avidamoeba

Perhaps this is how they're trying to solve the factuality problem of their LLM? Limit the sources to a known good allowlist. Train the AI answers model on those. If that's what's happening, it would be ironic that they'd have to undo their search results enshittification in order to overcome the LLM's inherent flaws. Of course the regular search results could keep being shitty. In fact they might get worse depending on the cost of the AI Mode.

I am again shocked at this semblance of accountability. Would the victims get some of these fines or is that not the kind of case?

I don't doubt your parents' anecdote.

My anecdote is that I've worked for 3 major Canadian private corporations (10000-100000 employees). Currently working in a major American corporation. My wife works for a Canadian public institution. I know a few others who work at large Canadian corporations. The waste and inefficiency at the corporations (banking, telecom, food) is staggering and similar between them despite the different sectors they operate in.

The one I'm currently at (a well-known name) just canceled a 4-year old program that 75% of the software engineering teams worked on, because the leadership finally figured out that we can't deliver it fast enough with sufficient quality to support hardware going in production. That has been obvious to a lot of us in engineering for at least 2 years and we've kept raising the alarm. The new hardware is now going to ship with the old software stack, some people would get reassigned, others laid off *, no one outside would ever hear about what happened. The profit figures are going to go up due to the decreased labour cost and there would be no new costs down the line since the program was mostly pointless to begin with. Our current software stack is more than good enough to last us at least another decade. The initlal program was sold to us by a third party corpo as something that could reduce cost over time. It was a lie that worked because of massive undercount of labour. Things like this happen all the time in large private corporations. I've seen it more than once myself.

In contrast, my wife's public institution is way more careful with how money is spent, how and who they hire, and they pay less for most positions. They're constantly understaffed and overworked.

And yet all of the above deliver decent products and services to people. The corporations I've worked for make significant profits. They keep their prices as high as they can be and their product design and marketing is such that they sell as much as they could, trying to increase every quarter.

Given my experience and what I've read so far in my life I've come to the realization that inefficiency isn't dictated so much by the type of institution or even the market pressures it faces. Rather the primary driver is complexity. I think all large enough organizations develop similar structures and problems as they're inherent to any social activity with imperfect information.

* Actually the cost of that mismanagement doesn't end at the layoffs. Layoffs externalize a part of the cost of the mistake to the rest of society - workers families, social programs, healthcare, political stability, etc. I.e. the cost is socialized.

The same way decisions are made in large public (state owned*) or private firms as well as public service institutions in our systems. People get hired and do these kinds of decisions across such institutions all the time. All of these issues are tackled and there are imperfect solutions that with all the pitfalls and mistakes work alright.

* Actually I'm not sure if the US does state corporations but we do in Canada. For example in rail, air travel, nuclear power (design and implementation), telecom, housing, etc. Of course we've privatized many of them with not so great results for the public, just like the UK did. They've made a lot of money for their private shareholders though.

From Wiki:

The company began operations on November 1, 1996, when the government sold and moved the country's air navigation services from Transport Canada to the new not-for-profit private entity for CAD$1.5 billion.

Looks like another privatization success story. At least its structure doesn't seem predatory.

Dies it cancel out like a double negative?

The thing is, you don't have to centralize the entire economy in order to be positioned to solve for the difficult problems facing us. But I think having a robust long-term looking economic planning in the sectors everything else rests upon, like energy, natural resources, transportation, logistics, education, care, housing, banking, finance, basic research, defense, food, electronics components, chips etc. itself produces small firms that can do new things much easier due to the availability of materials, equipment, capital and labour at low costs. This doesn't mean that for example every chip made would have to be planned by someone in the capital. Nothing of the short. It means that the planner says, we need to have the designs and ability to mass produce low cost, high performance RISC-V cores for PCB integration by 2025. Then Huawei and SMIC get their shit together and assemble teams to do this. If they need more capital they get it. If there's someone who wants to start working on a design with a new team, there's going to be capital available for a startup. Once the core is in production, that core becomes an input for other large and small firms, or individuals who want to do something with a low cost processor in it that now have a viable path to form new firms. This is why there are a shit ton of small Chinese firms making innovative consumer items. This is why US firms keep explaining how they can't possibly make this or that product in the US in the context of tariffs. When everyone downstream from them is profit maximizing, their inputs become prohibitively expensive. Someone was talking about how much it would cost to source a small neodymium magnet motor for a consumer pump made in the states and said it's so expensive that it's only viable for defense, aerospace and such. And then the neodymium still comes from China.

The consumer parts of the economy where you have smaller firms with interesting products often sits at the tip of the existing supply chains and infrastructure. Perhaps use them differently. That's also true for local breweries as they rarely grow their hops, wheat or build their own equipment from bare metal, or mine the metal.

But even in the consumer sector here (Canada), most of the aisles in our grocery stores are filled by the products of a handful of companies. PepsiCo, Coca-Cola, Unilever, Colgate-Palmolive, Kraft, Nestle, Kellogg's, Danone, Mars, Mondelez and the store brand. Then you have Big Ag product filling the produce and meat section from the usual suspects. Outside of who provides these firms with direction and who collects their profits, they're what state-owned enterprises look like.

