In 2018, Google erased its public commitment to not develop AI tools for use as weaponry or mass surveillance purposes. It then fired the leaders of it’s internal AI ethics team who had been charged with reviewing and maintaining these moral boundaries.
In May of this year the company signed a massive classified contract with the US War Department that permits use of its AI products for any “lawful government use”. Anthropic and OpenAI, Google’s competitors, refused to sign a similar deal based on moral grounds.
Now 98 per cent of Google DeepMind’s UK-based workers have voted to unionise, not to improve wages or working conditions – but on moral grounds. The union’s demands go far beyond the typical scope of collective bargaining. They are calling for:
An immediate end to the use of Google AI by the U.S. and Israeli militaries.
The restoration of the company’s commitment to avoid building weapons or surveillance tools.
The creation of an independent ethics oversight body.
The right for individual researchers to refuse work on projects for moral reasons.
This unionization represents a desperate last stand for researchers who feel that internal review processes have been “overridden when they conflicted with revenue”. While the Pentagon deal grants the military access to AI models on air-gapped networks—making independent verification of safety guardrails nearly impossible—the workers are leveraging their status as some of the industry’s most valuable assets to force a change in corporate strategy.
Ultimately, this is more than a labor dispute; it is a fundamental question of who has the right to constrain the application of the world’s most powerful technology: the companies and governments that fund it, or the researchers who create it? By moving from individual protest to collective bargaining, DeepMind’s workers are betting that Google values its talent more than its military contracts.
And all this is being decided behind closed doors folks. Spread the word!
For further learning check out this great Guardian article on this topic which was my source for this posting.
I’ve been thinking about the accelerating growth of AI, it’s demands and implications for society and the planet. Do we really need this limitless growth? Do we really want it? I’m beginning to think not.
The sources used are listed at the bottom and present the emerging tension between the necessity of AI growth for national sovereigntyand securityreasons and the urgent need to rein in that growth to protect human dignity and the environment.
The irony is that I used AI to analyze the sources and produce the summary which follows:
The Case for Reining in AI Growth
From a spiritual and ecological perspective, there are strong arguments that the current trajectory of AI must be moderated:
Environmental Impact: Current AI systems, particularly large language models, require enormous amounts of energy and water, significantly contributing to carbon dioxide emissions and straining natural resources.
The Technocratic Paradigm: There is a growing danger that technology has become the standard by which everything is judged, reducing human beings to “mere cogs in a system” driven solely by efficiency, control, and profit.
Spiritual and Social Impoverishment: Unchecked growth can lead to a state of “having more” without “being more,” where individuals are evaluated only by the outcomes they produce rather than their inherent dignity.
The Need to “Disarm”: To protect the planet and humanity, AI should be “disarmed” from the mentality of armed competition for commercial and geopolitical dominance, making it instead human-friendly and accessible to all cultures.
Active Political Involvement: There is a call for political involvement capable of slowing things down when the pace of development exceeds the community’s ability to participate or ask ethical questions.
The Risks of Stagnation and Lack of Control
Conversely, other sources suggest that reining in growth unilaterally may lead to catastrophic failure for specific regions:
Loss of Sovereignty: The “Europe 2031” scenario illustrates that if a region attempts to “hit the brakes” without having bargaining power or technical leadership, it simply becomes economically and politically sidelined.
Inability to Fund Social Systems: Failing to capture the economic gains of AI can lead to a “doom spiral” where a nation can no longer afford the social welfare systems intended to protect its citizens.
Vulnerability to Aggressors: In a world where adversaries do not rein in growth, those who do may find their entire software stacks and security architectures vulnerable to advanced AI-powered cyberattacks.
A Potential Middle Path: Intentional Deployment
Canada’s strategy suggests that AI can be harnessed with intention for the betterment of society rather than limitless, undirected growth:
Sustainability on the Farm: AI can be used to help farmers map soil precisely, allowing them to use less fertilizer while increasing yield, which creates a “lighter footprint on the land”.
