__ __ _ _____ _ _ _
| \/ | ___| |_ __ _| ___(_) | |_ ___ _ __
| |\/| |/ _ \ __/ _` | |_ | | | __/ _ \ '__|
| | | | __/ || (_| | _| | | | || __/ |
|_| |_|\___|\__\__,_|_| |_|_|\__\___|_|
community weblog
AI is the major driver of innovation in Canada and around the world.
Canada's National Artificial Intelligence Strategy: AI for All. Big Capital Shoeshine Boy (and Prime Minister of Canada) Mark Carney and Artificial Intelligence Minister and literal Empty Suit Evan Solomon on Thursday unveiled Canada's new AI strategy. One day after a rousing speech about True Canadian Values, Carney tells the nation we are going all in on replacing well paying but unfulfilling desk jobs with de facto unemployment without a word towards increasing taxes on corporations or providing a basic income policy. As a bonus, the document itself reads like pap from last generation LLM models.
In theory we're not supposed to editorialize on the front page but I defy you to say I'm wrong about any of this.
posted by seanmpuckett on Jun 05, 2026 at 3:52 PM
---------------------------
It's truly awful. No mention of compensation to artists and writers who have had their work stolen by these corporations, either.
posted by jordantwodelta at 3:57 PM
---------------------------
I remember Solomon getting fired from the CBC for his problematic art dealing but didn't realize one of his main customers was Carney. Sweet Jesus.
posted by brachiopod at 4:25 PM
---------------------------
Fuck all of these fucking ghouls.
posted by adrienneleigh at 4:36 PM
---------------------------
I'm only halfway in, but this seems awful in every way but one: sovereignty. Canadian ghouls are preferable to American ghouls.
posted by justsomebodythatyouusedtoknow at 4:45 PM
---------------------------
To link to Julie LaLonde (a famous Franco-Ontarian feminist) on Blue Sky:
Canadians want housing, affordable groceries, a robust health care system, quality education, lives free from violence and to enjoy the beautiful nature this country has to offer.
We do not want fucking data centres gobbling up our water supply, poisoning our air, and making us dumber.
I think of all the water we will lose and all the good things we could do with that kind of money and I can't decide if I want to flip a table or start sobbing.
What a waste.
posted by Kitteh at 4:59 PM
---------------------------
One of the cleanest electricity grids in the world makes Canadian compute among the most sustainable on the planet
And hydroelectric is pretty much the cheapest per MW, if not one of the cheapest, which is what will attract development.
Another advantage of sourcing AI local to Canada is that it also keeps data sovereign, housed within the country, instead of hosting it in unfriendly and privacy-violating countries, like its unfortunately now-disreputable neighbor to the south.
Data centers use fresh water out of legal expediency. They can run on so-called brown water (albeit with higher operational costs) but their operators use fresh, because there isn't regulatory guidance to force them do otherwise.
Not saying that AI is great, but that if it is inevitable, Canadian leaders would be irrational not to consider the use of a technology to benefit its people to whatever extent is possible, while also managing the downsides through regulations.
posted by They sucked his brains out! at 5:23 PM
---------------------------
Look, AI is not fantastic for all of the reasons that have been mentioned here in this thread, on MetaFilter in general and in many other places. I don't disagree with anybody that AI is more than problematic. AI scares me and will likely cause me to lose my job in the near future which is obviously very concerning to me.
I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away. Individually, we can make our choices about how much we'll use AI, how little or not at all but from an economic and political perspective, governments have to engage with it and make it work. The sooner governments start making laws on how AI can be used the sooner people who are against AI can push to ban it or change it.
This strategy that has been put together reads badly but It doesn't read like LLM to me, it reads like all of the other "pillar" strategy documents that I've read over the years. They all suck, this one more than most. It doesn't address a lot of things that need addressing but it also addresses things that are importantly to Canadian sovereignty such as owning our data farms, making our own equipment and many other things. Do I think that it's mostly empty promises? Yes! But again, it's a start.
posted by ashbury at 5:33 PM
---------------------------
We're not supposed to editorialize on the front page?
posted by subdee at 5:36 PM
---------------------------
Question, accusation, or realization
posted by runsrealgood at 5:40 PM
---------------------------
I don't want AI to become the Gaza 2026 that sinks the mid term Democrats and puts PP into power, because that will be exponentially worse by every measure.
