As We May Think (1945) | Hacker News Reader

As We May Think (1945)

theatlantic.com

217 pointstosh55 comments

https://web.archive.org/web/20120822122558/http://www.theatl...

This article by Vannevar Bush inspired JCR Licklider to write his "Libraries of the future" [1] which in turn inspired the young researchers working on DARPA computing projects, who then went on to invent a large fraction of modern computing fundamentals at Xerox PARC.

Beyond just the indelible influence on the field of computing, Vannevar Bush was instrumental in setting up government institutions to support R&D in the build-up to WW2. Once the war finished, he also realized [2] that need for a post-war science research funding ecosystem to carry forward the institutional momentum & talent that had been developed during the war. He also considered his public engagement important enough that he wrote his memoirs to present his perspective and the whole arc of his activities over the decades [3]. It's hard to overstate his influence on the institutional setup for modern science & technology.

[1]: https://archive.org/details/librariesoffutur00lickuoft

[2]: https://www.goodreads.com/book/show/1024680.Science

[3]: https://www.goodreads.com/book/show/14290284-pieces-of-the-a...

He was also the one bringing the https://en.wikipedia.org/wiki/Frisch%E2%80%93Peierls_memoran... to the attention of the president.

I find him to be incredibly inspiring and "As we may think" even more so. The gravitas is palpable and reading this paper is one of these moments when "standing on the shoulders of giants" feels very much true.

edit: Regarding the memorandum, i found the following quote by Peierls especially chilling:

>Any competent nuclear physicist would have come out with very similar answers to ours if he had been asked: "What is the likely fission cross-section of pure U235? What critical size for separated U235 follows from this? What will be the explosive power of such a mass? How much industrial effort would be needed to do the separation? And would the military value be worthwhile?" The only unusual thing that Frisch and I did at this point was ask those questions.

Its hard to overstate how important it is to not make asking these kind of questions necessary again. I am doubtful about humanities chances for survival otherwise.

When I studied for my bachelor's in applied physics in the mid-70s several of those questions were standard exercises in the nuclear physics course.

> Its hard to overstate how important it is to not make asking these kind of questions necessary again. I am doubtful about humanities chances for survival otherwise.

Here's a problem though: those are the kind of questions I'd expect physicists to ask and give back-of-the-napkin answers to while killing boredom in a cafe or a bar. So I guess it's really more about what you'll do with the answers afterwards.

Dont ask, dont tell? And dont scan the napkin

Don't forget that Vannevar Bush also inspired Douglas Engelbart (it saddened him that Bush would not accept his and his team's digital computer innovations). The Mother of All Demos, by Engelbart's Augmentation Research Center (ARC) group was far ahead of its time and was a major influence on all of modern computing. It was a major and more direct influence on Xerox PARC than the also influential "Libraries of the future".

Engelbart was primarily motivated by augmenting the human intellect [1].

> Increased capability in this respect is taken to mean a mixture of the following: more-rapid comprehension, better comprehension, the possibility of gaining a useful degree of comprehension in a situation that previously was too complex, speedier solutions, better solutions, and the possibility of finding solutions to problems that before seemed insolvable.

Licklider also wrote:

> It is often said that programming for a computing machine forces one to think clearly...it disciplines the thought process. If the user can think his problem through in advance, symbiotic association with a computing machine is not necessary.

>...many problems that can be thought through in advance are very difficult to think through in advance. easier to solve...solved faster, through an intuitively guided trial-and-error procedure in which the computer...[revealed] unexpected turns in the solution

We have yet to realize the core vision laid forth in As We May Think and by other pioneers of modern computing. And if a certain group that thinks Tools for Augmenting Human Intellect are too dangerous to be spread widely has their way, we may never do so; at least not without first regressing for a substantially long time.

[1] https://en.wikipedia.org/wiki/Intelligence_amplification#Dou...

>It is often said that programming for a computing machine forces one to think clearly...it disciplines the thought process.

Gerald J. Sussman's "Structure and Interpretation of Classical Mechanics" (2001) also focuses on programming exercises for physics. In the preface they write:

>The requirement that the computer be able to interpret any expression provides strict and immediate feedback as to whether the expression is correctly formulated.

---

>Don't forget that Vannevar Bush also inspired Douglas Engelbart

Interestingly he read the July 1945 issue with "As We May Think" while stationed as radar technician in the Phillippines on an remote island. Having the time while waiting for redeployment and the radar technician background greatly contributed to his (and his teams) later inventions.

