Toronto: sometimes good, sometimes bad

A lotof good events have happened in Toronto since my last newsletter in September. Some bad, too.

First up, the biggest of good events: It. has. finally. happened. Finally. After more than a decade of delays, the Eglinton Crosstown LRT launched in February! I honestly wondered if it would ever open. People didn’t know what to believe about possible opening dates. But it happened. After 15 years of disruption on Eglinton, it happened.

Should there be an investigation into this construction fiasco, given it only took 6 years to build the Chunnel and 40 months to build the CN Tower? I think so, though it’s doubtful. I don’t even think residents will get an apology. People like Doug Ford and others will just shrug and move on or say we should be grateful, I guess. As someone who had to live with this mess for all of that time, I am feeling disgruntled and I expect to stay that way for awhile.

In other transit events, this was good: the Gardiner Expressway reopens over a year ahead of schedule. Yay! Not sure if it is good or bad but the Dundas subway station just changed its name. However I noticed that Dundas West still is…Dundas West. Perhaps that will change some time in the next 15 years.

In non-transit events, the other big bad thing that happened since my last newsletter was the Toronto Blue Jays’ heartbreaking loss of the World Series to the LA Dodgers. Hey, I am glad for Los Angeles: that city has suffered much recently and could use the win. But I would have been delighted if we had won it. I got to go to the Skydome and see the Jays win the last two World Series: I would have loved to see them win again, even if — like everyone else in Toronto — I was watching it on TV.

For more on that plus other good things that happened in 2025, go here.

Another bad thing that happened in the last while was eight current and former Toronto police officers bein charged in the Project South probe as part of an organized crime bust. You know things are bad when it gets worldwide attention in newspapers like The Guardian.

Homelessness is not a new thing in Toronto. There’s never enough resources, it seems. To see what I mean, you can check here and see: Daily Shelter & Overnight Service Occupancy & Capacity – City of Toronto Open Data Portal over at open.toronto.ca. However, one good thing that happened was the Slaight family donated $25M to fight Toronto homelessness.

The Ford government recently privatized the blue box recycling program in Toronto. It was bad at first, with many complaining about the botched roll out of it. Even on my street things were not picked up for a time. It seems to have sorted itself out though.

Speaking of the Ford government, it seem they will not be  proceeding with changes to rent control laws. That’s likely a good event for tenants. In terms of other events, Mayor Chow and the city unanimously  passed their recent budget. John Tory toyed with the idea, but it looks like he will not be running for mayor in the next election. But maybe another member of the Ford family will? Who can say.

I can say that these are all good: Toronto’s best diners: 5 local icons picked by Toronto Star readers. Also good? This piece on five of Toronto’s oldest restaurants. New restaurants are talked about all the time. It’s good to see the older establishments get more recognition.

Finally, I must say, it’s been a winter. I am looking forward to spring. But before I go, here are some good Toronto links I want to share:

See you ’round six months from now. In the meantime, go see the Cumberland Terrace while you still can.

Be aware and avoid URLs that start with https://storage.googleapis.com/

No matter what I search on these days, be it software or salad, I will see websites that start with the following: https://storage.googleapis.com/

In my case, it happens when I ask to see just the images of my search results. The images look appropriate, but when I click on the link, it will take me to some site which has nothing to do with the search or the image. Who knows what else is happening when you go there.

Based on this, it looks like it is not a new problem. You might think Google would want to clear it up, but apparently not.

So keep an eye out for that URL in your search results, and make sure you do not click on it no matter what.

On the iron triangle of project management


Everyone who works should be aware of the iron triangle of project management. In essence, it says:

  1. The quality of work is constrained by the project’s budget, deadlines and scope (features).
  2. The project manager can trade between constraints.
  3. Changes in one constraint necessitate changes in others to compensate or quality will suffer.

Another way of looking at it is with this formula:

  • Quality = (Time X Resources) / Scope

If the scope increases without increasing either the time it takes to get done or the resources to get it done, then quality will decrease.

I think about the iron triangle a lot as a consultant. Clients often want to fix the values for time and resources while increasing the scope and not decreasing the quality. Sometimes consultants will want to increase the time and resources on a project without increasing scope in order to provide higher quality. It’s a balancing act, keeping all four measures aligned such that everyone is happy.

Anyone who provides a product or service to a client or customer should be aware of the iron triangle. Work makes more sense once you are aware of it.

 

There are greater inventions than the Internet or AI

What is the greatest invention since the 19th century? If you’re someone like Om Malik, you might say the Internet. Others might say the personal computer, or computers. AI might be the choice of others.

While these are all great inventions, I’d nominate two others: medical technology and farm machinery.

 

Medical tech like vaccines and blood transfusions alone have saved billions of lives since they were created. And that’s two out of the many medical inventions that have increased our life expectancy according to this. It’s why we now can expect to live to be 80 as opposed to 40.

If that technology allows us to live longer, farm machinery enables us to do more than work the land in order to survive. Tech like tractors and the McCormick Reaper make it possible to harvest more food than ever before with much less effort. Other inventions like fertilizer enable us to get much more out of our harvests.

The Internet and other IT are grand inventions that help us in many ways. But I don’t think any IT can hold a candle to the benefits brought to us by medical and farm technology.

P.S. For more on farm technology, go here.

On Canadian conceptual art

When you think of Canadian Art, it’s likely you might think of artists like the Group of Seven or Alex Colville or other representational artists. But Canada also has some very good conceptual artists. I came to that conclusion after do some research on the late great Kelly Mark and all the artists in her sphere. Here’s some links highlighting the work of Mark and more:


I would be remiss if I didn’t point out one of my favorite Canadian artists, Micah Lexier, whose work can be seen on tumblr and who also has a good Instagram account. And if you live in Toronto, you no doubt have seen his work without even knowing it is his. Check out his tumblr and you’ll see what I mean by that.

The arts on Wednesday: more VanGogh plus graphic design masters, Futurists and more

I’m always reading about artists in such places as the New Times, the Guardian, Wallpaper and more. These 2 dozen pieces below capture just some of the good things I’ve found in those publications since my last post on the arts in October.

First up, some good pieces from the Times. I highly recommend the first two, on Van Gogh and David.

Each Friday the Guardian posts a newsletter on the arts, and I have to say, it is consistently rich and full of good things to check out. We are lucky to have such art coverage in such a major publication. Here’s just a small sample of things I found there:

Wallpaper also has good arts coverage. It tends to be contemporary artists, like these pieces:

 

While I love good painters and sculptures, I have a fondness for artists who work in collage and other forms that involve graphic design. Kalman, Heartfield and Kruger are just three examples found below, with some other artists mixed in:

The following don’t have any particular idea tying them together: I just thought they were interesting and worth a look:

(Top image from the piece on Van Gogh, middle image from the work Kalman did for the restaurant Florent in the 1980s).

To deal with aspirational clutter, aspire to something better (think Japanese minimalism)

As someone who struggles with a lot of clutter, I find some of it easier to deal with than others. Anything that can go in the recycling or the garbage I find easy to toss out. But even when I do that, I still have too much stuff lying around.