Until the orange number is higher than the blue number, I will still support Carney. Asia is going all in on data centres with no public pushback to speak of, due to China's nihilistic competitiveness on the topic.
posted by CynicalKnight at 5:54 PM
---------------------------
"have to engage with it and make it work"
I think people are broadly aware of this. However... Why does this always look like uncritical embrace, forced use, indoctrination, and cash and resource transfers to AI companies? Why does it never look like, say, nationalized infrastructure, skeptical adoption, and/or carefully-tested, domain-appropriate deployment, if such a thing can be done?
This does not, to me, feel like a government response that's going to lead to a significant improvement in the ability of Canada to "benefit its people to whatever extent is possible, while also managing the downsides through regulations" or of "people who are against AI can push to ban it or change it". Instead, it cedes a bunch of ground and will make it harder to create effective regulations and/or push back against LLM usage.
posted by sagc at 6:04 PM
---------------------------
ie, it's good to regulate technology, but it's less good when that regulation is pre-captured by the industry it's supposed to regulate.
posted by sagc at 6:07 PM
---------------------------
can somebody run this mess through an LLM to summarize it please, my eyes just slide right off of this shit
also what is up with the arrows in the middle of this diagram
posted by egypturnash at 6:13 PM
---------------------------
All of this ends as a net loss for all of humanity. Even the billionaires hiding on their Disney+ Bunker islands, after having consumed the last avocado and the innards of their supplicants, will be forced to regard their own failure to measure the worth of what they've "achieved" versus the intangible numbers attained. Abject moral failure that seems tantamount to "evil" in the most abstract way. Void of any construct or basic presence of conscience, of any consideration for other beings at all, save that which can be attained for supposed gain. And in no way am I saying I'm not complicit. I like my air-conditioned nightmare doomscrolling and live music.
tl;dr: the sociopaths won and everyone else is cat food now
posted by Token Meme at 7:06 PM
---------------------------
>people who are dead set against AI are not acknowledging the fact that it's not going away
Why must people keep repeating this? If the idea of exchanging your labor for wages can go away, AI can most certainly go away. You're just parroting talking points for the trillionaires. Stop it.
posted by Sing Or Swim at 7:08 PM
---------------------------
I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away
Racism and bigotry aren't going away either but that doesn't mean I'm not going to do my best to fight against them every single day.
posted by jordantwodelta at 7:14 PM
---------------------------
ashbury: "I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away. "
We totally know! This is mentioned over and over and over again and I can't decide if I want to flip a table or start sobbing.
posted by tiny frying pan at 7:25 PM
---------------------------
Shitty stuff is worth fighting against, yes!
posted by tiny frying pan at 7:27 PM
---------------------------
Marketers chose the slogan after consulting an AI tool, looking for suggestions, Shinsegae Group said.
posted by lalochezia at 7:39 PM
---------------------------
One problem with using hydro power as a reason to promote data centers is that the largest number of these centers will be in Alberta powered by natural gas.
Yes, we have to bend with technology, but we need to monitor and regulate this tech.
posted by CCBC at 7:55 PM
---------------------------
ashbury: "I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away. Individually, we can make our choices about how much we'll use AI, how little or not at all but from an economic and political perspective, governments have to engage with it and make it work. The sooner governments start making laws on how AI can be used the sooner people who are against AI can push to ban it or change it."
Why exactly are we supposed to wait until the government makes laws before we start pushing the government on what those laws should be? Isn't that backwards?
posted by adrienneleigh at 8:03 PM
---------------------------
I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away
There's a sucker born every minute.
AI isn't going away in the same sense that bullshit, swindles, and cons aren't going away. But the only times I ever see AI is when some fucking asshole cons me into seeing it, when some real person is complaining about the harms it does*, or when I feel like learning about it in the same way I sometimes find it interesting to read about criminality.