I have been reading The Dream Machine by Waldrop recently (great book!) that discusses Bush's pivotal role in the development of computing...which has a remarkable irony:

> Doggedly, and without success, [Bush] kept on trying to come up with a workable analog design for his Memex until his death, in 1973. And until the end, his colleagues could hear him grumbling about the "damn digital computer."

He wasn't wrong that digital computers weren't doing what he wanted. Still aren't.

What is beautiful reading this, in 2023, at my age, is the perspective for me. I actually grew up with people of that (1945) era. In the early 1970s most of the old folk in my village had seen at least one world war, some both, and as a kid I talked to them. The wild eyed wonder in the article, a mixture of hope and fear, was around then and is around today. The things that people - young or old - say, seem equally prophetic or naive. Plus ca change, plus cest la meme chose. The things they saw coming had not yet happened, and much they feared inevitable never did.

Very true, although the digital-ness of computers isn't holding anything back at this point, now that storage is ~infinite and the virtues of immutability are widely recognised.

Its worth mentioning that analog computing is making a revival with its applications for neuronal networks. Veritasium has a great video about that.

I wonder if this should be aggressively ignored these days as a fruitless path. What if we were to do a clean room reimagining. It's hard to ignore the history here but sometimes sharp turns are needed to move forward.

I've spent 20 odd years trying to piece all this history together as a side hobby and met everyone from Engelbert to Sutherland to Atkinson to Ted Nelson. I'm wondering what a radical departure could gain us.

Base it on say, Karl Popper and Media Literacy theory instead along with webs of trust.

The TBL system we currently use is too susceptible to information pollution with profiteers and propagandists as a perfect memex would also be as well as everything Nelson's ever cooked up.

We've maybe let it run its course. Kill your darlings and smash your idols to build tomorrow

> Base it on say, Karl Popper and Media Literacy theory

How would this work?

The problem is as cartoonist Gemma Correll quipped in 2016, "the fountain of knowledge versus the geyser of crap off the internet" (https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRqTgBk...)

The more serious analysis is that veracity and reliability of the net are worse than they are with paper and books because the barrier to entry and distribution is effectively zero. We haven't leveraged technology to increase the quality of information, just its accessibility and cost.

The quality and veracity should be part of the mechanism.

I'm trying to propose systems that have yet to be imagined, feel free to dismiss them, it's fine.

I should probably write an article on it. But we're talking 10,000 words, probably 4-8 weeks (I'd do interviews, citations, ask for reviewers, etc.) I can't really do that here in a comment thread.

The problem with the current veracity systems is that people with bad epistemologies put us in the same place as before. If millions of people thought astrology charts and dream interpretations were more valid than germ theory than our current veracity systems would hand out medical degrees to fortune tellers.

It's insufficient. You need theory of science to take a bigger role in the design.

I'm on vacation now but when I return I'll seriously consider it.

First off, thanks for mentioning Karl Popper and the media literacy theory. Will have to look into that. I wish you the best here.

Have been thinking about the issue as well and have been drifting more and more in the direction dealing with human error / aiding human thinking by just raising the bar abit without trying to find completeness and focusing on acceptability. Fixing the stuff that should be easiest. So looking at the description of errors and and how to communicate them in a sensible way. A minimum consensus based safety focused approach with capsuled complexity seems to be the way to go.

A few perspectives you may or may not have heard of and find interesting in this context. In the spirit of capsuled complexity and accessibility with good podcasts / videos:

Example for the concept and power of capsuled complexity with the 5 level explanation series from WIRED https://www.youtube.com/watch?v=raHM3k-uR0E

You already mentioned information pollution so you might be familiar with the whole military perspective, from operations in the information environment / cognitive warfare, Diss/Miss Information and missing context, psychographic advertisement, ODAA Loop and the research coming out of IBMs Watson by Juliane Gallinas with her Ted Talk about "(General)Patton in a Box" https://www.youtube.com/watch?v=NOQkPNlTTBY https://www.youtube.com/watch?v=-rg7uKsWlGs If things shift into a game theoretical perspective its also always worth while reminding people that even psychopaths like drinkable tap water. There is some win-win stuff.

(Air) Accident investigations and crew resource management for how to design processes dealing with human error. Mentour Pilots youtube channel or the series https://en.wikipedia.org/wiki/Mayday_(Canadian_TV_series). https://www.youtube.com/watch?v=PhO-87wwLJs about https://en.wikipedia.org/wiki/Eastern_Air_Lines_Flight_401 for example

Taiwans gov0 (zero) is an approach of consensus finding in politics. Outer alignment is also relevant here. I generally find totalitarianism (corrupt dysfunctional systems) a useful frame of mind in terms of why consensus finding has giant benefits. Joost Meerloos "Rape of the mind" has a good description of how these systems become dysfunctional. Also https://sproutsschools.com/bonhoeffers-theory-of-stupidity/ Noble lies and the problems in how to deal with them are also noteworthy here.