The stuff I find hard to get rid of is stuff I find useful and that will make my life better somehow. So I was happy when I found this article touch on that: What is aspirational clutter, and how do I get rid of it? It’s a particularly difficult form of clutter to banish from your home because you have to adjust how you think about yourself in the decluttering process.

You probably have lots of aspirational clutter at your home now: those art supplies you never use, those books you will never read, those tools you never pick up, those clothes you will never get to wear. And much more. The usefulness of things and the hope you can become better means they are especially hard to get rid of.

One way to do so is to aspire to something better than also helps you get rid of things. For me, I am inspired by these designs found here: 13 Productive Japanese Home Office Designs – Edward George – edwardgeorgelondon.com. As something of a maximalist, I doubt I will ever get that disciplined in redesigning my work space. But having a goal to become more like that could help me banish so many of the things I have lying around now that someday — SOME DAY — I will get to (but probably won’t).

If you find you still have clutter around your house even after big declutter sessions, ask yourself: is this aspiration clutter? And if it is, ask what do you have to change to be able to deal with it? Perhaps aspiring to live simpler is a way to do that.

For more on Japanese minimalist, check this out: How to Embrace ‘Ma’ (間) and Japanese Minimalism In Your Home – theartofzen.org

A really good AI prompt to help you get the most out of AI

If you do anything with AI, then knowing how to write a good prompt is essential. It doesn’t matter if you’re doing complex vibe coding or merely sending a question to ChatGPT or Google Gemini or Microsoft CoPilot, the prompt you use is the key to getting good responses.

Recently I was happy to come across this post on swiss-miss.com on the perfect AI prompt. I took the liberty of copying it below so I could add my own comments in parentheses. A good/perfect AI prompt should have:

  1. Goal – Define what you want clearly. (Use action oriented verbs and imperative sentences when you send a prompt to AI systems. Limit your goal to one thing. If you want several things, specify it in such a way that it seems unified. Or ask for one thing, and then put more details in the context dump).
  2. Return Format – Specify exactly how you want the response structured. (Do you want your output in list form, ask for it. But be creative and expansive. Do you want it in PDF format? Or an XLSX file or a CSV file? Maybe you need a markdown file or a Word document? Then ask. Also specify length, like “provide a one page markdown document” or “produce a list of 10 items”).
  3. Warnings – Highlight any important details the AI should double-check (e.g. if you are asking for resources, specify you don’t want out of date resources)
  4. Context Dump – Provide background info to improve the response (this is a good one. The more additional information you can provide, the more likely the AI system will provide the answer you want. Warnings and the context dump help frame things for the AI system. Never assume it knows what you know or that it even remembers what you asked it just recently.)

Prompt specs will continue to change as generative AI systems change. For now, this is a good guide to get better results from your AI queries. You can still specify the role you want the AI to take on, though I see people doing less of that. Regardless, detail is important in your query.

P.S. Thank to the blog swiss-miss.com for this. You need to go there: it has decades of goodness you should know about.

From vibe coding to no-code / low-code AI generated coding


People are beginning to stop writing code, though lots of code is still being written. The reason for this is “vibe coding”, or more accurately, AI generated code.

I’m seeing examples of AI generated code frequently now. For instance, Spotify says its best developers haven’t written a line of code since December, thanks to AI. Over at a company named StrongDM, they’re going pure AI and the humans don’t even look at the code. Claude Code has plenty of references. I’ve even seen Reels on Instagram showing this in action.

I am finding this is happening with me as well. I can write code so well now using AI tools like Roo Code and IBM Bob that I am focused more on the prompts I need to come up with in order to generate the working code, the documentation, and the test code that insures the working code is of sufficient quality (which it usually is). Code that would have taken me a day to write I can get done in less than an hour, and software efforts that would have taken me a week I can get done in half a day. (As always, YMMV.) I’ve written about my experiences here and here.

Given these recent developments , I think we are approaching a new stage of software abstraction. Just as we went from machine code to assembler to third generation languages (3GL) like Fortran and C, we are now at the stage of abstraction: AI generated code. Developers like me are shifting from a coding role to a governing/controlling role. The coding role of generating code is now done by the AI tool: the human role is insuring the code is correct.

In some ways this is not unlike no-code / low-code. But instead of using a tool that provides “visual building blocks like drag-and-drop and pull-down menu interfaces” to write code, the developer uses AI to do it.

We all depend on software every day, from the apps we use to the online systems we work with. What’s changing is who will write that software. Going forward, less and less of that code will be written by people.

P.S. here are some links to pieces on writing software using AI I recently came across that were good:

For more on low-code and no-code, see this. For more on cybernetics, see this.

P.S.S. As an aside, I think it’s easy to imagine going from having a human developer using AI to no human at all, just AI. Easy or not, I think we always need humans in the loop. We still need humans to direct the AI. More importantly, we need humans to govern and control AI, now and in the future. Think of the cybernetic loop above. Replace the Sensor with AI. Now think of the other system. If humans ultimately depend on the other system, then a human should be the controller. Imagine the catastrophes that can occur otherwise (think of a spiralling downward feedback loop).

Disclaimer: As always, these are my opinions and not necessarily those of my employer.

 

The Webcam of downtown St. John’s is a delight

I love this webcam so much I decided to post it here so more people can know about it. It’s a camera provided by the CBC that focuses on downtown St. John’s in Newfoundland, Canada.

(Link here)

It’s a perfect webcam if you ask me. The setting is as picturesque as anything. There is just enough activity happening to make it seem real (as opposed to a still photo): cars go around constantly, birds fly by, and occasionally you’ll see a boat in the harbour.

CBC has other webcams, but this seems by far the best. Check it out if you have a minute. It’s perfect especially for anyone from Newfoundland feeling homesick.

On exercise for the mind instead of the body

I often talk about exercise from a physical point of view, but not from a mental point of view. I thought I should expand my idea of exercise when I came across this quote from G.K. Chesterton:

It is a good exercise, in empty or ugly hours of the day, to look at anything, the coal-scuttle or the book-case, and think how happy one could be to have brought it out of the sinking ship on to the solitary island. But it is a better exercise still to remember how all things have had this hair-breadth escape: everything has been saved from a wreck.

It is an exercise not of the body but of the mind. Mindfulness is an exercise of the mind too. In both cases we are training our brains to do something different than it normally does, and by doing so, we become more fit and up for the challenges life can throw at us.

Athletes do mental exercises all the time, especially in the area of visualization. We should all do such workouts on the regular.

Quote via this piece which I found on Austin Kleon’s newsletter. Image of Chesterton from Wikipedia.

Apple is marching to the beat of a different drummer

Apple is going in a different direction than many of the other big players in IT. I thought of that when I saw this news:

AmazonAMZN $210.05 (-5.56%), AlphabetGOOGL $323.78 (-2.51%), MetaMETA $663.95 (-1.32%), and MicrosoftMSFT $403.00 (1.82%) all spent record sums last quarter on purchases of property and equipment — largely tied AI chips and data centers. And for the companies that offered forward-looking guidance, their capex plans for the year blew analysts’ already generous estimates out of the water.