Generative AI is not relevant to my life, and in that sense it has already gone away, probably permanently, the shrill and distant bleatings of the hucksters notwithstanding.
*The other day a thruhiker told me he had asked AI where to send his resupply box, and it had given him the address of the free hostel in the next town. Almost perfect bullshit: believable, a correct address, a place hikers stay, and completely wrong. They don't accept packages. He was somewhat worried about the mistake. I didn't get on his case about the perils of using AI, I just told him that tiny town post offices are often very much on their game, and if so the package would probably be held for him there. Which it was. There are and will continue to be people who want to do good work and avoid destroying everything with hokum and balderdash.
posted by surlyben at 8:47 PM
---------------------------
> > people who are dead set against AI are not acknowledging the fact that it's not going away
> Why must people keep repeating this? If the idea of exchanging your labor for wages can go away, AI can most certainly go away. You're just parroting talking points for the trillionaires. Stop it.
The problem is that people are grasping at straws for reasons that largely revolve around ego-protection. I was at a technical conference today in a field very vulnerable to AI-disruption and spoke with a table of people at lunch *absolutely convinced* that token prices are subsidized such that that any day now the "true cost" will reveal itself and everyone will stop using LLMs because it's too expensive. But Moore's law has *never* worked that way in history and the reality is that cheap open/self-hosted models impose a strict ceiling on the maximum inference price that the market will bear. Normalized to current (or any time frame's) costs and ability, LLM inference will only get cheaper, that's something that should be obviously axiomatic but a huge number of people are waiting for a deus ex machina that I can sincerely promise you *is not coming*. Framing this reality as "just doing the trillionaires' work" is just more cope.
posted by haycorns at 9:38 PM
---------------------------
[One deleted; being snarky at another user (who's not even in the thread!) is just not going to add to the conversation.
Also, seanmpuckett: yeah, you're not supposed to editorialize to this extent in an FPP, and you know that. This got a few flags but it's also got a reasonable amount of conversation going on, so i'm going to leave it alone, but please don't make a habit of this? (Note: i was an early commenter, without my mod hat on. If it hadn't picked up steam, though, it would likely have been deleted earlier.)]
posted by mod_adrienneleigh at 10:15 PM
---------------------------
lalochezia: "Marketers chose the slogan after consulting an AI tool, looking for suggestions, Shinsegae Group said."
So i've been following this particular story, and this is actually an interesting example of a different kind of accountability laundering than the usual: These folks are blaming 'AI' for something that was a deliberate human decision. The decisionmakers on this were absolutely dogwhistling to fascists on purpose; there's an explainer thread over here on Bluesky. (It wasn't just the slogan; there were a number of deliberate choices made including the numeric value of the volume of the special cups, which was like a S. Korean equivalent of "1488".)
posted by adrienneleigh at 10:24 PM
---------------------------
people at lunch *absolutely convinced* that token prices are subsidized such that that any day now the "true cost" will reveal itself and everyone will stop using LLMs because it's too expensive.
They are convinced about it because they literally are subsidized.
posted by Pyrogenesis at 10:46 PM
---------------------------
I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away
"...it is easier to imagine the end of the world than it is to imagine the end of capitalism. That slogan captures precisely what I mean by 'capitalist realism': the widespread sense that not only is capitalism the only viable political and economic system, but also that it is now impossible even to imagine a coherent alternative to it."
-- Mark Fisher
"We live in capitalism. Its power seems inescapable. So did the divine right of kings. Any human power can be resisted and changed by human beings. Resistance and change often begin in art, and very often in our art, the art of words."
-- Ursula K. Le Guin
"The Butlerian Jihad, occuring ten thousand years before the events described in Dune was a war against thinking machines who at one time had cruelly enslaved humans. For this reason, computers were eventually made illegal by humans, as decreed in the Orange Catholic Bible: 'Thou shalt not make a machine in the likeness of a human mind.'"
-- Brian Herbert
posted by overglow at 11:32 PM
---------------------------
They are often subsidized, especially for subscription-based pricing, but as far as I've seen it's still broadly true that models at a given level of quality have predictably been getting cheaper to run (both from improvements in the the models themselves and improvements in the hardware).