Hostage negotiator are a in hindsight not so surprising great source for consensus finding mechanisms and mindset. https://www.youtube.com/watch?v=8EguLJgkc54

Tim Urbans echo chambers vs idea spaces

You mentioned epistemology, ontology in computer science is a great counterexample for the dangers of aiming for completeness as well as Humes guillotine.

You might also like the stuff from Daniel Schmachtenberger on the sense making crisis. But i always feel always bad about recommending this since its hours and hours of videos. But from the types of information pollution to mementics to a description of the everything crisis we are in through the frame of finding a third attractor its quite valuable. https://www.youtube.com/watch?v=7LqaotiGWjQ

Tristan Harris might also be relevant for how technology (like social media and AI) is causing problems due to humans interacting with it. https://www.youtube.com/watch?v=xoVJKj8lcNQ

I would love to hear more about this!

Have you written up anything about your interviews and research?

What do you see as the major failings of the Memex / Web model?

What would you rather see instead?

What do you mean by "media literacy theory"? Sources or references on that?

> What if we were to do a clean room reimagining

That's a lost cause. Maybe not on a personal level, or among a circle of close knit friends (although it is still extremely difficult), but or the grand societal and historical level you refer to here: impossible. All you can hope for is to make dent in the trajectory of a relative isolated domain, since it will be an effort that competes with all the other efforts that want to accomplish change as well.

But better a dent in the right direction than no dent at all.

Pro-tip to avoid show-through when scanning documents on thin paper with text on both sides: put a piece of black paper on top so the reverse side luminance is constant.

Very cool pro tip. What sort of improvement can one expect?

> “Consider a future device ... in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.”

Incredibly prescient until the last sentence. In aggregate, we have foregone the supplementary nature of technology to humans, as it is now humans that supplement technology.

Why do you say this? The fundamental nature of carrying smartphones with us seems to me to be that we genuinely are able to explore and enjoy culture, shared experiences and knowledge in a completely new and powerful way. Of course, there are aspects of exploitation, surveillance and manipulation. But are they really more essential to the experience of internet technology than the huge possibilities that they have opened up?

> Of course, there are aspects of exploitation, surveillance and manipulation. But are they really more essential to the experience of internet technology than the huge possibilities that they have opened up?

A great question. Unfortunately today, I think they are.

What I see as the principle difference between the Zeitgeist of 1945 and the spirit of 2023 is our relation to struggle.

By '45 we'd probably realised that the war had tipped toward Allied victory. From then until perhaps the 1970s we were possessed of a mind that "Things are worth fighting for", and that science and technology are a struggle "towards good". We had not yet processed Auschwitz, Dachau and Belsen.

Cold-war notwithstanding, not only was the wonder of science on a heady incline, the fruits were magnificent, benevolent and cheaply available to all, from the highway to the boardroom to the kitchen. There's the "dream". Not only were we the benefactors of technology, but we were in control of it.

Today I find I must agree with David Graeber and Peter Thiel, that Francis Fukuyama's conceit of the "End of History" has seeped infectiously into the Western mind and extinguished optimism.

One of the things we must urgently fight - and it is as "worth fighting for" as ever - is the re-emergence tyrannical abuse through technology, and therefore to keep it in balance as a beneficent force.

Looking through these pages I'd say that sadly most of us, perhaps numbed by convenience and complacency, have lost the will for that fight.

On the contrary, they knew exactly what the concentration camps were. It's is us today that have no idea. Ask anyone in the west how many people died in the holocaust and they will tell you 6 million. Half the number. We don't count the other half and celebrate the people who killed Russians and Poles: https://www.wsws.org/en/articles/2023/09/25/lapr-s25.html

My grandfather, who devoted his life to nonprofit work, liberated concentrations camps firsthand as a US Army officer, and observed in his later years how common it was for people to remember the six million Jews who died but forget the six million Poles.

(The crime of antisemitic genocide is very very great, but if you doubt the Polish suffering, recall that the cry for “Leibensraum” was not purely or even primarily about the Jews, but it was undertaken with great military force, right?)

I have had a hard time independently verifying the Polish numbers to even share them with others so I agree the west ignorance on this runs deep.

