "hembiträde", and there are no matches for it on that job site.

The idea that anyone is entitled to a secretary, that a secretary is a kind of status symbol, is rightfully falling out of fashion, and not just in Sweden.

But it's not being replaced by nothing, it being replaced by (non-personal) function staff that offloads specific tasks. The department has student relations officers that deal with scheduling, enrollment, tracking grades etc, the university has a travel office, email and digital calendars, available on mobile devices, simply removes a significant part of the need for secretaries in the first place - and professors have TAs (who would probably be responsible for a good chunk of the tasks described today). They just don't have general-purpose personal staff.

replaced by (non-personal) function staff that offloads specific tasks

And it's more practical since it works even if people in those positions gets sick or are away for holidays.

> The idea that anyone is entitled to a secretary, that a secretary is a kind of status symbol, is rightfully falling out of fashion, and not just in Sweden.

Indeed, the opposite error, seen in many ultra-high power distance societies, is arguably worse: when a secretary and a desk not facing a wall becomes an end in itself and nothing of value is produced.

> But it's not being replaced by nothing, it being replaced by (non-personal) function staff that offloads specific tasks.

Those things have been around for a long time, but treating it as a replacement rather than a complement to general-purpose secretaries means there will be no-one to do the more open-ended tasks, or tasks specific to the work of a particular person or group.

> there will be no-one to do the more open-ended tasks, or tasks specific to the work of a particular person or group.

Staffs still exists. It's not unheard of for a manager in a company to instruct his department to undertake some particular task. Sometimes that's practically all "business development" does. In larger firms, there is often an "office of the CEO" which has a number of people working on particular tasks for the top management, often with a title like "special projects", and often seconded from staff offices with the relevant skills.

The general idea of a secretary does make economic sense. In theory it's good if a professor spends more time researching/teaching and less time typing stuff and managing their appointments. I'm a generalist and would feel rather odd having a secretary but I can understand the general economic argument. It also makes sense to have a maid and someone who mows your lawn etc. etc. (except if you find the work inherently entertaining of course)

In Sweden and in other places, everything, including the universities, is nowadays run by economists. Since university professors don't produce anything that has any immediate monetary value it of course make perfect sense to save money by replacing administrative personel that costs money with the professors time, which is free since the professors pay their own sallaries with external funding.

> run by economists

I am confused. Perhaps you meant to say managers, administrators or bureaucrats ?

Formally they're the ones with the job descriptions, but under New Public Management it's the economists that are the ultimate arbiters of truth.

One amusing aside in that article was that one of the professors who appear to have received the budget to pay an assistant to nothing was Oswald Veblen.

His uncle Thorstein Veblen surely would have had a field day with this idea.

It is worth noting that Paul Halmos was an assistant of von Neumann's:


Thanks, this was a good read.

In the German Wikipedia there is a photograph of the lecture timetable, University of Berlin, 1928.


Von Neumann's scheduled his lectures as follows

- Wed and Sunday (!) 9-11 am; - Friday 3.30-5pm - Tue 7-9pm (Invited Talks / Faculty Seminar) - Mon 4-5 pm

I think that was uncommon even in the old days.

While academia is a place of good ideas and discussions, the power-distance and the use of students as very cheap labour with the promise of a higher place of importance someday is a dark side of it

Low skilled workers are treated better and more fairly than Graduate Students in a lot of occasions

Academia is stuck in the past in that sense. And it is already hurting them in quite a few disciplines where 'being published' is great but trumped by 'being able to afford a normal life'. So it filters out people that are willing to put their lives on hold in order to have their names at the top of a couple of papers. Some of them find that the lives they put on hold can not be put together again after a decade.

Graduate students are as a rule underpaid, overworked and if they're unlucky harassed to boot and all that for the privilege of having someone else then finally take most or even all of the credit for all the hard work.

Academia is similar to many other fields in that you roughly get what you pay for. That doesn't mean you can't save some money here or there or that money is not wasted here or there, but as a rule of thumb this generally holds.

My biggest quirk about the universities and institutes I've worked for so far is not the way they deal with grad students or postdocs, it's about seemingly 'unimportant' issues. Things like large, noisy, shared offices, lack of whiteboards, unnecessary bureaucracy, and lack of paid linguistic editing. At universities money tends to be saved at the wrong end, private businesses are much better at creating productive environments. Especially the constant lack of office space is hampering productivity. The priorities tend to be wrong (outside of a few centers of excellence), focused on getting as many researchers as possible. Instead they should spend a bit less on human resources and a bit more on a good work environment.