"Subsidize it until it either actually gets cheap or until you're indispensable" is one of the most well established tech game plans.
posted by atoxyl at 11:35 PM
---------------------------
You have to see how a Canadian AI company called Croptimistic is helping farmers precisely map their soil.
Croptimistic
Tiredly rubs forehead
posted by BrStekker at 12:01 AM
---------------------------
So much for Carney's "values". Marc Lee calls it here in his review of Carney's 2021 book.
"For example, solidarity is not about the role of unions in supporting workers to get better wages and working conditions. Instead, it's more of a vague appeal for education and training so that workers have the skills needed to thrive. Similarly, fairness and responsibility are not about progressive taxation and ensuring a more just distribution of income, but an appeal for markets to work better through prudence and disclosure.
For Carney, the banker, resilience has more to do with identifying and preventing systemic risks to the financial system rather than ensuring our buildings and infrastructure can survive fires, droughts, floods and heat domes. Meanwhile, sustainability is about green investment opportunities shaped by a strategic direction set by governments through carbon pricing, regulation and financial disclosure. Ironically, Carney's support for carbon pricing is among his strongest policy recommendations.
It's not clear how Carney would come to grips with the massive inequalities in our society, the rapidly declining state of the climate, and the dark side of new technologies and their potential to displace mass amounts of workers. Nor does trade factor in, as the second Trump administration collapses the whole basis for Canadian trade with the United States, and the post-war global order, with the United States as hegemonic power, starting to crumble.
At the end of the day, Mark Carney's values proposition is not enough to "build a better world for all." You can take the boy out of Goldman Sachs but the imprint of Goldman Sachs lingers. Carney offers up a solid understanding of how we got here, and the complexities of building a modern mixed economy, but skirts over more fundamental economic challenges. As for Carney the politician, that chapter remains unwritten."
posted by cotton dress sock at 1:13 AM
---------------------------
lol, and people say *my* politcal FPPs are biased. This whole post is the antithesis of reasonable good-faith discussion. Worst of the web.
posted by mrjohnmuller at 1:23 AM
---------------------------
I was dead set against AI (photographer, so...) but decided I don't like being one of those people who renounces something as evil without learning more about it or trying it. I got into it to at least have a semi informed opinion. AI has turned out to be, well, empowering in ways I did not expect. I'm not only hella more productive at my day job (senior insurance broker) because of AI, but it has helped me - a non coder - develop some pretty damn amazing software that I need and no one else was making for me. I'm chuffed.
Yeah there are a lot of problems to work out. There's also a lot of potential that people are overlooking is how I feel.
Cheers.
posted by snapsmack at 2:13 AM
---------------------------
It's disappointingly sentimental for a statement of the Carney government which is usually better, but that was probably politically unavoidable. This all could have much better been two sentences of the King's Speech: "As earlier governments have always done in the face of technological and economic change, Our government will modify the implementing regulations and enforcement strategies for consumer and labor protection, environmental, and domestic content laws so that AI vendors and users will continue effectively to be obliged to comply with those laws. Our government recognizes that AI, like prior major technological changes, will likely reduce the human labor input required per unit produced of extant goods and service, and accordingly Our government will prioritize policies supporting (a) an increase in unit production capacity and demand, (b) improvement in the quality of goods and services, and (c) readiness of Canadians for the new and changed jobs that will be required for this incremental production."
Still, Canadians should be happy to have this much. Could you even imagine a US "National Strategy for AI"? A Trump white paper's creaking intersection of awkward misreadings of lobbyist memos and right-wing social-media sentiment, or a Democrat's white paper's not-even-trying-to-intersect mix of left-wing media social sentiment and very accurate readings of a slightly different set of lobbyist memos...
posted by MattD at 5:42 AM
---------------------------
jordantwodelta: " No mention of compensation to artists and writers who have had their work stolen by these corporations, either."
Maybe someday we'll preface every AI produced whatever with one of those pious land acknowledgements reminding everyone that all this is built upon someone else's property.
posted by chavenet at 6:29 AM
---------------------------
that it's not going away
The issue is that it doesn't exist, but it will ruin everything anyway
AI is too good at disguising the decisions that the powerful want in "inevitability." It's just the "natural order" for white men to be in charge of everything, don't cha know.
