Sunday, September 26, 2021

The first of the missing Second Millennium BCE paintings

The Lion Gate at Mycenae
©2021 Charlene Brown

In a blog post a couple of weeks ago, I mentioned I only had one painting of an archaeological site from the second millennium BCE, and that I was going to start painting the other three highlighted entries in the timeline for that millennium. 

The Lion Gate was the entrance to the Citadel at Mycenae in Southern Greece. It was built in about 1350 BCE and featured a relief sculpture of lionesses, symbolic of Mycenean royalty and the goddess Hera.  The now-missing heads may have been of composite beasts, possibly sphinxes.   

Sunday, September 19, 2021

One of my first computer-painted videos

The Garden at Government House, Victoria, BC
Watercolour sketch
©2009 Charlene Brown

Using this mid-October sketch as a starting point, I have computer-painted in both directions through all seasons and put the series together in a video that runs just over one minute. If you’d like to see it, click on  A Year in a Victorian Garden 

I haven't had time to paint much lately, so we're into reruns. Happily, this is because my family is making up for many lost months of travel and get togethers, and have been visiting us here on the island in waves for the past few weeks. This post is a repeat of one of my favourite blog posts from 2009, the first year I produced this blog.  


Saturday, September 11, 2021

Cross-cultural Time Capsule of the Second Millennium BCE


Click on image to enlarge

A few years ago I put together a cross-cultural 'History of Design' timeline covering art and architecture from prehistoric times to the beginning of the twenty-first century.  Three years ago I painted pictures of some of the entries in the pages of that timeline

Then I highlighted the illustrated entries in the timeline to show where these ‘time capsules’ fit into the big picture. 

Recently, I noticed that I’d left the second Millennium BCE out of that series of paintings, so I have started it now.  I looked through my files to see what paintings I might already have of that period… and found exactly one Deir El-Bahri, as seen from KarnakThis painting is based on a photo of the mortuary temple of Queen Hatshepsut at Deir el-Bahri on the west bank of the Nile.  I like the idea that Hatshepsut’s temple lines up with the huge Karnak Temple across the river, and took the photo from the entrance to Karnak. In fact, you can barely see Deir el-Bahri from that point, and you can’t see the Nile at all – but it seemed a good time for some artistic license.

Starting next week, I’ll paint the other three highlighted entries in the timeline above.

Wednesday, September 8, 2021

Heuristic input to Predictive Analytics

Cover Illustration for Inventing the Future with Haiku: Whistler P2P
computer painting
©2016 Charlene Brown

Because Artificial Intelligence (AI) has the ability to process ‘big data’ it can be applied to huge problems involving complex systems.  However, in order to be reliable in forecasting the future, it needs to incorporate the intuitive aspect of human intelligence.  

We need to find ways to build common sense into artificial intelligence.

It is possible that properly coded ­algorithms might eventually enable a computer to execute heuristic processes, but it is more likely that heuristic processes can serve as a good first step in data analytics by synthesizing data into a form AI can handle.

Glean as much as you can intuitively before you start quantifying and fitting number-crunching formulas.

I figured this out almost fifty years ago while studying what is now called business analytics.  It was called management science when I got my MBA, and operations research before that.  There was probably a lot less data to mine in the early `70s, but it seemed like a lot at the time. 


1.      The Delphi Technique is a method of arriving at a group opinion or decision by surveying a group of experts (usually in diverse fields) who respond anonymously, then have an opportunity to reassess their answers after seeing the aggregated response. It is especially useful when there is no true or knowable answer, such as in policy decision-making, or long-range forecasting.

·         There are apps that make political forecasts by using AI to comb through Twitter – sort of like a really big Delphi study without having Delphi participants’ opportunity to reevaluate their input.

·         The World Economic Forum Global Risks Analysis, “Visualized: A Global Risk Assessment of 2021and Beyond describes a rigorous method of quantifying expert opinion that sounds similar to a Delphi study, except input is not anonymous. See box below.


2.      Debating at a Policy Workshop: Policy resolutions were raised at the Liberal National Convention, held April 8 – 10 on Zoom. Resolutions had originated with various Commissions and Provincial branches, and were presented and workshopped on the second day. Several dozen were put forward for debate and voting on the final day. 

There were over 6000 of us attending the convention (virtually) and, for each resolution put forward, everyone had a chance to request debate (four debaters were selected and debates were conducted only if there were at least 50 requests – a necessarily arbitrary process) and then vote on whether or not to advance the resolution to the Election Platform Committee. That Committee finalized the Liberal Platform for the September 20 Canadian General Election.