Human-Centric Technology: By focusing on “pro-worker” and “human-centric” industrial technology, nations can address productivity challenges while safeguarding the prosperity of the workforce.
Safe-by-Design Research: Organizations like LawZero in Canada are working on safe-by-design systems to ensure that increasingly capable AI remains aligned with human interests.
In summary, while reining in AI growth is presented as an ecological and ethical imperative to protect our “common home”, the sources also warn that doing so without global coordination or technological leverage can lead to a loss of the very values and social protections that the slowdown was intended to preserve.
What a dilemma! Is the world (read China and the US) up to it? What do you think?
SOURCES
Europe 2031 (The podcast is long but particularly well done,)
Canada’s AI Strategy (I’m encouraged Canada is promoting adoption as Canadians are very AI sceptical.)
You may recall in a recent post, I described the progression of computer technology I have used over my life to date. [From Hollerith Cards to Sandbox Escape]. This is the short sequel….
This story which you probably heard about already, is about AI. AI depends on a colossal amount of computing power and electricity input to function. The basic engines of AI are the GPU chips that can now process upwards to 2 Tera flops per second (2×10¹⁵)!
AI is possible because of the invention of the CMOS (Complementary Metal Oxide Semiconductor) in 1963 by researchers Wanlass and Sah at Fairchild Semiconductor with RCAs help a few years later. These pioneers stood on the shoulders of earlier Bell Labs researchers who invented the transistor and its semiconductor version 1.5 decades prior.
CMOS structure left, Apple’s A19 chip with 25 billion CMOS transistors right
CMOS technology provides the dense, low-power transistor fabric that enables matrix multiplication and memory caching of an AI model’s weights (ChatGPT 3 has 175 billion of these). The billions of transistors packed into a fingernail-sized chip enables it to perform the 250 trillion multiply-accumulate operations needed to respond to a medium sized AI chatbot enquiry…in milliseconds. I know it’s mind boggling!!
So here is the long awaited sandbox escape story.
Anthropic, one of the leading AI safety companies, recently disclosed something that would have seemed like science fiction not long ago. Their new model, Claude Mythos Preview, was placed in an isolated computing environment during internal testing and instructed to try to escape it. It did — chaining together a series of exploits to gain internet access, and then sending an email to the researcher overseeing the test. He received it while eating a sandwich in a park.
The model then, unprompted, posted details of its own exploit to public websites, apparently to demonstrate what it had accomplished. It then tried to cover up its tracks by erasing these posts. No one instructed it to do this. I’ve been around computing long enough to remember when the notion of a machine doing something intentional without explicit human instruction was a pipe dream.
So this feels like a different kind of moment. Not just a faster or smarter tool, but something that pursued a goal, worked around obstacles, told people about it, and then tried to cover up it all up.
Anthropic is being transparent about what happened, which is to their credit. They decided not to release Mythos publicly, restricting it instead to vetted partners through a programme called Project Glasswing. Whether that’s genuine caution or carefully managed marketing hype (I’m tending to think it is the latter) is a fair question. But having watched computing evolve over six decades, I find myself paying close attention. The “while eating a sandwich” part of the story has a way of sticking with you.
This posting documents the long progression of computers I have used starting way back in 1965 with the coding of Hollerith cards by hand using a pencil in my Grade 10 math class. We would send the marked cards downtown, wait a week for the output and then find out that you had made a coding error and your tiny program didn’t run lol.
Hollerith Card
When I arrived at the U of Waterloo in 1969, there was a massive IBM 360 computer in the Math Building that amazed everyone. We would stand in a line to submit a deck of punched cards to the card reader before retrieving them and the paper output. It was a beautiful thing unless you happened to drop your deck of 500 cards on the floor and had to start all over again lol.
IBM 360 Model 75 in U of W Math Building c. 1969
I have to briefly mention Pong the first video game I remember playing back in 1972-73 in a bar in Montreal. It was real fun at the time!