Amazon expects its 2026 capex to surge to $200 billion. Google is aiming for $175 billion to $185 billion. Meta estimates it will spend between $115 billion and $135 billion. All of those figures came in well above expectations and, for the most part, have weighed on their stocks. Microsoft didn’t give a formal 2026 capex outlook, but if its peers are any indication, spending will likely exceed the roughly $114 billion Wall Street expects for the calendar year.

Of the Big Tech companies, just one stands apart this earnings season. Apple’sAAPL $277.08 (0.77%) capital expenditure, already just a fraction of its peers, actually declined in the December quarter from a year earlier.

All those other companies have capex increases rising like a hockey stick: Apple’s has been relatively flat in the last 10 years.

So what is Apple doing in the way of AI?

Apple has struck its own path with AI…  it’s embracing AI but is not an AI company… Apple’s decision to use Google’s Gemini, rather than an in-house model, to power the next generation of Siri and Apple Intelligence. The Google deal, reportedly worth about $1 billion a year, gives Apple access to a top-tier AI model for pennies on the dollar compared to what other Big Tech companies are spending to build their own.

It makes sense to me. Apple is not going to be an AI company any more than it will be a search engine company. It’s betting if anything that AI will become a software/SaaS commodity locked in to one company due to network effects and it will just be part of the overall solution that Apple’s products offers it’s customers.

Speaking of going in a different direction, I’ve recently updated my OS on both my Mac and iPhone and was surprised by the introduction of Liquid Glass and removal of  the flat (bland?) UI I’ve become accustomed to. It reminds me of the skeuomorphism of the older versions of the UI I once liked. Sure it ain’t Aqua, but it does make the screen more lively. I like it, but not all do. As the French say: plus ça change, plus c’est la même chose. For a thoughtful analysis of this new UI and what’s good and not so good about it, I recommend this piece.

 

On the rise and fall of Scott Adams, cartoonist of Dilbert

There were two Scott Adams: the one known for being the creator of Dilbert, and the one who recently passed away who was beloved by right wingers in America. People like me who knew the first Scott Adams and who loved his Dilbert comic in the 90s were dismayed when he morphed into the second Scott Adams.

If you ever wondered about him as I often did then I recommend this piece by Andrew Farago. It does a really good job covering the two Scott Adams and explaining the transformation. If you were bewildered as to how the transformation could even occur, Farago helps with that bewilderment.

For more on Dilbert, see the Wikipedia page (which is where this image resides).

On lawgivers

In the House Chamber of the U.S. Capitol are 23 marble relief portraits depicting historical figures noted for their work in establishing the principles that underlie American law. This website shows the portraits and has a short description of each of the men displayed. It’s a great place to get learn more about historic law givers, and a good reminder that the law is not static and should not be presumed.

It’s February 2nd. Happy Candlemas to you!

February 2nd is Candlemas and if you are a devout Christian you likely know about it and celebrate it. For most, however, it is likely a forgotten event, unlike Christmas or Easter.

I think that’s too bad. I like the idea of a ceremony that creates blessed candles “to be lighted at times of stress – during storms, in sickrooms and at the bedside of the dying” (as this piece explains). We could all use such candles, whatever our beliefs, to get us through hard times. Even to get us through winter.

If you are struggling to get through this difficult winter and this difficult month, might I suggest you acquire your own blessed candles — even if you have to bless them yourself — and get them out and light them when the time arises. You deserve no less.

P.S. When it comes to February, Austin Kleon provides plenty of ways to look at the month in a positive way. I’ve written about that here and here. I also recommend you check out this post he made in 2023.

P.S.S. If you are wondering if there are other -mas holidays you’d like to know about, I’ll direct you to this. There are ALOT, including Michaelmas, Marymas, Hallowmas, and many more.

Image is of the painting “Candles” by Gerhard Richter.

 

How to reward yourself after a difficult task

How not to reward yourself after a difficult task is to immediately move on after you do it. That’s what I would do, and later I would wonder why I always had to use some much effort to do it.

What I try to do now is to take time after the task to savour my accomplishment. I think positively about what I did and reward myself with positive talk. If that sounds dumb, ask yourself: do you “reward” yourself afterwards with negative self talk? Because THAT is dumb. (Also neutral talk is not much better: it’s the “bread and water” of self talk). (Image: Philip Oroni)

On the quote: never give in, never, never, never

You may be familiar with parts of this quote:

never give in, never give in, never, never, never, never-in nothing, great or small, large or petty — never give in except to convictions of honour and good sense. Never yield to force; never yield to the apparently overwhelming might of the enemy.

The source of the quote is here and was given by Winston Churchill on October 29, 1941 to an audience at Harrow School.

It is often abused and misused. For example:

WRONG. 🙂

Never give IN.

In these difficult days, in this difficult time in the world, it is a good quote to remember.

So too is the reframing of events at the end of his speech: “These are not dark days; these are great days”. How we rise to the occasion on dark days can transform them into great days.

On vibe coding with Roo Code

I’ve been using AI tools to write code lately. Vibe coding I guess, though I am not a fan of the term. A few months ago I was using Copilot and Claude to do it. In the last few weeks I’ve been leaning heavily on IBM BOB (which I’ll write about soon).  I’ve also been experimenting with similar tools like Roo Code to vibe code.

Roo Code can do several things according to the docs,  but the main things I’ve been getting it to do are:

  • Generate Code from natural language descriptions and specs
  • Refactor & debug existing code
  • Write & update documentation

There are a number of ways you can put it to use. Because I am a fan of VSCode, I am currently using the Roo Code VSCode extension. Installing it is straightforward, though I should mention I am currently using version 3.30.3 because I was informed by colleagues that it was a stable version, and the latest version 3.42 had issues.

Once installed, you will need to connect to an AI provider. Roo recommends Claude Sonnet 4.5 and so I went with that, albeit an internal work version. I expect the external version of Claude Sonnet is fine.

Roo Code has a number of modes it can operate in: Architect, Code, Ask, Debug and Orchestrator. Today I wanted to get it to help me write some software so I opted to work in Code mode.

I gave it the following task: Can you write software that will look at my Google Chrome browser’s history for the last week and turn it into an HTML file that is a summary of what I looked at on the web this week?

Roo first set up a list of things it is going to perform. In my case, the Todo List was:

  • Analyze Chrome history database location and structure
  • Create Python script to read Chrome history
  • Extract last week’s browsing data
  • Generate HTML summary report
  • Test the solution

Roo then went off and did these things and tested it by running the following command: open chrome_history_summary.html

Pretty good! Alas, I clearly used too much social media this week, so I gave Roo a follow up task:  Can you have an option where I can have a list of web sites to exclude from the report? For example, I don’t want to see all the times I went to instagram or bluesky.

Again Roo proceeded to create and execute a new Todo List which included modifying the python program it first created, adding a separate chrome_history_config.json file which can be used to filter out web sites from the report, and finally a README.md file that explains the programs features, requirements to run, how to use the program, and much more.