The inevitability and unaccountability is the product.
The problem is not that the AI 'accidentally' sent the rocket to kill those children, it's that the generals can plausibly point at the computer when the Hague comes for them for killing the children.
posted by eustatic at 7:12 AM
---------------------------
This isn't great, the statement on AI, but I think it is largely bullshit anyways. Here's the thing - if you think that's awful have you heard Evan Solomon trying to defend it? I've heard two interviews with him on the radio (at the moment I can only find this one online) and trust me you will hate AI and him even more. Who on earth thought this guy was a good idea as a mp, as a minister and as a minister of AI? As a Northern Ontarian I can smell a southern Ontario carpetbagger from a kilometre away. The worst.
posted by Ashwagandha at 7:27 AM
---------------------------
If this puts some safeguards, thought, and/or limitations on AI then it's not all terrible (and better than we were at last week).
posted by mazola at 7:42 AM
---------------------------
I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away.
I'm more concerned that people are dead set on the idea that this "fact" is in fact factual.
posted by brundlefly at 8:29 AM
---------------------------
[Two deleted.
runsrealgood: The user in question wasn't in the thread at the time of my previous mod note, but "not being in the thread" is not the most salient issue with taking cheap shots at other users. Please stop.
The other comment was deleted for quibbling about a moderation decision. I might have made a bad call here, but as always, please take it to MeTa or use one of the other available pathways (like the MOC contact form).
mrjohnmuller: I didn't delete your comment because it's as much about the post itself as about the moderation decision, but please do stay on topic.
[edited because i forgot to include something]]
posted by mod_adrienneleigh at 12:23 PM
---------------------------
"Carney assures Canadians that AI the fastest way to give all of our money to American Billionaires" - Beaverton
posted by ovvl at 1:10 PM
---------------------------
AI nationalism is a symptom. AI is an existential risk to humanity, Canadian AI means nothing if Carney is selling this as if it were a problem analogous to Canadian healthcare.
posted by polymodus at 3:36 PM
---------------------------
Think i got deleted as well...
Sorry about that.
EDIT: But once you are deleted, it's hard to figure out what bad thing you said?
Metafilter is just the best!
posted by Windopaene at 4:48 PM
---------------------------
"Maybe someday we'll preface every AI produced whatever with one of those pious land acknowledgements reminding everyone that all this is built upon someone else's property."
I put an ethical statement in software I developed using an AI pointing out that the LLM was trained on sus data and I give away what I make for free and as Copyleft because I feel that something I created based on code that (essentially) came from The Commons must be returned to The Commons. I don't regard it as pious so much as necessary and honest.
posted by snapsmack at 5:51 PM
---------------------------
I think the difference here is that you're pairing the ethical statement with an ethical action while land acknowledgements are mostly empty gestures.
posted by Space Kat at 6:15 PM
---------------------------
"while land acknowledgements are mostly empty gestures"
Empty to you. Not to everyone.
posted by snapsmack at 6:55 PM
---------------------------
I have misgivings.
I've lived in places where increased development has affected the water table (and non-controversial development... hospitals, housing), so while I think a lot of the environmental problems with data centres are solvable ones, I also know that even something that looks "green" and "good" on paper can cause significant disruption to the surrounding people who live there. Data centres are needed, not just for AI (data residency is a privacy issue, and one we don't think about enough). But I don't think the people designing them really understand the needs of the surrounding residents, nor care too.
I don't like how the document really doesn't spell out what we're talking about when we talk about AI. It gives all the usual "good news" stories, but the language on what kind of AI they are talking about isn't precise. They need to be specific otherwise AI = gen AI for most laypeople. And it is difficult to have that conversation because it's already been hijacked by big tech to make AI = whatever venture they hope to profit from.
I am also worried that this strategy will be an excuse to loosen regs around things like procurement and consultation. There's already been a movement in Canada of industry rich people who want weaker regulation - and since Carney at his heart is a capitalist (I think he does have a moral compass, but I think every solution for him is firmly rooted in capitalism and global finance), I'd guess they probably have his ear on this one.