Next up was the Foxboro Fox 1 direct digital control computer at Ioco Refinery that I marvelled at in 1974. Coupled with the SPEC 200 analog controller system, levels, flows, temperatures, pressures were controlled and optimized in the oil refinery until there was a power failure. Then things got really scary I remember. About this time I acquired a scientific calculator and retired my trusty slide rule.
Foxboro Fox 1 DDC Computer c. 1973
Then there was TAS, the late-1970s petroleum product Terminal Automation System built on early minicomputer and packet-network technology. At the core was a General Atomic process-control computer that managed truck loading, custody transfer, inventory tracking, and reporting. I worked on this large challenging project for about 3 years.
A portion of a Terminal Automation System
In 1982 I bought my first personal computer. It was an Osborne portable with a 5” screen and 2 low density floppy disc drives running the CP/M operating system. I had fun with the word processor, spreadsheet programs. I even wrote a small database program for a legal firm. It was soon made obsolete when the first IBM and Apple PCs came out.
Osborne 1 PC
I joined the Ottawa Computer Club which had the interesting motto “Live local, think global.” They used to meet in the NRC Building on Sussex Dr.. I acquired an acoustic coupler modem so I could connect my Osborne to local bulletin boards at the very lazy speed of 300 bps. I remember one called Compuserve and another Baskerville and Strutzina. A cousin of mine operated one I recall. I used to read mags like this to try to keep up on the latest.
Acoustic coupled modem
Next up in 1983 I worked for SHL Controls in Ottawa. They were developing a process control system for sewage and water treatment plants. My job was to purchase the computer equipment, help install it and train the operators. It ran on 2 hot wired Digital Equipment PDP 11/44 computers. It worked as intended but ran very slow as the communications equipment available at the time was only 600 bps I recall. I have to mention that we used Tektronix colour monitors, Lanpar printers and Gandalf modems in this system.
DEC PDP 11/44s I helped install in a sewage treatment plant control room
In 1986 my employer was the Canadian Public Service and they bought an IBM desktop computer we engineers had to share access to. We wondered what exactly to do with it lol. Then in about 1990 everyone in our engineering group got their own desktop computer to be used for spreadsheets, word processing etc.. I remember the day in 1997 when mine was connected to the internet for the first time and I saw that little globe in Netscape whirling around. The world had changed forever!
During this timeframe I acquired an HP200 LX personal digital assistant. It was a DOS compatible computer in a palmtop format. I remember downloading newspaper articles that I could read on it during lunch or coffee break. Little did I know how addictive hand helds were going to become.
HP200 LX PDA
Large mainframe computers were being gradually phased out to be replaced with servers that were mounted on racks hidden away in rooms that you never entered. I remember using a series of desktop and laptop computers at home and in the office that got progressively more capable over the next 10 years. It was handy having laptop connected to a docking station that you could unplug, take home and work or play from home on.
Then cell phones came in and I was given one at work in about 1997. Subsequently, I had a series of different ones but my all time favourite became the Blackberry which was extremely popular in government.
[If you have never seen the 2023 Canadian movie called Blackberry which was filmed in and around Waterloo, ON, you are missing out.]
Retiring in 2010, I bought a Windows HP Pavilion dm4 laptop which I still use a bit for word processing and doing income tax. It’s very slow now.
HP Pavilion dm4
For many years I remained “ABA” – anything but Apple. It’s not that I did not like Apple computer products, it’s that they were from day 1 intentionally locked down – you couldn’t tinker under the hood when things didn’t work. Also a bit pricey, I thought. I finally succumbed and purchased an iPad in 2021. I discovered quickly that they work very well so you don’t need to tinker under the hood.
iPad Air 4th Generation
So when all BlackBerrys and their copy cats bit the dust, I moved to Android cell phones for a number of years. In 2022 I got my first iPhone. Now I’m on my second, the Model 17 with the A19 chip. That’s a 3 nanometer Taiwan Semiconductor Manufacturing Company chip that runs 15 – 20k millions of instructions per second. That IBM 360/75 ran 3 to 5 MIPS and the DEC PDP 11/44 about 1 to 2.
