202109080847 As We May Think

A seminal essay on machine augmented thinking written by Vannevar Bush in 1945.¹

He urges that men of science should turn to the massive task of making more accessibly our bewildering store of knowledge. For years inventions have extended man's physical powers rather than the powers of his mind. [These inventions] are new results, but not the end results, of modern science. Instruments are at hand which will give man access to and command over the inherited knowledge of the ages. The perfection of these pacific instruments should be the first objective of our scientists as they emerge from their war work. Like Emerson's famous address of 1837 on The American Scholar this paper calls for a new relationship between thinking man and the sum of our knowledge.

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.

Knowledge evolves and endures throughout the life of a race rather than that of an individual

Publication of research has been extended far beyond our present (1945) ability to make real use of the record.

Some thoughts on this quote:

• Is this still the case? Has the modern internet and computing technology succeeded in bridging that gap? Or has it merely accelerated it?
• How can I ensure that — at a minimum — the time I spend accumulating knowledge can be valuable and retained? How can I ensure that it all won't fall out of my head like a sieve?

A record — if it is to be useful to science — must be continuously extended, it must be stored, and above all it must be consulted.

He talks about storing the record via photography and puts great stock in that medium and its future. Some of the general principles came true but there was obviously a great emphasis on physical mechanisms and less about electronic information transfer (though there were impressive hints at it).

Then he moves on to extending the record — or what I'm referring to here as the act of writing — and the economy of producing new valuable information. He discusses text-to-speech and speech-to-text and paints a picture of a researcher in a lab or the field walking around taking pictures, speaking his observations into the record and tying the two together with timestamps.

One bit in this section that I want to think on is how he describes revisiting this record later at home or in the office. He suggests that you could watch the photos and listen to the record and then add more commentary via similar speech-to-text or something to augment the record.

• How does this relate to the idea of repetition and review?
• How can we iteratively enhance the record? Is there a digital native way that shows the evolution of thought over time in a given record?

There's some talk about automating rational thought that doesn't seem to have come to anything yet. The idea of programming axioms and getting conclusions out of them for more and more complex ideas. The critical issue we've seen with this in practice is the great infinite swath of useless, nonsensical yet logically correct manipulations you can make of the symbols.

The reason that mathematics is different than mere tautological symbol pushing is because of its use: because of the constructivist knowledge that it enables when interpreted by sapient beings. Yes the construction of mathematically proven theories may be tautological, but that doesn't reduce their significance.

• Is there a world in which we could really "solve" mathematics through machination?

Selection, recall, and extraction of valuable information from the record is critical above all. If the great heritage of human knowledge is expanding rapidly and we can never find the good thoughts entered into it, how will those thoughts every amount to anything. We will all be in silos of discovering things others have discovered.

• Are we already past the point where information is too wide and fast for people to be able to meaningfully use it to make the world or even their own lives better?

Some stock is put in photo recognition and its use as a helpful selector for his imagined increase in photographic records.

• Could photographic records be more useful than text? Are they hampered in practice because of inability to programmatically extract information from them (sorting, meaning, recall, etc.)?

The article gets further and further away from present reality by discussing systems for mechanical punch card selection and dry photography reading. Though examining someone's well considered thoughts from that time — even if they ended up being wrong — is intriguing.

• How can we read thoughts that go down the wrong path differently than thoughts that are predictive? Can we identify that quality of truth and prescience in our own thoughts?

Section 6-8 gets into a great discussion about the way our minds use associations or graph organization schemes for knowledge. Trails that aren't fully followed are forgotten, memory is transient. The description of the Memex is nearly identical to the idea of the Zettelkasten or Wikipedia that we're so familiar with.