What’s nice is it doesn’t just create a program to do that, but it provides different ways to run the program:

# Basic usage (uses config file if present)
python3 chrome_history_summarizer.py

# Create config file
python3 chrome_history_summarizer.py –create-config

# Exclude specific sites via command-line
python3 chrome_history_summarizer.py –exclude instagram.com bsky.app

# Custom output file
python3 chrome_history_summarizer.py –output my_report.html

I could have asked it to change aspects of the page like the fonts or the background or more, but I was fine with what it produced.

In terms of cost, Roo can you give you a breakdown of how much tokens it uses and how much it costs. In my case it estimated the cost at 27 cents. Of course I could have written a similar program for free, but instead of taking me minutes I would likely have spent hours. My experience to do with AI coding tools like IBM BOB and Roo Code is I get about a 5-10X productivity gain. YMMV.

What’s nice about using Roo in VS Code is that you can have the code execute in VS Code. While I had no issues with the code, if I did I could have put some breakpoints in the code and debugged it on the spot. And unlike when I was using Claude and Copilot, I don’t have to keep copying and pasting the code into an editor before I get it to run.

Speaking of that, here’s a tip: start in the directory you want the code to appear in before asking Roo anything. I changed to my directory using Terminal in VSCode and then I started asking Roo questions. Otherwise it may write into a directory you don’t want it to.

All in all, Roo is easy to install and use. They have good documentation too. Interested? You can get Roo here: https://roocode.com/

P.S. As for everything on this blog, the opinions expressed above in this post mine only and not necessarily those of my employer.

 

In 2026 I am still not shopping at Loblaws or Metro. Why are you? Here’s why you should switch to another grocer

It’s 2026: why are you still shopping at Loblaws or Metro (or Farm Boy or other expensive grocery stores)?

In 2022 I switched from shopping at my local Metro to shopping at Walmart via Instacart (since I don’t own a car and the Walmarts are far away). In 2023, as prices at Walmart shifted upward on Instacart, I started shopping more and more at Food Basics (essentially much of the same food as Metro).  In 2025 I still mostly shopped at Food Basics because it has the best selection and prices, but I still shop at Walmart for some things, but I also shop at other low cost places like Giant Tiger, Fresh Co and occasionally No Frills.

So how do things stand in 2026? I decided to do a price comparison again on January 17, 2026 of these Canadian staples (selected from this list) to see how cheaply I could get them at the various grocery stores using Instacart. Here’s my results:

Food Food Basics Walmart Metro Loblaws
Chicken breast (1 kg) $15.41 $14.34 $22.00 $23.00
Eggs, dozen, Lg $3.99 $4.13 $3.99 $4.35
Butter, 1 pound $5.89 $6.13 $5.99 $6.55
White Bread $1.99 $2.60 $2.48 $2.75
White Rice, 1 kg $1.62 $1.77 $2.06 $2.33
Bananas, 1kg $1.30 $1.30 $1.74 $1.90
Carrots, 1kg $2.76 $2.27 $3.30 $3.24
Vegetable oil, 1 L $3.00 $4.18 $3.33 $4.03

Things to note: Food Basics and Metro have in store prices on Instacart: Walmart and Loblaws do not, so things may be cheaper if you go to Walmart or Loblaws directly. In each case I tried to find the lowest cost item shown, which meant the cheapest items in Loblaws were no name items. Also to get the best bargain, I went with rice sold in 8Kg bags and oil sold in 3L bottles. For some foods like carrots or chicken, they aren’t sold in 1 kg units, so I converted the unit price for a 2 lb bag of carrots into a 1 kg price.

For at least these staples, Food Basics and Walmart are cheaper than Metro which is cheaper than Loblaws, taking into account the things I noted.

Now this is just 10 staple items. Over a week you buy more than these things. And you buy them over and over again in the course of a year. That’s likely thousands of items, with markups at the higher priced stores being anywhere from pennies to dollars more on each good. I figured I was saving over $1500 a year in 2022 which made up for the Instacart fees for me. You might find something similar in 2026.

Of course you may have many good reasons for shopping at Loblaws or Metro. They both may have items you can’t get anywhere else. You might think their produce or meats better. You might find them more convenient, especially if you don’t have or want to use Instacart. All valid.

But if you do have options to shop at lower cost grocery chains like Food Basics, Walmart, Giant Tiger or Fresh Co, then I recommend you do so. Save your money and/or make your life easier.

P.S. I will also note that it’s good to keep a close eye on the fees of Instacart. For example in 2022 the service charge was 5%. Now it’s 8% with a minimum, making it worthwhile to get bigger orders of at least $50/store. Also the heavy bag fee seems to have gone. Instacart still makes my life easier and saves me money, but YMMV.

It’s winter. Here’s some hot IT links to warm you up (What I find interesting in tech, January, 2026 edition)

I haven’t posted on tech in awhile, which I think is a shame, because I like to see these posts myself, if no one else does. Especially I like to come across them years later and think: oh yeah, I was working on that then. Or think: I forgot about that…I need to take that up again.

So here are dozens of good links on IT that I have found interesting over the last year while I’ve been developing software that uses much of the technology you see highlighted below.

As always, almost all of these links are self explanatory. The ones that aren’t…well…few if anyone but me reads these posts anyway. Just treat it like a collection of potentially good resources that I found worthy of sharing.

AWS: I was doing some work on pushing containers into AWS (which I wrote about, here) so I found these useful
Okta/Authentication: before I pushed the containers into AWS, I had to integrate them into Okta. So I did a ton of research on Okta. Here’s some of the links I found useful:
API Dev: my code has a FastAPI interface, so I also had a few issues on that front I had to figure out first:
Milvus: likewise my code interacts with Milvus, which took a moment to master. Ha! Master! I have barely scratched the surface: 
Langchain: Langchain is at the core of my application, so I have been doing lots of studying on this software:
RAG: RAG is another essential part of my application, specifically agentic RAG. Hence these links:
Other good stuff: I’d be remiss if I didn’t keep these links and share them….some of them are especially good.

Finally here’s the obituary for Lou Gerstner. IBM and my place in it was greatly and positively affected by him. I wonder if there would have even been an IBM anymore if Lou did not come along. May he rest in peace. More on his impact on IBM, here.

On my pieces of advice written for the 99u

Around 2011 I wrote a number of articles for an online publication, the 99u (since folded into http://www.behance.net). The audience for the pieces were creative workers, but really they are for workers and people in general. You can find them here:

I am proud of these pieces, and I think the advice is good advice, still. Take a look when you can.

P.S. For some reason the last one is not on Behance but it is on jane.dallaway.com, hence the last link.

P.S.S. Speaking of taking a look, take a look at the art work of Oscar Ramos Orozco. He illustrated all the articles I did for the 99u, plus many more. In the background of the pieces above you can somewhat see them, but they are also distorted. To see some of the work he did for the 99u, go here. You can see his profile on Behance, here. The image above is also his.

Congestion pricing in 2026, from New York to London to Paris

Congestion pricing isn’t new: London has had it since the beginning of the 21st century, and Singapore had it before that.

What is new is that New York City recently embarked on an exercise in Congestion pricing, and after a year, the results are in. How good are the results? Well, traffic is faster, transit use is up, and quality of life has improved. All in all, it’s looking pretty successful.