I think it was in the book Dream Machine, mentioned elsewhere in the thread, where it described that there were largely two different philosophies of the relationship between the human and the machine/computer. One was "augmenting the human intellect", where technology serves to extend human abilities. The other was - if I recall correctly - cybernetics, where the human serves as part of a larger mechanical system. I might be remembering it wrong though.

I think Cybernetics was more like the primordial AI. But it was also expanded to study even larger systems like a collection of interacting agents like humans.

"If the user does not fully control the software, then the software controls the user."

I need to finish reading that book. But there's also Weiner's The Human Use of Human Beings which I think tries to address the human side of cybernetics.

This is a beautiful example of functional thinking. The ability to see the attributes and and requirements of a system rather than the items themselves.

Harvard offer a free online course called “lab to market” which among other things teaches a systematic way of using this to help uncover startup opportunities. In short, you use functional thinking to match the functionality of nascent technology “seeds” to the functionality required by an identified market’s needs.

Moreso, the implications, opportunities and hazards of such things.

It's a shame that this never got implemented, and now with the enshittification of copyright and "IP", it'll never happen.

The main feature was caching everything, and being able to make a copy of the "path" through things, along with a local copy of everything, in a nice exportable format, for your friends and colleagues.

---

Also, the fact that HTML does not allow you to mark up hypertext in a sane manner, doesn't help the picture. You can take a paper document, and add all manner of markup to it, without needing to alter the original source. Why can't we do the same with hypertext?

> You can take a paper document, and add all manner of markup to it, without needing to alter the original source. Why can't we do the same with hypertext?

This is a crucial point that I had never considered. Maybe we could mark this interleaved metaHTML with IDs or classes that make it easy to strip out. Or maybe HTML is a rendering format and we need something one step up.

Would it be possible to prototype something using the Wikipedia as a starting point?

> Two centuries ago Leibnitz invented a calculating machine which embodied most of the essential features of recent keyboard devices

Mmm, not exactly. The first calculating machine was invented by Blaise Pascal in 1642, three centuries before that article was written. Leibnitz "expanded on Pascal's ideas" [0] and produced his version of the machine in 1673.

[0] http://www.gwleibniz.com/calculator/calculator.html

Good point. Maybe send an email to the author to let them know.

Prior posts:

https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

Still. It's fun when the kids discover hypertext wasn't invented in the 90s.

It wasn't invented in the 1990s, but a strength of HTML was its relative simplicity. From my recollection, early attempts at hypertext added complexities like bidirectional links, which required more complex software to serve and develop documents. It may have been an amazing tool, but it was also one which a world accustomed to references (which is pretty much a unidirectional hyperlink) was not prepared for. It is also a facility that the modern web has proven largely unnecessary. (It has few uses outside of research.)

> (It has few uses outside of research.)

People loved trackbacks and other forms of bidirectional links, like monitoring where a link is shared on social media or checking quote-tweets. They're awesome for commentary and discussion.

The problem was never having 'few uses'. The problem was having too many uses - to spammers, as well as big tech companies trying to enclose social media to profit off walled gardens.

Over 2 dozen of 'em. Here's the only time it got more than 2 comments:

https://news.ycombinator.com/item?id=21736200

Wikipedia's summary of the author's importance in 1945 isn't great, but: https://en.wikipedia.org/wiki/Vannevar_Bush#World_War_II

Very interesting article. I hope my tangential comment is not inappropriate: Writing style in the Atlantic Monthly seems to have been considerably dumbed-down over the decades.

i have a vivid memory of reading this after kevin kelly (or possibly gary wolf) called this "the ur-document of the internet"

There are other antecedents.

H.G. Wells, "The World Brain": <https://archive.org/details/worldbrain00wells>

Paul Otlet's Mandeneum: <https://en.wikipedia.org/wiki/Mundaneum> <https://daily.jstor.org/internet-before-internet-paul-otlet/> <https://web.archive.org/web/20051227195507/http://people.lis...>

Generally, the Encyclopédistes, most especially Denis Diderot. <https://en.wikipedia.org/wiki/Encyclop%C3%A9die>

Consider a future device which stores books, records, and communications (1945) - https://news.ycombinator.com/item?id=34454819 - Jan 2023 (2 comments)

Highlights from the Memoirs of Vannevar Bush - https://news.ycombinator.com/item?id=32112040 - July 2022 (3 comments)

Vannevar Bush: A Public Sector Entrepreneur - https://news.ycombinator.com/item?id=30738264 - March 2022 (13 comments)