 3.      Visualization: Data visualization, which can distill large data sets into visual graphics, can make it easier to understand complex relationships.  However, as determined in Visualizing Unforeseen Results, visualizations are seldom 'stand alone' documents.  Annotated visualizations may provide more easily understood explanations than detailed text-only analytics.

4.      Ideation Sessions: By employing ‘design thinking’ which considers input from experts in different fields – marketing, design and  engineering – working together, disruptive innovation ideation sessions can enrich discussions.  These sessions help participants to imagine 'what if?’ disruptions such as black swans, wild cards and events such as tipping points occur.

o   Black swan events: unpredictable, massive impact, highly improbable, – eg. 9/11, collapse of the Soviet Union, Covid-19

o   Wild cards: imaginable, low probability, high impact   The difference between Black Swans and Wild Cards is that Wild Cards are imaginable because they have precedents (ie predictable to a certain extent – temperature increases, Halley's Comet, 2008 financial crisis, religious conflicts, financial unicorns or alicorns).  

o   Tipping points: action of a system which has become unstable – eg. effect on crop yields of temperature increases.

Conclusion: Synthesizing information through soliciting wide opinion, debate, visualization, pattern recognition, trend analysis and extrapolation, in other words, going as far as you can in parsing the problem intuitively (heuristically), increases the likelihood of formulating a solvable optimization – and increases the chance the answer will actually make sense.


Saturday, September 4, 2021

Visualizing Unforeseen Results

Exponential Interactions
Computer sketch
©2021 Charlene Brown

In trying to convince some of my grandchildren that data analytics skills will be important for everybody, no matter what their career choice or field of study, it’s occurred to me it might be worthwhile to update my own understanding of data analysis -- and to explore the advantages of data visualization. 

Much has been written in recent years about the unexpected results of climate change. And during the current wildfire season even more has been written about the unforeseen results of climate change exponentially combined with other factors such as decades of fire prevention policy. Too much to comprehend, sometimes.

In Limiting data in search of information, Seth Godin points out, "It’s easy to be in favor of more data. After all, until we reach a certain point, more data is the best way to make a better decision. But then, fairly suddenly, more isn’t better. It’s simply a way to become confused or to stall."

Could data visualization, which can distill large data sets into visual graphics, make it easier to understand complex relationships?

The computer sketch above attempts to show some of the factors in the flow of results, both expected and unexpected, of the following.

Interactions between Forest Policy and Environmental Conditions

One of the unforeseen results of many decades of successful forest fire prevention in North America has been thousands of square miles of overmature, tightly packed, highly combustible conifers, particularly in National Parks — a perfect storm of wildfire hazards.

Climate change is also worsening wildfire conditions, in every possible way — increasing temperatures (especially in the North), heat domes and other extreme weather, dry lightning, pyrocumulonimbus clouds, and a longer fire season.

Meanwhile, over the twenty-year period of significant temperature rise in the Arctic, an unexpected phenomenon, a conifer-to-deciduous shift, has been detected in the taiga region of Canada. This increase in the proportion of deciduous cover has also been noted in Alaska as well, following severe and frequent fires in the boreal forest.

Quite possibly, this vegetation shift will reduce wildfire susceptibility because, as a rule, deciduous trees don’t burn as quickly or intensely as conifers, as I mentioned in my blog post, A word about deciduous trees. Over time this shift to deciduous forest may even mitigate the rate of climate change by improving carbon sequestration. Deciduous trees remove CO2 from the atmosphere much faster than conifers.

Another eventual good outcome of worsening wildfire conditions is the gradual elimination of problematic reforestation practices.

In the past, for reasons of cost effectiveness and efficiency, timber management policies have allowed:

  •       herbicide destruction of uneconomic deciduous trees
  •       clear-cutting 
  •       reforestation by planting softwood (conifer) saplings in evenly spaced rows on clear-cut land.

Instead of this, multiple species should be planted in clusters (to allow each species to develop and benefit from fungal networks among their roots) with deciduous clusters acting as fuel breaks interrupting vast swaths of conifers. The result would be a healthier, more fire-resistant forest — at a much higher cost than the old way.

Annotated computer sketch
©2021 Charlene Brown

Conclusions: The visualization at the beginning of this article, ‘Exponential Interactions,’ helped organize my thinking but is not exactly a stand-alone document. Annotations are required.

Annotated visualizations may provide more easily understood explanations than detailed text-only analytics.