It didn’t get much notice in the piece on NYC, but I suspect air pollution also was decreased. After all, Paris has reduced its air pollution by 50%, and part of that was reduction in cars in the city.

If anyone says congestion pricing doesn’t work, show them some of those pieces. For more on congestion pricing in NYC, read this.

How to get fit in 2026: start with your workout


Welcome to the new year. One of the biggest problems with getting fit is knowing what to do. These links can help.

If none of these work for you, I have quite a number of other fitness write-ups, here. Take a look: you might find something you need there!

P.S. Not workout related, but related, here are some sources of information on new gear that caught my attention over the last year:

 

On the Left and Datacenters


People on the right are often criticized about their paranoia about things like vaccines and wifi signals. Left wing critics will say to them: trust the facts, trust the science, ignore the conspiracy theories springing up everywhere. I think that’s good advice.

I think the same advice should apply to the left when it comes to data centers. Two cases I came across this week makes me think something else is happening.

First off, I read this piece about a research group is on the hunt for hidden data centers. This group, epoch.ai, is using satellite imagery and other sources of information to do this. I found this surprising, since there are other organizations (e.g. CBRE) that track data centers throughout the world. It’s pretty common to do. And unlike CIA black sites or nuclear facilities, it’s in the interest of companies building data centers to let the world know they are doing so. Keeping it a secret is not very good for them.

You could argue that tracking data centers makes sense in terms of their power and water usage. That’s a valid claim, though I think there are better ways to do it. I’m also not convinced this is the best way to push back on resource hungry hyperscalar datacenters. I think communities, cities, provinces and states need to understand the impact of such facilities and push back when these facilities harm them, but to act like its a mystery that can only be dealt with using spycraft does a disservice to readers who are concerned about such things.

The second piece I read about was this one which criticized the building of data centers in places where it was too hot. While reading it I was looking to see if the writers would draw the connection between the placement of data centers, undersea internet cables and Internet Exchange Points (IXPs or IXs or NAPs) Alas I did not find any such references.

There are many reasons why data centers are placed where they are, but one of the key ones is proximity to IXs! Even with the blazing high speeds of current Internet connectivity, the closer your data center is to an IX, the better it will be. A data center built miles and miles from IXs (which will be close to Internet cables) makes no sense. The fact that I could read that second piece and not see this pointed out makes me wonder how much the authors understand the technology, and if they don’t understand the tech, what benefit such a piece is, other than as a hit piece on data centers. It’s not a piece people can come to and gather facts and a better understanding of data center placement.

There are lots of ways for the left to criticize data centers, but if they don’t do it with facts and reason but come at it conspiratorially, they’re no better than the right and their paranoia about vaccines and wifi.

On the great Frank Gehry and 10 of his best buildings

The great Canadian-American architect Frank Gehry passed away last month. To mark his passing, the website Yanko Design did a feature highlighting 10 of his iconic buildings, including the Dancing House in Prague (seen above). Well worth checking out.

No doubt you will say: it doesn’t have building X or Y? It’s a testimony to Gehry that he designed so many great buildings that 10 items is too short a list.

Here in Toronto, we are lucky to have his work incorporated into the Art Gallery of Ontario. It’s interesting to see other buildings featured in the list like the Vitra Design Museum and recognize design elements that it shares with the AGO (e.g. the twisting stairs).

The frugality of Bernie Sanders (some brief thoughts on politics and fashion)

Is Bernie Sanders a frugal guy? Does his fashion sense indicate he is a left wing guy? Let’s consider this winter jacket he’s wearing. For Mayor Mamdani’s NYC inauguration, he wore this to give a speech and swear in the mayor. This particular jacket he’s had for 8 years now, which seems pretty economical. Frugal even.

How did I know he has had this coat for 8 years? Because the picture above was taken in 2018 at Bill de Blasio’s swearing-in ceremony. And what happened then? There was an entire controversy about him wearing a supposed $700 jacket. (More on it, here.)

Was there anyone writing about Bernie’s frugality this time? Not that I could find. What I did find was lots of people writing about the wife of the mayor wearing expensive boots to the ceremony!

Some thoughts:

  • These aren’t really valid criticisms of political figures on the left. They’re more fake gotcha moments to try and sour readers on an individual by highlighting some supposed hypocrisy. They don’t hold any more weight than criticizing former Mayor Bloomberg of bein a hypocrite because he would occasionally choose to ride the subway even though  he can easily use a chauffeured limo.
  • Unless a candidate asks people to vote for them because they are poor, there’s no hypocrisy anyway. Most left wing candidates ask people to vote for them because of the actions they are going to do, not for the things they wear. If anything left wing politicians put less stock in fashion choices than they should. Fashion doesn’t matter much, but it does matter a bit.
  • Left wing candidates are policed for their fashion much more than the right. Obama was criticized for not wearing a flag pin and once — once! — wearing a tan suit. Michelle Obama was constantly scrutinized for what clothes she chose to wear to events. In the end it doesn’t matter much to most people and the people who disliked President Obama moved on to other things to criticize him for once he donned the pin and got back in the dark suit lane.
  • Fashion does signal something about an individual. But it’s a very unclear signal for the most part. Bernie’s supposedly expensive jacket might have been picked up at a discount mall for a small fraction of it’s list price. The expensive boots Rama Duwaji wore were apparently borrowed, so essentially they ended up costing nothing, indicating nothing other than she has good taste. People who depend too much on fashion to tell you something about a politician should give that a rest and focus on their words and deeds instead.
  • If you want to know something about the new mayor of NYC and his words and deeds, I recommend you read about his inauguration. And if you want to know more about his wife Rama, you can check out her website or her instagram page.

 

White Out! Here’s a blizzard of good links to end the year (i.e. the December 2025 edition of my not-a-newsletter newsletter)

The second half of 2025 is done, so it’s time to follow up on what I wrote at the end of the first half of the year.  Having just had a good dump of the white stuff (sleet) all over my yard, I’ve decided to stay warm and dry indoors, make myself a big cup of coffee to sip on, and sort out what’s happened since June. Let’s take a look.

As with the previous edition of my not-a-newsletter newsletter, this one is somewhat preoccupied with what is happening in the Trump Administration, though less so. Like the man himself, we are all weary of his presidency, even though he’s been in office a year. (Though technically this is his fifth year as President.)

Other than trying to get a Nobel Peace Prize, he seems to be spending more time focused on his supposed strength: real estate. Of all the things that Trump has done, somehow the fact that he tore down the East Wing of the White House shocked me more than I thought it should.  As some are saying, it is emblematic of his wrecking-ball presidency.  Meanwhile, Trump plows past concerns over East Wing demolition — and envisions an even bigger ballroom than initially planned. The Guardian said the new ballroom is giving off  ‘Dictator-for-life vibes’, although it will be some time before it’s completed, and his plans for the East Wing keep changing. In the meantime, he’s having a go at smaller renos, like the Lincoln Bedroom bathroom 

When he’s not doing damage to the White House, he’s wrecking other parts of Washington, like the Kennedy Center. The Center has seen ticket sales plummeted since Trump takeover. The fact that he’s renamed it the Trump-Kennedy center won’t help matters.