At least that's my personal experience and opinion.

Moving from a top research group to a fortune 500 company just last year, I couldn't agree with you more.

I had great resources in some places (access to the finest minds in my field, for example) and was completely hamstrung in others (a hard salary cap as an assistant researcher, extreme difficulty getting a new whiteboard installed in my office, etc)

Industry isn't perfect either but seems to be better at recognizing that researchers are best used in the role they were trained for and not doing ancillary work.

How is the money?

Significantly better than academia, of course. I get RSUs, annual bonus, etc. I actually got a fair amount of non-recurring money from patenting activity last year, as well.

I suspect I could make more money as a 'pure' software developer, but it would not be a factor of 2. And I'm doing what I love and largely choosing my own research direction. It needs to have an ultimate business application, but finding that alignment is not particularly difficult.

As with many scientists, having the resources I need to accomplish my research is often the stickier point. I have money for a few coops to do more speculative research, and annual equipment (I need clusters to do what I do). I don't have trouble getting funding for 2-3 conferences each year, which is all I need (more is a distraction, IMO). I'm hoping to start building out my group in a few years.

And I can do largely autonomous research work, very much like what I did as a researcher in academia. Arguably I am even more autonomous here, as I do not have a research adviser/boss.

I assume that having gone through academic hell, at least as a grad student so as to get your PhD, is basically a hard requirement to do the sort of thing you do now in industry, though?

As a person without a PhD, my sense is that "research" as a field is Not For Me, even in industry, because the PhD is the critical factor that determines whether one is "researcher material".

Yes, the position I filled explicitly required a PhD. I would be skeptical it would be a good fit for the vast majority of people who do not have a doctorate.

While I certainly would not call it academic hell, getting a PhD was very hard for me. However, while it was hard, I loved it. If you don't love it you probably[1] do not love research, and then you would absolutely hate the day to day.

[1] Probably because many things can make life miserable. A bad adviser being #1. I was lucky to have an absolutely incredible adviser who really helped me through. Of course, the other challenge is the very poor salary. I had a wonderful support network and no college debt so I could survive, and even then it was hard.

Excellent! Congratulations, it looks like you have a very nice mix of elements there.

Already hurting them, and has been hurting them for almost a decade, even more at this point.

I was thinking about continuing my education to do research in distributed computing, but I learned midway through my undergraduate, majoring in computer science with the "Software Engineering" track, that our field was already playing catchup with industry. I learned this because I worked for a startup my senior year. Their pace, learning, and new innovation was nothing our dusty class with out of touch professors were able to keep up with. Even the ones that were trying to be "cool" by teaching stuff like Ruby on Rails (mind you, this was in 2008 so it was the hip thing at the time) instead of Java Server Faces, it was just bad. You could tell the professors had no idea what it took to build software. After being in the industry about 10 years now, probably the biggest joke of a course I took and arguably a complete waste of money was CS 3100 -- Software Engineering. It was an entire class on Object Oriented programming with Java, and using UML diagrams to model everything, and CRC cards and all this object oriented cruft. Everything was an object nail to drill with your object hammer. It was at this point I was becoming more familiar with LISPs and dynamic/functional programming languages, and I started to care more about behavior rather than creating a tangled web of leaky abstractions.

Anyways, why go get my masters in distributed computing when I know it will be hilariously out of date and all the best innovation in distributed computing is coming from large, internet-scale companies such as Twitter, Google, Netflix, Facebook et all? But unless I could get into Stanford, which is essentially the big tech firm's R&D arm, I don't see the point. I was thinking of the Georgia Tech OMSCS, but I think I am too jaded at this point.

I have a similar story for a college course also titled "Software Engineering". Essentially, it was a course on how to properly execute the waterfall model of design.

We were assigned teams of four, wrote hundreds of pages of requirement papers, design papers, everything papers. Then we wrote the code in PHP in one night.

For the final section of the class, 3 teams of 4 were merged into teams of 12. It's obvious that no college students have the desire or ability to manage a team of 12, as it's a huge task even for an experienced manager. The overhead of having 12 people involved was incredible. Fortunately, more than half the students in the group were happy to be "fired": we didn't talk to them at all, did the work in a smaller, more nimble group, and everyone got a good grade.