As We May Think (1945) - https://news.ycombinator.com/item?id=30687828 - March 2022 (2 comments)

As We May Think - https://news.ycombinator.com/item?id=22431079 - Feb 2020 (1 comment)

As We May Think (1945) - https://news.ycombinator.com/item?id=21736200 - Dec 2019 (15 comments)

The Technical Evolution of Vannevar Bush’s Memex (2008) - https://news.ycombinator.com/item?id=20900693 - Sept 2019 (4 comments)

As we may think - https://news.ycombinator.com/item?id=16444620 - Feb 2018 (2 comments)

HN Classics: As We May Think (1945) - https://news.ycombinator.com/item?id=16091725 - Jan 2018 (1 comment)

The Technical Evolution of Vannevar Bush’s Memex (2008) - https://news.ycombinator.com/item?id=12139602 - July 2016 (3 comments)

As We May Think (1945) - https://news.ycombinator.com/item?id=9787567 - June 2015 (12 comments)

1995 Vannevar Bush symposium: Closing Panel - https://news.ycombinator.com/item?id=7457056 - March 2014 (45 comments)

As We May Think (1945) - https://news.ycombinator.com/item?id=7381573 - March 2014 (1 comment)

As We May Think (1945) - https://news.ycombinator.com/item?id=4577865 - Sept 2012 (34 comments)

As We May Think by Vannevar Bush - https://news.ycombinator.com/item?id=3287344 - Nov 2011 (1 comment)

As We May Think (1945) - https://news.ycombinator.com/item?id=1565764 - Aug 2010 (20 comments)

Bonus: this is a treasure trove:

https://archive.org/search?query=subject%3A%22Vannevar+Bush+...

Thanks especially for that bonus!

People keep posting these inspiring, wide-eyed early visions of technology, a secret hope maybe that someone, somewhere can be enticed to go against the runaway dystopia of modern surveillance capitalism.

On a good day I think the trick will work. On a bad day I think it wont.

Nostalgia for optimism is latent radicalism and resistance.

So I think the days of "surveillance capitalism", or more generally the abuses of benevolent technology by the few against the many, are numbered.

But there's a long way to go from here to there. And it will take more than posting starry-eyed nostalgia.

Yeah, especially when Vannevar Bush also founded Raytheon thereby helping to forever entrench and fortify the military industrial complex in the USA. Optimism is a good and powerful thing but it must always be tempered with reason.

It's an interesting bind isn't it?

Which is better? To live in times of hawkish fervour, with the pointy side of technology facing some external foe. Or to live in "times of peace" when the bored, paranoid and over-zealous folks who are "good with tools" turn upon their fellows to exploit and control.

Personally I am more comfortable when I can't see the words "This side towards enemy".

Makes one think the best thing that could happen to humanity is facing some hostile extra-terrestrial force.

But William James already explored that in the "Moral Equivalent of War" [0]. We just can't seem to switch to seeing things like climate change, or AI-driven MegaTech as an enemy, because we look at it, and we see us [1].

[0] https://www.britannica.com/topic/The-Moral-Equivalent-of-War, https://en.wikipedia.org/wiki/Moral_Equivalent_of_War_speech

[1] https://en.wikiquote.org/wiki/Walt_Kelly

He was also the first head of the NSF (the American National Science Foundation, a government organization that funds much of the research that goes on in universities). Vannevar Bush did a lot in both the public and private sectors.

I’m all for dreaming of and working in on technology that actually improves people’s lives, and I also think we do a lot of that already (besides the obvious examples of all the medical technology that, along with the test of modern medical practice, made death while giving birth a rare tragedy instead of a common one, just think of music recording software, really, just imagine if Hendrix had access to a cd-quality porta studio on his pocket!), but let’s not forget the past was far from perfect, and had its own dystopias, far worse than getting tracked to be shown ads.

Specifically, we’re discussing a text from 1945, by this time the Hollerith tracking system had been used for years to manage concentration camps. The system was used on all aspects of camp management, including who was murdered, and when a prisioner was available for a specific work. That’s as bad of a technological dystopia as I can think of and it hasn’t been modern for about 70 years already…

So tl;dr: in agreement with you about trying to make good things, but let’s not fall for the rose-colored trap when remembering the past.

I just have to pause for a moment and shake my head at the fact that they put a nearly 80 year old article behind a paywall. That just strikes me as ridiculous.

Is it? I don't have a subscription to The Atlantic and the PDF version works for me:

https://cdn.theatlantic.com/media/archives/1945/07/176-1/132...

Reply on news.ycombinator.com