[FoRK] Non-speech, non-keyboard direct communications will create a new class of humans

Stephen Williams sdw at lig.net
Sun Jun 23 21:56:26 PDT 2013


On 6/23/13 9:06 PM, J. Andrew Rogers wrote:
> On Jun 23, 2013, at 5:12 PM, Stephen Williams <sdw at lig.net> wrote:
>> Many of us think, at least at times, much faster than we can communicate.  At the moment, we can only leverage and communicate stored information in external ways.  We could probably communicate several times faster, plus boost that with shorthand and keyword-like inclusion of packets of existing knowledge.  Given a much better visualization and representation method, we should be able to communicate a lot of information quickly in an absorbable way.
>
> I am skeptical of this assertion.
>
> Humans are able to rapidly retrieve arguments and facts that have been previously thought through and packaged but I can't think of an example of communication requiring novel and carefully constructed thought that did not move slower than the speed of communication. It certainly does not work that way for me and I have not seen an obvious example in other people. There are many subject matters where I appear to communicate extemporaneously with great detail and at a pace that is limited primarily by biological bandwidth but I am not *thinking* that fast. It is a speed of delivery borne of practice and repetition rather than thinking.
>
> Polymaths can do this around so many topics that it creates the illusion they can do it with every topic. But it is an illusion. Or at least, I've never met anyone who could actually think up new ideas and arguments constructively faster than they could communicate.
>
> And no one absorbs what I am communicating that fast, and I expect this is true for other people. Even when there is a considerable level of common context and intelligence, understanding of new information moves much slower than the speed of thought. For most people, most time is spent building the contextual groundwork.

Information that is well-known to someone can be dumped quickly without much thought.  It can be absorbed by well-trained people 
if the speaker is able to get enough feedback to provide information in an order that can be put in place immediately.  
Communication falls down when there is a big mismatch between internal representation of the subject area in the receiver vs. 
the order and model used to communicate.  Or when there isn't a model yet in that area and too much thinking is required to 
understand and relate.

Sometimes I can absorb nearly everything quickly and I'm impatient, usually because there isn't much of a delta but sometimes 
because I'm able to fit new information to an existing structure while staying focused.  Often I can't do this and I have to use 
repetition, but it happens often enough.  If you start rattling off various specs to a true car buff, they'll remember, picture, 
and validate the information very rapidly.  Being able to do this in many different areas is probably a good definition of a 
polymath; rapid understanding and absorption (and chaining to related and deduced / induced thoughts) is probably more 
indicative than spouting stored info.

In any case, it is well-known that people can generally understand much faster than most people speak most of the time.  Even if 
you simply compressed everyone's speech for faster, consecutive playback, you'd speed up communication.  Long ago I considered 
creating an audio conferencing system around this, but people would have to work with the resulting structure, so it isn't time 
for that yet outside of maybe standards group status calls.

If information could be dumped in a structured way, and especially if there is a convention for structuring information, it 
could allow the receiver to digest what is communicated in a personalized way. I think I have some useful ideas in this area.

On thinking, there are a number of cases where it is currently difficult to explain concisely what is being thought of so that 
communication lags behind the queue.  The simplest case is when you realize that you have to explain dependent concepts to 
explain something, which is mostly an information dump.  The next is doing small amounts of some type of algebra / calculus (as 
in "distributed systems architecture algebra") that takes a lot to explain but is fundamentally about manipulating a small set 
of symbols.  The mode that seems closer to full thinking is creative problem solving, although that also is manipulating symbols 
and analogies.  In these situations, direct communication links would allow polite and efficient negotiation of the need for 
explanation of dependent or referred to concepts.  Additionally, you could build up a library of explanations to be delivered 
without significant think time, thereby freeing up even more cycles.  Even in a group, you could spend your effort surfing the 
absorption / understanding status, addressing issues.  Call it "flipping the conversation".

I guess it depends on what you mean by "thinking".

I'm sure that you've had the experience of discussing something with someone who shares an overlapping knowledge background.  
You can use concise terms and words to imply whole architectural pieces, principles, modes, or other idioms at different levels 
without detailed explanation.  Perhaps one of the easiest communication improvements would be delta transmission (cached) of a 
wiki-like body of glossary, with explanation, links to relevant and source material, etc.

Even the simple case of referring to, representing, and communicating links would be very helpful.  Absolute identity of people 
or companies mentioned.  Establishment of social links and communication channels via easy mental decisions would be valuable 
shorthand.

>
> A lot of stuff pops into my head continuously but it is half-baked. Fully baked ideas emerge at a much slower rate and those are the more valuable ideas.
>
If you are collaborating on brainstorming, design, and engineering, sharing the half-baked ideas is valuable also.

sdw



More information about the FoRK mailing list