Sounds quite outdated with the excessive documentation, etc. However, believe that learning a sequential process first has some advantages in shaping the mental model and experience with pros/cons later when learning agile. What do you think?

My embedded professor waged a little guerilla warfare against the Object Orientated Everything doctrine. Subtle little sentences like- fired mid lecture: "Yes, it helps us programmers, but the compiler has to rip it all out again, to not ruin the cache performance-"

which got us to question other lectures, and find out that the exact truth - depends on the circumstances.

PS: If you strip university down to its basics- its basically society accepted basic income for three/four years - to do whatever you want- if one does not use that, because s/he is chasing some gamified credits in exchange for real credit - is it really fair to blame that on the professors of that institution?

They must teach something, in a field, where half the innovations are basically secret and the other half is outdated the moment you write exams.

So what can you expect from a university?

We learned C/C++/Java.

Object Orientation is still a valid crutch for most programmers, who would otherwise fall back into procedural decay.

Functional programming languages are great, but hard on the novice mind.

Software engineering was one of those courses that are not usefull until you are out in the trenches. Still the Prof had valuable and valid war stories of theire own time in companys and lots of stuff you can find here on HN as advice including:

"He who writes down everything, is the last man standing."

Nobody claimed to have a silver bullet.

> PS: If you strip university down to its basics- its basically society accepted basic income for three/four years - to do whatever you want- if one does not use that, because s/he is chasing some gamified credits in exchange for real credit - is it really fair to blame that on the professors of that institution?

I am finding it difficult to unpack this, but the basic assertion that it is some free time for you to explore and figure yourself out -- that's great, if it is on your dime. If it is on someone else's dime, I would hope you would be somewhat a little more responsible with other people's money (That's how we ended up with all those memes about millennials: because there is a little truth in every meme).

Put another way, I worked my ass off delivering pizzas and working doubles on the weekend to pay for my nice state college. I am not trying to state some survivorship bias here, but rather illustrate the point of view you have when you work your literal ass off for every dollar. Spending it wisely then becomes your #1 priority. For all the sweat I put into being graced with the privilege of attending the "prestigious" state college of somewhere, USA, I wasn't about to spend another minute in there than I had to, because I saw waste and uselessness everywhere I turned and because I think I calculated my costs personally -- it came out to something like the rate of 200 dollars an hour per hour of actual education I received there -- That's a pretty nice consulting fee if you ask me.

Honestly these days, I think it's all a big racket, but I have hope for the future generation. These kids going to college now and racking up these student debts that will never be repaid, they are slowly but surely realizing they were duped.

Sorry, wrote very uncomprehensible.

In Germany college is free- so waste wherever you look- but i found i could invest that time and learn something there. And the professors where quite down to earth- also the sort of college i visit has a requirement that professors spend every fourth semester in the industry.

In contrast, as a consultant i met many companys, incredible proud on there tech- and only few of them really where as golden as they saw themselves.

I've been interviewing the products of many of those types of classes lately and discovering that none of them seem particularly interested in computers outside a classroom.

If you haven't even booted up a VM on a cloud provider to try stuff out, why do you expect to have a career in it?

True in most cases I'd say. In this case, even for $200/month and working hours of 60-80h/week I'd still argue most people would be proud to have this position. The learning, the networking effects, the constant intellectual challenge. I think this is also value, and in some regards a lot more rare than money.

Eh, I think my pride and job satisfaction would drop off quickly if my boss was constantly mad at me for not completing the work I was supposed to complete; and my loved ones were constantly mad at me for not making time for them.

Even learning and networking are not guaranteed. Your life is basically dependent on your relationship with your advisor. He/she can make or break you as a PhD student...

In my case, it's a bit different, but it has been disappointing enough that I'm trying to ditch the program with my MSc in hand.

I wonder how long his assistants lasted.

This was posted within today, not 1993. Title fix?

It almost entirely consists of a quote from a 1993 document, the fact that someone quoted this in an e-mail today is comparatively unimportant.

Hmmm. If you think what the source (Plover’s shitpost blog, not an email) provides is not beneficial, then I’d think the link should be to the original document.

I, on the other hand, think his selection and publication of this excerpt has merit and is useful. His framing of the quote is what drew me to read it in the first place.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact