I saw this on Twitter but it needs sharing. About men who think they are supportive of women, but only see women as appendages or extensions of themselves. No. They are someones, not someone’s someones. Why this is even an issue in 2024, I do not comprehend.
Back in gentler times, Parliament Hill here in Ottawa offered a unique attraction: a cat sanctuary.
Alas, the sanctuary is long gone. The cats were neutered and spayed many years ago, so their numbers were dwindling. And when the volunteer-supported sanctuary’s original caretaker passed away, the remaining cats were adopted and the sanctuary was abolished.
One of those cats was all-black Coal, adopted by a gentleman named Danny Taurozzi. Coal is 16 this year, still alive and kicking, the last remaining Parliament Hill cat. Coal is also a cancer survivor.
Cancer can kill cats quickly (as we sadly know all too well) so I am glad to learn that Coal is doing okay for now. I hope he has a few more years left, because when he closes his little eyes for the very last time, it will truly mark the end of an era.
Yesterday, August 9, 2024 marked the 79th anniversary of the last time nuclear weapons were used in anger. Or to put it another way, today began the 80th year without a nuclear war.
Can we make it through the end of this year, too, without major confrontation?
Russia vs. Ukraine. Iran vs. Israel. China vs. Taiwan. These are just the major flashpoints, but probably because they are on everyone’s front pages, they’re also the least likely ones. Fecal matter tends to encounter propellers in the most unexpected places. After all, who would have thought on June 28, 1914 that a student stepping out of a coffee house just as a car carrying a VIP made a wrong turn in the distant city of Sarajevo will change world history? Who would have thought on July 7, 1937 that a Japanese soldier with a belly ache in faraway Beijing would trigger an incident on a bridge informally named after a European explorer who visited centuries prior, leading to the first shots fired in a global calamity that eventually concluded with the aforementioned last use of a nuclear weapon in anger?
War. War never changes, goes the motto of the Fallout family of computer games. And I am increasingly worried that the Fallout games, just like a text game from the dawn of personal computing, A Mind Forever Voyaging, foreseeing a society with rising inequality, political polarization, authoritarian rule, will be treated not as cautionary tales but as instruction manuals…
It’s been nearly two years since the world has become feverish about GPT and its cousines, large language models that for many represented their first real interaction with machine intelligence.
Yet misconceptions abound. Expectations against these language models are often unrealistic, which then result in damning evaluations, still often characterizing the LLMs as mere “stochastic parrots”.
In reality, they are neither just random text generators, nor true intelligences with reasoning capability. They are language models.
I keep thinking that our future would be in safer hands if we let AI-assisted cats take over the reins.
What does that mean? They model, by learning through terabytes of examples, relationships between words and phrases, sections of text. Associations, in other words. They know that apples are red, not liquid; that the sky is blue, not serrated. Which is to say, they model language but language itself models reality.
The sheer size of these models, combined with the tremendous amount of material used to train them, leads to superhuman capabilities. The models are fluent in many languages. They understand intent. They can help uncover gaps in your knowledge, something that happened to me on numerous occasions. They can translate solutions into workable computer code. They know tricks of the trade that even experienced programmers may not be aware of. They can teach you, as indeed the models have taught me a thing or two about specific details of modern machine learning architectures. They can even offer some insight into their own capabilities and limitations.
Throughout it all, however, they rely primarily on their associative capabilities. They are not reasoning machines. Reasoning for these models is as hard as it is for you and me to multiply large numbers in our heads, without the benefit of pencil and paper or a calculator.
And ultimately, they are still just language models. Imagine if the speech center of your brain was somehow excised, made to operate on its own, without being able to rely on other parts of your brain hardware. No sensory inputs anymore. No ability to visualize things, to recall sounds, to imagine anything. No sense of continuity, no internal monologue, no “self”. Just a speech center that, when triggered, responds by generating words, but without the benefit of the instant reality check that would be offered by other parts of your brain acting in supervisory roles.
That’s what GPT and Claude really are.
So to expect them to excel at, say, solving nontrivial logic puzzles is like expecting a suspension bridge to work well as an airplane. Wrong tool for the wrong job.
I can certainly imagine LLMs (and preferably, continuously trained as opposed to pretrained LLMs) in the future, working as part of a larger network of specialized machine learning components, forming a complex “artificial brain”. But LLMs are not that, not yet. They are just one part of the puzzle, though arguably, they might very well represent the most important part.
It is, after all, through language that we learn the ability to not just react to the world around us but to comprehend it.
Having read comments from some Brits who wish to get rid of the monarchy in order to turn their country into a “democracy”, I despair. It is one thing that, in 2024, most folks are illiterate when it comes to science and technology but apparently, history and the social sciences are also badly neglected subjects.
My point, of course, is that these commenters confuse the form of government with the sources of power and the nature of the state.
A form of government may be a republic (res publica, i.e., governance in the name of the public) or a monarchy (monarkhia, rule of one), among other things.
However, both these forms of government can be autocratic (relying on the might of the state) vs. democratic (relying on the will of the people) insofar as the source of power is concerned.
And neither the form of government nor that source of power determine if the state will be liberal (that is, respecting basic rights and freedoms, such as freedom of conscience, freedom of enterprise, or the rule of law) or illiberal/authoritarian.
Midjourney’s response to the mostly Claude-written prompt, “Three regal anthropomorphic cats sitting on thrones, representing the orthogonal concepts of government. One cat wears a crown (monarchy vs republic), another holds a scepter and a ballot box (autocracy vs democracy), and the third balances scales of justice (liberal vs illiberal). The cats are arranged in a triangle formation against a backdrop of a stylized world map.”