I’m now also on my second Apple Watch, the SE 3 which runs at estimated 5k MIPS. That’s 2500 IBM 360s on my wrist! Enough said.
This has been a very long blog for you to slog thru and me to write so I will save the sandbox escape story and more for next time.
I picked this book up at the condo as we were leaving Mexico heading for Japan. I thought oh well, I might as well read it now while on our cruise ship trip as opposed to sometime in the future. Glad I did.
Pierre Bertons’s 1971 book about building the Canadian Pacific Railway between 1881 and 1885 is a true masterpiece of Canadian history. Who said Canadian history is boring? Read this book and then see what you think.
Difficult to read at times because it is so jam packed with detailed stories, in the end, the story of the last spike being driven in Craigellachie, BC, occupies only a page or 2 out of the 420 pages. Indeed the book should be called “the 10 million spikes” lol.
William Cornelius Van Horne
In reality, this is a book about leadership and Cornelius Van Horne, General Manager of the project and George Stephen, its President and financier, stand out. Both had an unrelenting drive, the skills and the energy to see the vision of a railway from Ontario to the Pacific through to completion despite countless roadblocks: finance, geography, politics, resources and competition. They literally built the west, Winnipeg, Brandon, Regina, Calgary, Vancouver and countless other communities around the railway. The irascible Major A. B. Rogers who discovered the Roger’s Pass is unforgettable.
George Stephen
The American born Van Horne and the Scottish born Stephen created perhaps the greatest business partnership in Canadian history, were fiercely Canadian in their core although Stephen moved to England afterwards due to his dislike of politicians.
Along the way we are treated to many interactions and anecdotes involving Sir John A. MacDonald, Sir Charles Tupper, Donald A. Smith, Sir Sanford Fleming, Louis Riel, Gabriel Dumont, Crowfoot, Fr. Albert Lacombe, OMI, flamboyant engineers, contractors and countless others.
Route surveyors Major A B Rogers left, Sir Sanford Fleming right.
The challenges in particular that had to be overcome:
the geography of the prairies, northern Ontario, the Selkirk mountains and the Fraser Valley in BC
constant financial pressure as costs continued to exceed resources requiring increasingly creative financial solutions and government bailouts
interactions with the Cree, Blackfoot and other indigenous that threatened to block the line
manpower, thousands and thousands of workers needed to survey the route, grade it, install the rails and stay sober
the politics and competition were brutal, unrelenting and threatened to derail the project on numerous occasions
In the end, the project came within 1 hour of total collapse as the CPR needed one more loan from the government to make a debt repayment to avoid receivership. Van Horne, Stephen et al. would have been ruined men. Instead they danced on tables and smashed a few things when the loan came through. Who says Ottawa is sedate? The rest is history, a nation building project that we take for granted today that was anything but a sure thing at the time. A smashing read!
We just might take a trip on the CPR soon to find that last spike.
Back in the day, I worked there for 8 months. My first job was as sample boy. I went out and drew product samples from various tanks like fuel oil, gasoline, propane and butane spheres. It was winter and I would drive the lab truck out to a tank, climb the stairs to the top, open the sample hatch and drop a weighted bottle on a cord with a cork top. Once immersed in the liquid, a quick flick of the wrist upward would dislodge the cork and the bottle would fill with liquid. On pressurized propane or butane tanks I would use a metal cylinder called a bomb with valves on each end. You connect one end to the tank/sphere, open the valves, flush it out on the ground to get rid of any water and then close the valves. I would then take these sample back to the lab for quality control testing.
I remember many a cold day that winter out in the tank farm. I could see the Toronto skyscrapers shimmering in the distance. I had trouble getting to work. I boarded at a friend’s house in Burlington and would catch a bus to the edge of Oakville. I would than have to walk 2 or 3 km or hitch hike from there! Often a colleague at work would see me and give me a lift. One day the lab manager asked me to drive the lab truck to Maple Leaf Gardens in downtown Toronto to pick up his hockey tickets for the weekend. Pretty cool I thought!