Economic growth is just easy to check data. Many people are completely unaware even of that. Not saying you specifically are. If one's interested beyond that, could look into other indicators such as education, life expectancy, etc.

China's relaxed some sectors and not others depending on their importance and the competency within.

Generally markets and competition do well in figuring out how to do something we don't know how to do well and cheap. Once we figure that out for some product or category, profits fall competitors fail and consolidation sets in, cost of production falls further due to decreasing duplication and increased scale. At that stage, you have to re-establish control or the monopoly begins draining resources from the economy by raising prices. I think this is what China's doing. They do a high level plan on what they want to develop, get their centrally controlled bits needed in place, e.g. capital from banks, raw resources for batteries, then let existing or new companies enter a competitive market to develop the thing. We saw this occur with EVs. I don't think they've reached the consolidation point yet.

A sector that hasn't been relaxed for example is banking and for a good reason.

But beyond relaxing control, the other very important thing that was relaxed was foreign direct investment. Getting factories built in sectors you don't have by foreign firms, almost always as joint ventures with the clearly stated goal of knowledge transfer. Personally I think this is likely a bigger contributor to their economic explosion than planning changes although it also requires planning changes itself.

On Walmart, I think what you're looking at is the feedback mechanism of shop/factory data going up the stack. A Walmart store has no capability of doing the planning needed to get a requested amount of a certain good on its shelves, beyond requesting it from Walmart HQ. Walmart HQ computes and directs everything from telling how many thousand plastic trays a factory in China should make, to eventually getting the 50 requested by a store in Bumfuck Nowhere delivered. Data feedback mechanisms exist in every operation of meaningful complexity. They existed in the USSR, they exist in China, they exist in every corporation I've worked in, currently automotive. Every large corporation takes data from its operations, either through people or directly from processes, or both, or from their products themselves, computes projections, decides what to make more of, less of, what program to cancel or start, what input resources to get more or less of, how many people to hire or fire, all in order to support the desired new projection. Then they turn that into their expected growth numbers for the next quarter and spit it out during their investors call.

At this point buying Chinese made in China or perhaps European made in China might be the least bad option. At least they're cheaper and less money goes to an American multinational tumor. But you gotta make sure it's actually safety certified. I have some ZooZ Z-Wave units, but I've also been eyeing Aeotec.

And their products are actually made in Canada?

I tried making it as concise as possible while preserving the main bits needed to follow the logical arguments and using as little jargon as possible. I'm glad you appreciate it! ☺️

I don't think that's true.

Central planning ran the USSR and its satellites for some 40-70 years. They didn't even have computes for the majority of this period and many of these economies experienced high rates of growth. If I remember correctly, the USSR speedran economic development so that the GDP per cap of the USSR increased 10 times between the beginning and the end of the experiment. The US grew about 3 times during the same period while being the main world hegemon, profiting from the vast majority of the world. Of course there were problems, like the famines in the 30s, but they didn't repeat post-WWII. It's not like capitalism hasn't caused famines around the world either. So despite the standard criticism, I don't think planning did poorly overall.

China is also demonstrating how long term central economic planning allows to build an economy efficiently, with a long term focus and avoiding most crises capitalist economies experience on regular basis. They're clearly leading in development of solutions to climate change in a way that is above and beyond any other economy, in solar, wind, battery and EV production. Just earlier this month we saw their emissions fall despite higher electricity usage for the first time. And they're powering a lot of everyone else's renewables transition. Then on the ageing front, they're already doing a lot of manufacturing automation. I read they're also doing farming automation now. Apparently DJI's other job is spraying fleets for example. I don't know much about healthcare and elder care but I imagine they're either working on reducing labour needs or planning on it. So yeah, while we're afraid of automation because we know we'll be left jobless and/or deskilled by the capital owners (even if it eventually leads to a crisis), them socialist fkers don't have that problem. The more they automate, the less population they need to maintain and grow their standard living, the cheaper they can manufacture what they make, the easier the ageing population problem becomes. Given how many universities they're opening each year, growing the highly skilled research labour share, I think they're only going to accelerate these trends.

One more thing about planning - the largest capitalist corporations that deal with actual physical production and large supply chains already do the type of planning that's been done in past and present socialist states. In fact it's probably larger and more complex than some whole countries. A common example is Walmart. You'll find little market forces within its operation. In fact companies like this, that have complex enough products and/or supply chains do everything they can to isolate themselves from the free market in order to decrease uncertainty, therefore increase the likelihood of successfully producing and delivering the product, and of course maximize their profits. If you consider how every major sector of the economy is getting consolidated through competition into a monopoly or oligopoly, and similar economic planning process goes on in most of those, you could perhaps see how capitalism itself trends towards central planning. Of course for profit maximization and not social benefit.