Is he still engaging in extreme corruption? Yes. And not just for himself. The fact this is no longer considered “news” may account for why the media is focused more on other aspects of his Presidency, even though such corruption is a big deal.

As for the new year, could Trump’s tariffs be struck down by SCOTUS? I doubt it. Will he be able to stop the war in the Ukraine? Also doubtful. Will he be able to put the Epstein scandal behind him? Highly unlikely. Maybe he’ll stop using social media in a despicable manner the way he did during the No Kings marches? As if.

Though seemingly unstoppable, it appears 2026 will be full of obstacles for the man, based on the news and these charts examining his year in office. Big supporters like Elise Stefanik and Majorie Taylor Greene are bailing on him. GOP centrists are choosing things like the nuclear option in a fight with Speaker Mike Johnson over Obamacare. It also doesn’t help him that the Job Market is cooling (but not collapsing). More than any other problem he has to deal with though, the key one could simply be his health.

Musk

Like Trump, Elon Musk is someone who we can’t avoid thinking about, even if we want to. His bid for a pay package that could make him the first trillionaire helps to keep him in the news. He still acts like a Big Thinker, throwing out stale ideas like “micropayments”. Let’s not forget his last big idea, though: DOGE. Like the man who came up with it, DOGE was useless in its brief existence. In fact it seems that DOGE ‘doesn’t exist’ despite having months left on its charter.  It’s legacy is that it failed to save any money; all it did was leave refugee children to go hungry.

One of the few ideas that Musk successfully implemented was turning X into a platform that fosters antisemitism. Mind you, Musk is not alone in foster hate: Tucker Carlson’s gave a substantial platform for far-right antisemite Nick Fuentes to share his views. Sadly, we are seeing an ascendence of far right leaders everywhere. For example: In the New Trump Era, the Proud Boys Are ‘Not Apologizing Anymore’ (talkingpointsmemo). Perhaps that’s why Pantone names its Color of the Year for 2026: White. (Technically Cloud Dancer, which led me to revisit this 1979 art work from the Canadian Artists, General Idea: Nazi Milk.)
Mamdani
Among Americans opposing Trump, there’s been one new name that’s been attracting much attention over the last year: Zohran Mamdani, soon to be mayor of New York City. Ironically, he and Trump got along great when he visited the White House, much to the shock of many. For why people — especially young people — voted for him, read this. He seems ready to hit the ground running come January, based on this. Could be an exciting year for NYC.

Canada

Canada has also been opposing Trump. One way they’ve done it is by banning US liquor sales in provincial liquor stores. While the ban has been successful and harmful to some US companies, Canada is left to figuring out what to do with all the stockpiles of US booze.

Ontario also have been running ads in the US using Reagan to complain about the tariffs, leaving Trump angry. I’m not sure Americans ever got it enough, despite it being an effective ad. It didn’t help that the Reagan foundation did Trump a solid on tariffs – at the expense of Reagan’s ideals.

Finally/In other year end news:

The first quarter of the 21st century is over. Time to leap into the second quarter. Mind the gap.

The decline of Starbucks

How is Starbucks doing these days? Well not good, according to some in the media:

And what should you expect in the first quarter of 2026? Key quote:

“Starbucks’ stock has lagged the broader market over the past year, driven by operational, financial, and sentiment-related factors, with the company’s share price declining roughly 10–14% while major indices posted gains. Key factors include persistent labor unrest and historic employee strikes, which have represented the longest and most widespread work stoppages in Starbucks’ history and raised investor concerns about operational disruption, rising wage and scheduling costs, and reputational risks that could depress sales and margins.

At the same time, Starbucks has faced slowing comparable store sales and transaction declines, higher labor and commodity costs, and the financial impact of restructuring initiatives, including store closures and layoffs, under its ‘Back to Starbucks’ turnaround plan, which have weighed on profitability and near-term earnings expectations.”

I thought Starbucks would turn it around as the pandemic declined and companies demanded RTO (return to office). Clearly they have more problems than I thought. It will be interesting to keep an eye on them over the next year.

As for me, I never forgave* them for abandoning so many neighbourhoods like mine during the pandemic, and then returning not as a Third Place but as a Grab n Go place. I now take my money to other coffee shops that did not do that. I suspect many others do too.

(*Forgave is too strong a word, but I said to myself when they closed my locals that I would remember that after the pandemic. By the way, the image above is of one of the Starbucks in my area.)

You might think you want band steering on your home router, but maybe you don’t :)

If you are having problems with your WiFi network at home, it could be a band steering problem. Let me explain.

If you have home Internet and a relatively new router which supports WiFi, it may support communications on two bands: 2.4 Ghz and 5Ghz. It’s also likely that you have band steering. What this means is that you only have one connection point in your network, and the router will steer your device to the appropriate band depending on which one can deliver better performance at a given moment. This piece explains it well.

Normally this is a good thing. But I am here to say that some times it is not.

One of those times happened on my router last week. I was getting terrible response time with my WiFi. I was pinging 8.8.8.8 and instead of it taking less than 30ms it was taking 300ms or 500ms or even greater than 1000ms. I tried everything to reset the router, but nothing worked. I thought maybe some of the older devices were causing the problem on my network somehow. I wanted to put them on the 2.4Ghz band and have my computer joing the 5GHz band, which is something that worked for me years ago. Alas, my router has band steering, so we were all getting lumped together.

Fortunately it was easy to go into the router settings and disable band steering. I did that and created a default connection point and a 5GHz connection point. When I did that, I discovered an interesting thing.

It turns out the 2.4Ghz band range was fine! When I joined it, my response times for my pings were down in the 20-25ms range. However, when I joined the 5GHz band, my response times were terrible again! I don’t know why that is. I suspect there’s a problem with my router. But as some folks like to say, that is a January problem.

So the next time you are having WiFi problems and you have the choice of accessing your network through different bands, consider turning off band steering and see if that helps isolate the problem.

Happy Christmas Eve, 2025


Christmas Eve has always been the best day of the Holidays for me. It’s the time with great anticipation. It’s also when most of the effort of the season has been done, and what remains is Comfort and Joy.

Here’s some Christmas lists – new and old – that you might find useful this year or in future years:

  1. The LCBO’s best bottles of sparkling wine and champagne from Toronto Life
  2. The Best Made-In-Canada Gift Ideas For 2025 from Chatelaine
  3. Tips for decorating — and staying organized — for the holidays from The Washington Post
  4. Chic Outdoor Christmas Decorating Ideas from | Apartment Therapy
  5. 5 Christmas Garland Ideas for Every Room in Your Home also from Apartment Therapy
  6. Nerdy Hanukkah Card from MakeItHackin on Tindie.

Happy Holidays, and Merry Christmas.

 

You might think you want to run your code in a Docker container on AWS ECS vs having it run on a AWS EC2 server, but you are wrong :)

You might think you want to run your code in a Docker container vs having it run on a server. Next you might think you want to run it on AWS’s ECS (Elastic Container Service) because you assume that is easier than going full blown Kubernetes.  Let me try and dissuade you from this and try and persuade you to stick with a VM running on EC2 (Elastic Compute Cloud).