I do think everyone needs a basic level of AI literacy. My hope is that it's through a critical lens, focusing on risks to privacy, democracy, environment, etc. and how to lessen those impacts if not avoid them entirely. But hope springs eternal. My worry is that it will approach literacy with a bias towards boosting the technology rather than the "health and safety" talk that's needed.
I also think that the US tech sector controlling all things tech in Canada (which it does) is bad. Especially right now. So Canada taking control of domestic AI development I think is necessary, even if I don't particularly like gen AI as a technology.
I think in the scheme of Canadian tech news, I'm more worried about the impacts of Bill C-22 (the lawful access bill) on Canadian privacy (and which the committee just voted against having the privacy commissioner speaking at). I'm also more worried about the lack of online harms legislation - apparently in the works since it died in parliament the last time, but still not here yet. While the AI stuff is worrisome, I think we need to pay attention to things more holistically. Gen AI is just one segment of technology among many others that need regulations to protect Canadians from surveillance (by the state or otherwise), deceptive design, privacy breaches, etc.
posted by eekernohan at 5:02 AM
---------------------------
I just engaged in a big argument over at Reddit/gardening because a poster pleaded with people not to bring gardening questions to ChatGPT & waste water.
The mods of that forum deleted the whole conversation and I think that was super heavy handed so I appreciate metafilter for working hard to keep these conversations on track.
I am on the side of 'don't ask a disembodied compendium of human knowledge questions about when is the right time to put out your beans'. Go ask a garden centre in your zone.
But anyways. I am irritated by everything the Canadian government is doing right now because it feels like the response to massive change on a geopolitical, sociocultural level (as always) is - pay for a bunch of useless consultants to offer 'training' to Canadians while in the back rooms we sell off all our natural wealth (in this case space, hydro electric power) to the highest bidder. The implication that Canadians need 'A.I. literacy training' is so embarrassingly out of touch.
The country managed to mess up turning a profit from legalized weed!!!
We never get this right, we sell early for what seems like a lot of money, the least controversial or innovative version of whatever's going.
It is absolutely a bankers response.
posted by MirJoy at 7:10 AM
---------------------------
ashbury: “I'm also concerned that people who are dead set against AI are not acknowledging the fact that it's not going away. ”
“In 1979, Mattel sold toy Battlestar Galactica fighter ships that fired a tiny spring-loaded "torpedo." One 4-year old choked on it, and the entire line of torpedo toys was recalled and never sold again.”
‘He asked ChatGPT for advice on a mass killing. Minutes later, two were dead. OpenAI is wrestling with how to handle its most chilling conversations.’
"ChatGPT Wrestles With Its Most Chilling Conversation: How Do I Plan an Attack?," Georgia Wells, The Wall Street Journal, 2 May 2026
— The Wall Street Journal (@wsj.com) May 2, 2026 at 9:11 PM
— Matt Zoller Seitz (@mattzollerseitz.bsky.social) May 2, 2026 at 11:42 PM
posted by ob1quixote at 7:42 AM
---------------------------
I started to read the linked "AI Strategy" document, but I will have to defer a serious read to later; I don't have patience for a box of gov't soft-soap right now.
I'll offer this anecdata instead, since it's somewhat related and this is the most current AI thread.
My (Canadian) doctor runs a large and pretty competent (Canadian) practice, with a few doctors, several nurses and medical technicians, and their own blood lab. He has been proactive in looking into ways to improve the practice's efficiency. One of the things he recently added is an AI "virtual receptionist" to deal with most phone callers. This service is mainly to deal with booking and changing appointments. It's one of the phone options to choose when calling in, so the caller does know that they're being passed to it. And it's easy enough to request a transfer to actual live desk people at any point. As far as I know, it does NOT offer medical advice or make suggestions, and will transfer to a live person if there's confusion or misunderstanding.
I had a regular appointment with my doctor last week, and I took the occasion to briefly discuss the new virtual receptionist. It has worked well enough at understanding me over the phone, and with offering and booking appointments, but one criticism I had was that they had deliberately added "office noises" in the background. To me this seems like fakery, an effort to make people think that they are actually dealing with a person.