To illustrate, let me offer a few examples. I live in Canada: a liberal, democratic constitutional monarchy. South of us is the United States: Also liberal and democratic, but a republic.
In contrast, the DPRK (North Korea) may serve as an example of a state that is an illiberal, undemocratic republic. Saudi Arabia is an illiberal, undemocratic monarchy.
Examples for other combinations are perhaps harder, but not impossible, to find. Orban’s Hungary, for instance, is rapidly converging on a state that is best described as illiberal, but democratic (the primary source of power is the people, not the might of the state) republic. I think some of the states in the Middle East (maybe Kuwait?) might qualify as relatively liberal, yet undemocratic monarchies.
These categories are not perfect of course, and do not cover all outliers, including theocracies, transitional governments or failed states. Still, I think it’s important to stress that the form of government, the source of power and the nature of the state are three fundamentally orthogonal concepts, and that all combinations are possible and do exist or have existed historically.
Understanding these distinctions is important. For instance, there are plenty of historical examples (e.g., the French Revolution devolving into the Reign of Terror, or the Russian revolution leading to the totalitarianism of the USSR) when the transition from monarchy to republic led to a significantly more autocratic regime. “Republic” is not a synonym for “democracy”.
It’s a civic holiday Monday that feels like a Saturday, reminding me of an old Soviet-era science-fiction novel, Monday begins on Saturday, by the Strugatsky brothers. It’s also a rather gloomy Monday morning, so it’s time for me to grumble about a few things.
For instance, how politics infuses everything these days. I signed up to follow a Facebook group dedicated to brutalist architecture, which for some inexplicable reason, I like. The comments section in one of the first posts I saw rapidly deteriorated into political bickering, as to whether or not it was appropriate to repurpose one of the Nazi-era indestructible flak towers in Hamburg as a luxury hotel. Because you know, politics is everything.
Speaking of which, I saw another post elsewhere about employees of a large US company who, after being told how successful the company was last year, were informed in the same breath that the company will cut their pension plan contributions. Needless to say, there followed comments about the evils of capitalism. Having experienced both capitalism and one of its alternatives, a socialist economy with central planning, all I can say is that capitalism works most of the time until it doesn’t; but when it doesn’t, victims are ever so eager to replace it with something that never works instead.
Then there was this post at an online news site claiming that it is practically impossible to run an ethical AI company. Well, what can I say? If you are telling me that allowing machine learning algorithms to learn from accumulated human knowledge is unethical, then sure, you are absolutely right. Then again, I suspect that what mainly drives such complaints is blatant ignorance of how machine learning works in the first place.
OK, well, never mind that, there’s good news. A fusion energy breakthrough: Neutron impact on tokamak components uncovered. Er… Say again? You are telling me that after 70+ years of research, we are beginning to understand why, or how, a heavy neutron flux rapidly destroys test equipment in the lab? Isn’t that like, say, greeting it as a “steam turbine breakthrough” when a prehistoric tribe manages to draw a spark from slamming together two rocks?
Oh well. On mornings like this, I feel I am beginning to comprehend the mood of the late, great Kurt Vonnegut who once told the CBC’s Anna Maria Tremonti to go jump in a lake.
The title of this blog post is used as the byline or catch phrase of the Canadian Centre for Experimental Radio Astronomy, a group operating a 12.8 meter radio telescope, a repurposed former NATO satellite communication facility, located in Carp, just outside of Ottawa.
One of the things they organize is a summer camp for students. Today, I was invited to talk to a small group of students, and indeed I did so, talking (mostly) about my work on the Pioneer Anomaly. It seemed like an appropriate topic, considering that detection and resolution of the anomaly was heavily dependent on radio science, specifically Doppler radio navigation.
It was fun, and my talk, I am told, was well received. I was also offered an opportunity to briefly tour the facility itself. It was fascinating, even though it was insanely hot inside the dome under the August sun. (I definitely needed a shower when I got back home.) The only memorable fly in the proverbial ointment is that I arrived late, thanks to a stupid disabled truck that blocked the Queensway, as a result of which it took forty minutes to get from Vanier Parkway to Parkdale. Fortunately, my hosts were understanding.
I mentioned this before: A Mind Forever Voyaging, a computer game from the 1980s, one of the text adventures of the legendary Infocom, a game in which you play an AI protagonist, sent to simulations of the future to explore the factors behind the decay and collapse of society.
As you venture further and further into the future, things get worse. Inequality, homelessness, violence.
I was again reminded of this game this morning when I saw the news: New mortgage rules are in effect, allowing borrowers less down payment and longer terms. As a result, the monthly mortgate payment for a $500,000 home is “only” around $2,700, give or take.
Has it occurred to anyone that perhaps the problem is not on the borrowing side but on the supply side? That if we lack affordable housing, making it easier for people to borrow money that they cannot afford to repay is not really a solution?
The same newscast again mentioned an increasingly frequent problem, “renoviction”, when people are evicted from their rent-controlled apartments because the landlord renovates, only to learn that they can no longer find a place of residence that they can afford.
Also on the news: yet another old business (opened 1954) is shutting down at the ByWard Market. In their case, it’s the changing nature of the business post-COVID but for many others, it’s the deteriorating public safety. Increased police presence only pushes the problem elsewhere, like Centretown. I had to drive across town today to our car dealership, for an oil change. I saw panhandlers at every major intersection. Not too long ago, such sights were rare, dare I say even nonexistent here in Ottawa. Now, downtown sidewalks are full of homeless folks.
I have said it before when I lamented about AMFV here in my blog: It’s a piece of (interactive) fiction. Please do not mistake it for an instruction manual. Let’s come back from the brink before it is too late. Unless it is too late already…