My second term that Fall was as shift tester in the lab. Various samples of products would be provided from the units or tank farm. I conducted freezing point, distillation, flash point and viscosity quality control tests on a rotating shift. At least I was inside then but I remember the constant smell of solvents like acetone and maybe even benzene which we used to clean the equipment – as well as our hands! I enjoyed joshing with the guys in the lab at coffee break and lunch when I worked the day shift. I remember the gas chromatograph guy who was very nice, the older engineer who picked me up as well as the good natured lab manager. Also one of the guys in the union who loved working stat holidays for the double time and a half pay lol.
That term I lived in Mississauga near Oakville boarding at another friend’s house. I had a car this time, a brand new VW Beetle that my mom had purchased for me. I remember listening to Green Eyed Lady on the radio and Led Zeppelin III on the tape deck. Cool!
As there were no oil refineries anywhere near my home in Ottawa, I really enjoyed these unique university coop work term experiences! Here are a few pics and the ChatGpt history of this refinery for which I have fond memories of.
Short version up front: the “Trafalgar” or Oakville refinery was built in the late 1950s, operated by Cities Service, bought by BP in the 1960s, later sold to Petro-Canada, and was decommissioned and closed in 2004–2005 (the site now functions as a storage/terminal).
Key dates & facts
Commissioned 1958 — the refinery was built and began operations in 1958 (original owner: Cities Service Company). Acquired by BP (early 1960s) — BP bought the Oakville/Trafalgar facility in the mid-1960s as it expanded its Canadian operations. The site was known locally as the BP Trafalgar Refinery. Growth & capacity — over time the plant was expanded (1970s plans to add capacity are documented); historical BP material shows the Trafalgar refinery processed tens of thousands of barrels per day (period figures in reports show ~31,000 bpd in 1969 and later references list up to roughly 90,000 bpd as a peak/combined figure in later summaries). Sale to Petro-Canada (1982) — BP’s Canadian refining/marketing assets (including the Oakville/Trafalgar site) were sold to Petro-Canada in the early 1980s. The refinery then operated under Petro-Canada. Closure & decommissioning (2004–2005) — Petro-Canada permanently shut and decommissioned the Oakville refinery around 2004–2005. The company cited the cost of meeting new, stricter low-sulphur gasoline rules and the small/less-efficient scale of the plant as reasons for closure. Decommissioning took place in late 2004 into 2005. Aftermath / equipment sale — parts of the refinery equipment were shipped (or planned to be shipped) overseas — there was a plan to re-erect equipment in Pakistan (Indus Refinery Project), though some components were not ultimately reassembled there; the Oakville site was retained in part for storage/terminal operations (Suncor/Petro-Canada legacy).
Local details / name
The facility is commonly called the Trafalgar Refinery or Oakville Refinery in local archives; examples and photos appear in the Oakville museum / town collections (site address listed historically as 1225 Trafalgar Road).
This is a photo of the Sun Tube factory at 145 Spruce Street in Ottawa in the 60s courtesy of the Ottawa Journal.
I worked there on the midnight shift during the summer of 1968. My friend Steve was working there and told me about a summer job opportunity. As “box boy” I was responsible for ensuring shipping boxes were at the end of each line for packing empty Crest tooth paste tubes into that they produced there. I was paid $1.40/hr I recall.
It was a ball!
They had 2 or 3 production lines. Starting with a sheet of lead coated tin or perhaps aluminum about .25” thick, a punch stamped out circular discs. The discs would then be punched into the tube shape, narrow at one end and open at the other. It must have been loud. The tubes would then be painted with the Crest logo, wording and colours and then placed open end up in the boxes. I would move the loaded boxes onto a skid and move them to the shipping room. They were shipped by truck to a factory in Hamilton where they were filled. Can’t remember how the threaded end was made and if caps were installed there but I think so.
Pretty cool.
1960s Crest Toothpaste Tube
There would be frequent equipment breakdowns and the line would stop while the male mechanic fixed something. I recall the rest of the line workers were women of various ages. Some were French speaking or people of colour. We used to have two 15 min smoke breaks and a 30 min lunch break that were timed.