Last week I wanted to take some code I had and run it on ECS. But before I did that, I did an experiment: I created a dockerfile that was just a simple web site that consisted of a Python Flask app and some HTML files. There was a wrinkle: I wanted the HTML files to sit on a file system outside the container. It’s easy enough to do with Docker and a command like this on an EC2 instance:

docker run -p 8000:8000 –rm  -v /home/bernie/mydata:/app/mydata:ro website:latest

My container listens on port 8000 and any files in my /home/bernie/mydata directory are accessible within the container at /app/mydata

For a container running in ECS, it is an entirely different beast. You have to:

  1. Create a container registry in Amazon’s ECR (Elastic Container Registry).
  2. Make a Docker image and upload it to your container registry.
  3. Create a cluster in ECS.
  4. Create a Task Definition in ECS.
  5. Create an ECS Service, including setting up an ALB (optional), configuring security groups and configuring the VPC associated with the cluster.
  6. Set up an EFS (Elastic File System) to store the HTML.
  7. Set up an EC2 server to create / upload the HTML to. You cannot directly access the EFS you create to the container: you have to mount it to your EC2 server, load the files to the EFS, and then attach it to the cluster/container.
  8. Once you do all that, THEN you can access the simple Flask web site and the web pages you created.

MOST IMPORTANTLY, you have to DEBUG DEBUG DEBUG all through this process. Did you forget to add port 2049 to your security group rules? Fail. Did you forget to set your IAM settings correctly? Fail. Are your subnets set up wrong? Fail. On and on and on.

If you insist on doing it, then I have some useful links below. But if you can get away with a simple micro EC2 server with a simple software firewall in front of it, I highly recommend you go with this approach.

Creating a simple EC2 server? Easy. Having Docker running your container on your server? Easy. Getting your first container running on ECS? Not easy.

P.S. If you still think this is a good idea, then I recommend this: Deploying a Dockerized Web App on AWS Using ECS and Fargate: A Step-by-Step Guide. Deploy a dead simple container first just to work out all the issues you will have with ECS and any other AWS services, and once you have the ability to do that, proceed with more complex containers.

If you want to do it using the AWS CLI, I recommend this repo: Developing and Deploying a Basic Web Application on Amazon ECS Using Fargate.

More good pieces to read: here

P.S.S. Another consideration: while a container running in ECS isn’t too expensive, it’s more expensive than a micro EC2 server. Which is also more expensive than serverless computing using AWS Lambda (although Lambda is more complicated than an EC2 server).

P.S.S.S. I went through this process because the client I was working on had a mandate to deploy code this way. Hence this exercise. The client has good reasons to do things this way: most enterprise clients do. If you are not an enterprise, make your life easier.

Blogging in the AI era

Traffic surge due to AI

Last week I wrote about AI and food blogging. After I wrote that, I got to thinking about my own blog. Above you see my daily traffic, with a big spike happening at the beginning of December. I looked at what visitors were looking at on my blog then: I could see someone looking at a bunch of links beginning with “A”, then beginning with “B”, then “C”, etc. It’s not something a user would do, but it is something a bot would do. I am assuming it is somehow AI related.

It made me wonder why I am still blogging if people are just going to bypass my blog and read it from AI. In some ways I don’t mind: if someone finds my content useful via AI versus my own blog, then it doesn’t matter to me. In other ways I do mind: I keep up a weekly discipline of posting at least 2x/week as much to keep the blog a living thing, but if AI is going to kill off my traffic, then what is the point of maintaining this discipline? Likewise, if I am getting traffic due to AI, then maybe it doesn’t matter to keep posting, since my site will still be getting traffic. It’s a pickle, I think.

I think I will still post because I choose and want to post, but the days of posting to help maintain interest in the blog may be over. Something to consider as I go into the new year.

1984: the year computing changed

The above photo is from a 1984 photo of a worker at IBM in Böblingen. It’s a fairly typical image from then (I know, I was there at that time, though in a different place.) Most of the computer equipment in that room is associated with the mainframe that he’s working on: tape drives (back right), storage (front right), network (front left), consoles (middle). The one exception is the IBM PC behind him.

1984 was the year Apple introduces Macintosh. 1984 is the start of the personal computer revolution that would sweep all that mainframe technology aside. And while the Mac was important in the PC revolution, the IBM PC was just as important.

P.S. I thought of that when I heard the IBM campus at Böblingen is being phased out.

P.S.S. IKEA is currently selling a version of that table, here (the BAGGBODA).

 

On the rise of AI, and the decline of food blogging (and potentially all user generated content)

Food bloggers are seeing a drop in traffic as home cooks turn to AI, according to Bloomberg:

Across the internet, writers say their vetted recipes are hidden by the flood. Pinterest feeds are stuffed with AI-generated images of food that the attached instructions won’t achieve; Google’s AI Overviews surface error-filled cooking steps that siphon away clicks from professionals. Meanwhile, Facebook content farms use AI-generated images of supposedly delicious but impossible dishes to the top of people’s feeds, in an attempt to turn any clicks into ad revenue.

How bad can it be? To give just one example (from Fortune):

For Carrie Forrest, who runs Clean Eating Kitchen, AI has been devastating: 80% of her traffic — and her revenue — has disappeared in two years. Although the views started dropping when OpenAI’s ChatGPT was released, it wasn’t until Google launched AI Mode in search that her traffic collapsed, she said. Since then, she’s gone from employing about ten people to letting everyone go. “I’m going to have to find something else to do.”

My thoughts on this:

  • There already was several problems in the food blogging area, like too many people trying to make a go of it by writing about cooking. Not to mention that there was a problem with plagiarism within the world of food blogging. Now there is a new plagiarism problem, with AI essentially copying recipes and putting them forward as another source of cooking information.
  • AI plagiarism is not limited to food blogging. I’ve seen the same problem within the IT domain, with vapid sites posting “solutions” to technical problems that are a waste of time.
  • For people and organization that generate content hoping to get people to read it, AI will be squeezing people from both sides. Human generated content with be squeezed out by AI generated content. And with things like Google Zero, search engines will use AI to summarize that content, so that the person searching for information will not go to the next step of clicking through a link to get to another site.
  • Like the person who runs the site Clean Eating Kitchen, I suspect many smaller sites will dry up as a source of content. I think you will then have these aggregators (either Google or large AI generated content sites) plus other quality sites that can still compete, be they cooking.nytimes.com or marthastewart.com or even specialized sites like alisoneroman.com. This is not unlike Travel or Hotel sites that can still maintain a presence even though search engines like Google will scrape prices and present them.
  • Eventually, like all things in IT, there will be a continuing swing towards centralization, until the classic problems of centralization occurs. Then you will have niche sites breaking out again and things will decentralize. Until then, the rise of AI both on the front end (Google) and the back end (AI generated sites) will force out these midtier sites like Clean Eating Kitchen.
  • I am not immune to this problem. I see this site being scraped on the regular, presumably by AI. The good news for me is I write this site as much for me as for anyone else. I am happy if I can post things that help people. However, if some AI site presents the information and it helps people, I don’t care that they didn’t get it directly from me. I still benefit from writing these posts. If I were paying my bills with this site, it would be a problem. But this site is just a hobby for me.

