The machine intelligence was asking Google's Gemini how to do something in gmail, something I thought it might know something about. The problem had arisen because in connection with the business with NatWest previously mentioned at reference 1, I had completed a form which came as a pdf document. Submitting the form involved giving Acrobat permissions to access my gmail account, permissions which I wanted to withdraw once the job was done. With the main part of Gemini's answer being snapped above.
Which got me to the right place and did the business. But I could not get to that place myself, from inside gmail, and I had to use the link provided instead. And when you get there you have not got step by step instructions which exactly match what you are seeing on your screen, you have to use a bit of your own intelligence. And if you are unlucky, this is not good enough.
So what would be even better than what we have now, would be a Gemini which knew all about the computer which I was using and its patch state - and from that could work out exactly what I should be seeing and instruct me accordingly. Perhaps not go so far as to do the necessary for me as that would be giving it a bit too much control, at least for the time being. After all, I - no doubt along with plenty of other people - trust Microsoft with my computer. So why not Google, in due course?
A reason why not might be that it was almost as much of a pain keeping Gemini up to date with all the relevant information as it is to keep printed manuals up to date. So Google might not judge that to be a good use of their resources.
A useful by-product of all this was learning that the digital library called JSTOR had these permissions too, no doubt for some good but transient reason which I have now forgotten, which I was able to turn off at the same time.
Turning to other intelligence, there is the Australian memory problem first noticed at reference 1.
We suppose that memory can be thought of as a graph where the nodes are things in memory which can be retrieved and the edges are associations between those things. Associations which bridge in the sense that if A is associated with B and B is associated with C, then A is associated with C. Sometimes, if A is associated with B, then A is identified with B in some sense.
My knowledge of matters Australian is summarised - or perhaps caricatured - in the line of seven blue boxes upper left, starting with convicts on the far left. The seven boxes are associated by being in my Australian cloud.
The association of the three right hand boxes - rich mining lady, coal mining and Macquarie - is strengthened by their mining connection. Macquarie having been wrongly thought of as a mining operation, probably through their holdings in metals, whereas actually it is more financial services. No doubt all helped along by all the 'M' words involved.
I did not remember much about the rich mining lady beyond just that. Perhaps that she is a lady in what was a man's business and that she was a power in the land of Australia. There was nothing in the slots for what made her rich, for which company she owned.
Coal mining is something which the Australians do a lot of - exporting most of it to China and to other countries in the neighbourhood - and for generating their own electricity. And coal generally is seen as a bad thing because of its strong association with global warming. Not to mention all the death and disease among the coal miners.
This puts all three of them in the bad cloud, along with Thames Water, this last getting into that cloud on account of its poor performance, particularly on waste water. And then there is the alleged asset stripping. All of which strengthened the link between Macquarie and bad.
The association between the rich mining lady and Macquarie was now strong enough for them to be conflated, to which local problem the brains's sensible solution was to make the lady the proprietor of the company, to populate the aforementioned vacant slot. And once that has been done, it takes some serious event or effort to undo it.
A spin-off
Maybe there is some mileage in the following.
Let G be a large graph with nodes and edges. If X and Y are nodes, then an edge connecting X and Y can be written either (XY) or (YX). Edges are not directed, at least not in themselves.
While G might decompose into weakly connected clusters, generally speaking G is connected. Completely disconnected clusters do not interact and are not of much present interest.
Nodes have the non-negative real property activity and edges have the positive real property strength. One might move to a discrete world by saying non-negative and positive integers. One might have an upper bound. It might or might not turn out to be convenient to allow edges with zero strength.
Our graph exists in time. Nodes and edges have a life history, activities and strengths have history in time.
If X and Y are nodes and (XY) is an edge with a positive strength, we say that X and Y are associated.
If X and Y are nodes which are active at the same time, then they tend to be associated. And vice-versa.
Let A be a set of nodes. Then if X is a node and there is exactly one node Y in A such that (XY) is an edge, then A is a property of X. If A is a property of the node X, Y is a node in A and (XY) is an edge, then Y is the value of A for X.
One can say things about nodes in terms of their properties. For example, if A is a property of the node X, then A is a property of the node Y. Maybe the values of the properties of X are nearly always the values of the properties of Y.
Work in progress. Or maybe just starting out.
PS 1: and while we are being technical, I can report that we have just been moved off our old-speak BT landline - perhaps no more, at least in places, than a twisted pair - onto full fibre. The work appears to have been pretty well organised and everything is now up and working in the new fibre world. The cost to me was supplying a (yellow) ethernet cable - it was lucky I had one about - and feeding the new wifi network credentials into the four devices that needed to know them. Maybe just in time, as a section of the old landline has come adrift from our gutter and is flapping about in the wind. Maybe tomorrow I will think about getting a ladder out.
The cable which made the old telegraph pole side connection to the old router can be seen left above. While the bits on the right kept the spare cable for the telephone handset the radiator below. Now no needed as the telephone is now connected to the new router which is rather further away, at least as the cable runs.
PS 2: following reference 3, concerning some of the perils of legalising recreational drugs, I was interested to read the piece at reference 4, about the perils of legalising marijuana in New York City. Perils which I believe arise from having made the regulations around licensed sale of marijuana tough enough that there is a regular epidemic of unlicensed premises, premises making improper sales of improper goods. More work in progress.
I associate to a correspondent in Ottawa who was most unhappy about a bunch of rather unsavoury weed smokers who had taken up residence next door to him. Not least about the strong smell of the stuff hanging about.
References
Reference 1: https://psmv5.blogspot.com/2024/04/trolley-669.html.
Reference 2: https://www.jstor.org/.
Reference 3: https://psmv5.blogspot.com/2024/04/drugs.html.
Reference 4: The Battle Over Zaza Waza: A Lifelong Liberal Joins New York’s Weed War: When an illegal smoke shop opened across the street, Gale Brewer, a local councilwoman, vowed to close it. What happened next was “like a Fellini movie” - Jack Begg, Nicholas Fandos, New York Times - 2024.
No comments:
Post a Comment