Once upon a time, in a scientific periodical, there was a column named "Connections." The author of this column would take a scientific discovery, and a piece of trivia from the life of the discoverer, and would then jump from life to life, and idea to idea, pursuing connections between trivial details of these ideas and lives. The schtick of the column was that, towards the end of the column, the author would circle back, and connect an idea from the life of one scientific discoverer, to the original scientific discovery, with which he had started the article.
I was reminded of this by a new tool that K has proposed to me. She noted that this tool seems to encourage people to make connections between their ideas as they are writing down thoughts and notes. And, of course, she is correct. Making connections is a major part of how we make new discoveries. It is a major factor in doing research, and coming up with valuable insights. It is vitally important in this process.
Unfortunately, I am not sanguine about the possibility of these tools providing us with such insights. The author of "Connections," whom I spoke about, did not rely obvious connections between people, or, indeed, connections that would have been apparent in the simplistic searching that our computer tools are able to do so far. No, had an encyclopedic memory of scientific facts, and scientific lives, and, I assume, was also backed up by a team of researchers, who were able to support his efforts. All of these researchers were, of course, people. They were not simply finding matching items by running through masses of text. They understood when ideas could be grouped together.
There are a number of these tools that have been proposed to me over the years. I haven't found that any of them are terribly useful in what I do. I remember, some years back, the fad for mind mapping. I examined those tools, particularly since a lot of people started to use them, and then started to produce mind maps, and, if I wanted to read the mind maps, I had to use the mind mapping tools to read them.
Which brings me to Slade's law of Computer History": Those who fail to learn the lessons of computer history are doomed to buy it again, repackaged. Mind mapping is simply a somewhat more complicated form of a binary tree in database structuring. Binary tree is one of the oldest forms of database structuring. The latest round of productivity tools seems to be recalling network database structure. Again, network database structure goes back at least to the seventies, and possibly earlier. These new tools are simply a repackaging, and reinterfacing, of ideas that have been around for a long long time.
The item K found is an interesting tool, and an interesting experiment. She is correct: it does attempt to enable people to do, with machine assistance, what I tend to do when I am researching, developing, or writing. It would likely do it more consistently, finding all the connections when I might miss some--but, I strongly suspect, not as creatively. As noted, it is extremely unlikely that it would find the connections between ideas, rather than strings. An interesting trade-off. The fact that even the person reviewing this tool, having used a number of related tools (I have never even used basic mind-mapping tools beyond initial tests) says it has a steep learning curve does give me pause. And I find it interesting that the developers are making it cloud-based, and also locking it up. (Given my experience with companies that go for the IP lock up approach and mentality, I have a fair suspicion that the developers would be wanting to do a lot of data mining of the datasets that users would create with it. It wouldn't worry me, since I've always been pretty much open with my IP [and it drove Gloria nuts at times], but I imagine that corporations would go absolutely bananas when they found out.)
It's sort of like an automated (and constantly updating) concordance, and I've been using concordances for Biblical research for decades. I also know how frequently they fail you, unless you can remember the exact string to match. Maybe it can be enhanced with some sort of artificial intelligence that can determine, to a very limited extent, close ideas, as well as words and text strings. But I doubt we have that to a very useful level, yet.
It might be useful, but it would have to have a huge AI component to identify related ideas, rather than just matching strings. Which reminds me of some interesting comments that researchers are starting to make about AI. Right now we are all consumed with watching the vote for the Speaker of the US House of Representatives. Someone used the ChatGPT AI tool to create a speech nominating Shrek for Speaker. This speech, of course, has been doing the rounds on social media. Which also reminds me that the relations and collections of ideas that "seem" to appear on social media are, in very large part, because of AI tools being used in the background of all those platforms. Which is, in turn, making AI researchers realize that you can no longer use social media as "datasets" for training artificial intelligence on what humans say or how they think, because so much of the material is now contaminated because it has, for the most part, already been collated and connected by AI tools. Which has a spillover effect in the larger society, since so much of even the formal media now relies on social media for basic data.
How's that for a connection?
You use different tools for different types of development, and different parts of the development process. Normally, in writing a piece, I will use dictation (and Gboard) at some point. But, starting this one, I knew that dictation wasn't right, yet. I used Notepad to just jot down some notes, so that I could get all of my ideas in front of me. Then I used dictation, while looking at the notes and ideas, to flesh out the material, and get most of the verbiage in place. But Gboard is terrible at some of the finer points, so now I'm back to Notepad, to make edits, corrections, and, looking at it in static form, make sure everything makes sense. Often I use PowerPoint, and not just when it will be a presentation. I find it handy to note that, when a slide gets too "busy," I obviously need to reformulate, and sometimes regroup, a set of related ideas.
As a simplistic example of the whole process, right now I am working on a sermon. I was moved to write this sermon shortly after I got here to Port Alberni, as a result of some of my experiences in church shopping. I was not happy with the way the sermon initially turned out: it felt unfinished. But, eventually, a chance, throw-away comment, in someone else's sermon, sparked a connection with not one, but two, areas of research in my own field of information security. Adding those two ideas seemed to bring the sermon much closer to completion. In adding Biblical references I remembered an obvious one. But, by coincidence and chance, at the same time that I was looking to nail down the Biblical reference, in my own daily Bible readings I came across first a supporting idea, and then an extended idea, in a completely different part of the Bible. I wasn't able to find these latter references in my initial concordance searches, because the wording because while the ideas are almost identical, the wording is completely different. Thus, these new tools would be highly unlikely to make that kind of connection.
No comments:
Post a Comment