My doctor told me that the system has been working well so far for the practice. By handling most of the daily burden of appointment-booking, it has freed up a lot of time for his reception staff, with a consequent improvement in staff morale. He told me that patient feedback about the new service has been positive, even to some patients saying that they liked that it sounded and acted like a receptionist. Including with the office noises.
It was a bit surprising to me that not many apparently shared my negative opinion about the office noises. Patients know that they're talking to AI, yet are just fine with touches that make it seem more like a real person. (well, my Mom named her Roomba... people are funny. FFS, get a cat, Mom.)
So - a real-world deployment of basic AI, that seems to match my shred of optimism about how AI will first impact the broader world of work (not counting software development). It's taken over a basic, repetitive, dull task that no-one really enjoyed, seems to be doing it effectively and is acceptable to clients. As far as I know, no staff were laid off because of this change. And people are responding positively to the humanizing touches.
Anyway, I think this is a good example of how careful deployment of AI-powered services into very well-defined roles can have a positive impact. And I still believe that AGI - a one-size-fits-all drop-in replacement for any human clerical function - is still a pipe-dream that won't be achieved shortly just by over-feeding ever-larger LLMs.
As I said, I intend to put some time into digesting that government AI strategy document. A national "strategy" is (hopefully) better than just cringing and letting multinationals dictate the rollouts.
posted by Artful Codger at 8:51 AM
---------------------------
snapsmack: "I put an ethical statement in software I developed using an AI pointing out that the LLM was trained on sus data and I give away what I make for free and as Copyleft because I feel that something I created based on code that (essentially) came from The Commons must be returned to The Commons. I don't regard it as pious so much as necessary and honest."
Using slopbots and giving the result away for free doesn't make using slopbots ethical, any more than stealing someone's property and giving it away for free makes the theft ethical. Land acknowledgements are hollow because there is no way to make land theft ethical while you're still profiting from the theft.
posted by adrienneleigh at 12:27 PM
---------------------------
Artful Codger: "And people are responding positively to the humanizing touches.
That makes me queasy, actually. Large numbers of people are fine with all kinds of things, and I'm not as willing to assume that's a positive overall.
posted by tiny frying pan at 12:50 PM
---------------------------
Canada Finally Has a National AI Strategy. Experts Hate It (the Walrus)
posted by adrienneleigh at 2:08 PM
---------------------------
That makes me queasy, actually.
... me too. Which is why I raised it with my doctor.
Discussing the clinic's experience with the new service suggested to me the possibility that maybe people can understand they're interacting with a machine, and still appreciate that the interaction conforms to the usual patterns and conventions of human transactions. The machines are meeting us on our turf; its not like we're going to speak to AIs in SQL statements. Self-checkout kiosks and banking machines say please and thank-you and wish us a nice day, but we seem to not be queasy about that.
Early days, of course.
posted by Artful Codger at 2:57 PM
---------------------------
Yuck.
posted by tiny frying pan at 3:05 PM
---------------------------
There's a big difference to me in pretending a machine is a human doing human things like existing in an office than a simple thank you at a kiosk. We know the kiosk isn't at work. It isn't pretending to have a work day. There's no "pretend" involved. I don't know why the people in charge of A.I. want to make it so creepy. Oh! So they can replace humans and some of us won't notice.
posted by tiny frying pan at 3:08 PM
---------------------------
Hamilton City Hall saw a massive crowd protest a Data Centre proposal last week, which strongly implied two different things: 1) that many people don't like the unrestricted expansion of Data Centres (especially when heavily subsidized by the govt!) and also 2) that many people don't like the rapid growth of AI. I get it, I don't like them neither.
posted by ovvl at 3:51 PM
---------------------------
Carney did not look happy with Arbour's comments about AI during her installation speech. Of course, she will have no real power to change or affect things, but maybe, if the PM takes the opportunity to consult with her for her counsel, she will at least remind him about the ethical obligations that should come along with jumping on the AI bandwagon.
posted by sardonyx at 8:15 AM
---------------------------