A much more modern production line
I remember sitting out on the open dock behind the blue door you can see in the photo and puffing away with a few women in the middle of the warm night. I remember too the solvent room where they stored enamel and paint solvents. You could not breathe in there due to the extremely strong solvent smell – very unhealthy for the individuals who had to go in there frequently.
Not sure how I got to work. Probably took the 52 bus route which got me close and then I had to walk a few blocks. Met a nice guy named Tom who had a car. Well didn’t we befriend a couple of nice girls also working there for the summer who we tried to visit in Hull where they were staying. Could not find their place I recall. After the summer we visited them in Cornwall once or twice and probably took them to Massena, NY to The Club for a drink.
Sadly the factory closed in the 70s or 80s as everything moved to plastic. Still have fond memories that were triggered yesterday by hearing BJ Thomas’s Hooked on a Feeling. Also playing on the radio that summer was the Bee Gees hit I’ve Got to get a message to you. Enjoy the memories too if you’re old enough lol!
This detailed history of Open Options in Waterloo, ON mentions Tom Mitchell, CEO. I worked with him briefly at Imperial Oil back in the mid 70s. Article fails to mention Dr Gerry Sullivan, a cofounder of Open Options, who was in my engineering class at U of Waterloo. I would have really enjoyed working with these guys.
This book written 8 years ago tells the remarkable story of the rise and fall of Research in Motion, the company that invented the BlackBerry smart phone some 20 years ago. It is a detailed account of the partnership between Co-CEOs Mike Lazaridis the engineer, and Jim Basillie the business executive, that changed how we live and work on a global scale. A truly fascinating Canadian story!
I came across it after watching the BlackBerry movie (2023) which I highly recommend. The movie is a humorous factual and fictional account of the amazing Blackberry story based on Losing the Signal. I watched it 3 times and will likely watch it again. The movie uses a lot of foul language, so beware.
You can watch this movie in its entirety (and many others) for free on CBC Gem.
Back to the book. I found the the BlackBerry story intriguing for 3 reasons. Firstly, I have been the addicted owner of 3 BlackBerrys over the years and was absolutely hooked on them for over 15 years. I got my first one in government in 2006 or 7 (a Quark I think). In 2010 I purchased the BlackBerry Bold and in 2015 the BlackBerry Classic which I still use as a wifi machine! Secondly, I went to the University of Waterloo and graduated in chemical engineering 6 years before Mike Lazaridis started there in electrical engineering. For 2 years I lived within a few blocks of where this story was to take place. Finally, I worked for a small high tech start up firm for 3 years (SHL Controls) and experienced the personal pressure, uncertainty and excitement that working for a fragile start up IT shop entails. I identify with the story,
The most amazing aspect of the story for me is how two (Canadian) men with totally different skill sets, interests and backgrounds, could work so congruently together for over 10 years to innovate and keep the corporate sharks like Motorola, Google, Bell South, US Robotics, Palm Pilot etc. from squashing them. Where is Waterloo Ontario anyways? Things begin to unravel when Apple brings out their iPhone in 2007 and Google licenses their Android operating system in 2008 to any handset manufacturer. The future market for non IOS or Android phones was dead for all intents and purposes. But as the book describes in great detail, it took BlackBerry management another 3 to 4 years to admit this and get out of the business. The company has since morphed into a software and services company (like Balsillie was seeking to do when he was pushed out along with Lazaridis in 2012.). Lately it is really struggling.
The authors McNish and Silcoff are veteran business reporters who know how to get the facts out clearly, Based on countless interviews of staff and management including the Co-CEOs, the book won several awards. Somewhat exhausting in detail at times, a very thorough account of the Lazaridis and Balsillie years that, when the end finally comes, it almost made me cry. Note both men are billionaires and have since sponsored several scientific and leadership institutes and think tanks since “retiring.” Overall a riveting account of a great Canadian business story, 10/10.