(Image from Pixabay, which appears to be AI generated. Pixabay used to be a good source of imagery, but it has been overrun by AI generated images.  I found this by looking for “ai generated cooking” images.)

Need a book to read? A gift to give? Let the New York Times and I help you

The New York Times recently published a helpful list of 100 books to read (gift link, here). One thing great about that list is that it is broken down into subcategories. Need a book you can read in a day? Perhaps you need something to make you laugh. Or how about a really good memoir or biography? Whatever type of book you need, the Times has guidance for you in one of the subcategories. So instead of staring down a long list of 100 books, you can narrow in on just what suits your reading desires right now.

Books also make great gifts, for those of you looking to get some for the Holidays or some special event. Go check out the list, either for yourself or others. You’ll be glad you did.

Mark Zuckerberg says Social Media is dead (and other thoughts on social media for Dec 2025)

Last year I wrote that Bluesky is making social media fun again, and that was true. It’s still somewhat true: of the remaining platforms, Bluesky and Instagram are still getting steady use from me in 2025.

That might change in 2026 if Instagram starts floods itself with AI content, though. Already I am seeing more and more people posting terrible reels that either are explicitly AI generated or probably AI generated (I’m looking at you, cute animal videos.) Once AI video slop takes over that platform, my Instagram days will be over.

Despite my use, is it true in 2025 that social media really dead for most othe people? In a recent court trial, Mark Zuckerberg of all people says so. It was in his interest to say so in the trial, though. I still believe that the 1-9-90 rule continues to apply to social media: 1% of users create new content, 9% of users will engage with that content, and 90% just consume content. The difference now is the content people consume is not so much from their friends but from organizations and influencers.

As for where they consume it, according to Pew Research Center, most people in 2025 get their content on sites like Youtube, Facebook, Instagram and TikTok. Those platforms are also the ones that have seen growth since 2021, according to Axios. Axios adds WhatsApp and Reddit to those sites seeing growth.

The one social media platform seeing decline since 2021 is X (Twitter).  It doesn’t help that X is the go to site for antisemitism, according to CNN. (More on that, here.) And that’s just one of many reasons for people to flee it for twitter-like sites such as Threads and Bluesky.

As for Threads and Bluesky and Truth Social, they just don’t have the juice that Twitter had in it’s heyday. Maybe Mark Z is right: maybe social media as we knew it is dead. Maybe now those social media platforms are just full of people watching TikToks and Reels and Youtube videos, when they aren’t streaming music and video. And maybe that’s not a bad thing.

P.S. For more detailed analysis, I highly recommend the Pew report I linked to.

On Gilbert and George, their Southbank Centre show, and conservative artists

Untitled

Following up after the Richter, I also got to head over to London and see Gilbert and George: 21st Century Pictures at the Southbank Centre. You can get a sense of the show by reading what the Centre says here, but you really have to see their work in person to appreciate both the scale of it and the detail of it. (No matter if I got close or far back from a work, I was always losing something in my photographs, as you can see above.) So see it before it leaves on January 11th.

One thing the exhibit had me look into was the prominence of conservative or right wing artists. These two definitely qualify for that label. But other than someone like Clint Eastwood, I couldn’t think of other right wing artists who were influential.

It seems I just wasn’t paying attention. Reactionary artists are everywhere, based on this list here and here. Of all of them, though, these two English sculpturers (as they refer to themselves) are my favorite.

For more on Gilbert and George, I’ve found a number of other good pieces here and here and here and here. They’re living legends. I’m glad I got to see a major portion of their work at this Southbank Centre exhibit.

On Gerhard Richter at the Fondation Louis Vuitton

Untitled

I was blessed recently not only to be in Paris, but to be there when the Fondation Louis Vuitton had on display a major exhibit of the works of Gerhard Richter. To say “major” is an understatement: it had dozens and dozens of work from every era he has been making art, from the 1960s until now. As a big fan of Richter, I could hardly believe the paintings that they had on display, from Tisch, 1962 to Birkenau, 2014 (seen above). It was incredible.

If you can get to Paris by 02.03.2026, you owe it to yourself to go. You can read more about the exhibit, here. Of course the building itself is fantastic, thanks for Frank Gehry:

Untitled

There’s so much great art to see in Paris, but I’d recommend you check out what the good folks at Louis Vuitton have on display. Before this they had a big David Hockney show too. Have some champagne and cake in the dining area after the show: you’ll be glad you did.

P.S. Here’s a good review of the show with many more images from it too.

Let gen AI and the Internet be your travel guide in this day and age

In the old days of travel (the 1980s), travellers would depend on printed guides from Fodor’s or Lonely Planet to be your travel guide. They were good, but very general.

Now with tools like ChatGPT and CoPilot, you can make your own highly specific guides just by asking them for it. For example, if you were travelling to Paris, you can ask it for walking tours of neighbourhoods like the Marais and prompt it to give you cafes to see or places to shop or museums to visit along your walk.  (Sure you can also Google that information, but by starting with gen AI, you can get it in a format you want.)

Likewise, you may think: what do I say to people once I get there? For that, I asked CoPilot:

Hello. You are a French to English translator. Please provide me with some common French phrases for that are useful for 1) when you arrive at the airport 2) when you come to your hotel 3) at a cafe 4) at a restaurant 5) at a museum 6) in a cab or uber

and it came back with such phrases as:

1. At the Airport

Bonjour, où est la sortie ? — Hello, where is the exit?
Où sont les taxis ? — Where are the taxis?
Où est la station de métro ? — Where is the subway station?
Je cherche la porte d’embarquement. — I’m looking for the boarding gate.
Où est la récupération des bagages ? — Where is baggage claim?
…
3. At a Café

Un café, s’il vous plaît. — A coffee, please.
Je voudrais un croissant. — I’d like a croissant.
Est-ce que je peux avoir l’addition ? — Can I have the bill?
C’est délicieux ! — It’s delicious!
Est-ce que vous avez du lait végétal ? — Do you have plant-based milk?

….

These were good, so I followed up with

Can you expand each list from 5 items to 10? Merci!

and it replied with this and more.

Parfait, Bernie ! Here’s an expanded list of 10 useful French phrases for each of the six travel situations, with their English translations:
…
4. At a Restaurant

Une table pour deux, s’il vous plaît. — A table for two, please.
La carte, s’il vous plaît. — The menu, please.
Je suis allergique à… — I’m allergic to…
Quel est le plat du jour ? — What is the dish of the day?
L’addition, s’il vous plaît. — The bill, please.
Je voudrais réserver une table. — I’d like to book a table.
Puis-je voir la carte des vins ? — May I see the wine list?
Est-ce que le service est compris ? — Is the tip included?
Je prends ce plat. — I’ll have this dish.
C’était excellent, merci ! — It was excellent, thank you!

Those are just two examples of how you can gen AI to help you with your travel plans. But really, the sky is the limit on what advice you can get from them.

N.B. Do I recommend you double check these guides outside of gen AI to make sure they are correct? Bien sûr! I found the answers we got back were good, but you’d be wise to use Google/Bing/Duck Duck Go  directly to confirm the advice you got.