John Stuart Mill's essay "On Liberty"[1] offered one of the most comprehensive defences of free speech (that is as relevant today as it was then).
You do raise a good point re: tradeoffs in a healthy society. Mill anticipated this objection and addressed it directly. He didn't advocate for free speech without consequences but developed a harm principle specifically to establish what limits are acceptable. Acceptable limits on thought and speech should be based on demonstrable harm, rather than alleged offence, discomfort, or the current popular opinion or cultural disapproval.
He recognises the need to set some limits, yet also the dangers of who gets to set them. Historically, those with power to restrict speech have restricted truth as falsehood. The bar for restrictions should be very high, not because free speech should be absolute, but because the dangers of overzealous restrictions far outweigh the cost of permitting speech we might personally find objectionable. Even completely false opinions might have their value as they force defenders of truth to better articulate their position or reasoning, and prevents beliefs from becoming prejudiced, or platitudes.
I thought I'd share since it's relevant and there may be some younger readers here that might not have come across his work. I really recommend reading it, even if it's an LLM summary as an introduction (as seems to be the trend nowadays)
> the dangers of overzealous restrictions far outweigh the cost of permitting speech
I don't think restricting things such Holocaust denial, or declaring whole ethnicities as subhuman to fan the fires of racial or ethnical hatred is really being overzealous in restricting speech. Which is essentially what people are objecting to here.
It's also worth considering that certain industries (fisheries and agriculture for instance) are subsidised. It's in our national interest to maintain production capacity, so profits are the least of our concerns. Both the UK and the EU's agricultural sectors are heavily subsidised mainly for this reason. It's cheaper to import than to produce locally, especially with our environmental standards and targets, but we need to keep producing. More so in the current geopolitical climate.
And whilst nobody wants to risk being starved to submission, it's also equally important to promote more profitable sectors, and tax accordingly, so that we can support our more strategic sectors. I wouldn't say we're doing a good job at that for what its worth.
+1, exactly. Focusing too much in the money makes you forget about the power. At a national leadership level, there isn't much power in having a local Warhammer industry, fishing is much more strategic.
In broad terms, this is related to the error the USA made. Manufacturing in China was a very profitable deal for the USA. A lot of companies view labour first and foremost as expense, wealth as as the goal, and power in wealth— so it's not surprising as a whole the industry opted to "contract out" labour across the globe.
A lot of power lays in labour though. Money doesn't produce, invent, move, feed, etc— money is only good if someone will take it at the amounts you have it to do that specific labour you need for you.
This is a bit harsh: the USA didn't devalue labor in general, just manufacturing. They hired software engineers from all over the world, along with a lot of higher value engineering and product development jobs. The error the USA made was in pushing the workforce up the value chain faster than everyone could handle, and a lot of Americans got left behind.
China is moving up the value chain also, they are being forced to by their demographics, and they are investing heavily in the change ATM (just like they started investing heavily in green energy 10 years ago) so I don't think they will make the same mistake as the Americans are making right now.
SE will one day realize they are as screwed as the average worker and unionize in some way. The endless money pits are not for every SE, even today the majority does not work for the creme-de-creme of well paying companies. While specialists and smooth talkers might profit from the current model, as always is the case, most won't.
Unions are a bad solution to the problem of companies not vaulting their employees. Unfortunately, without altruism by the corporations or governments action, they are also the only effective solution.
I'd prefer we look to other countries that have solved the problem in other ways, such as including representatives of the employees in the board.
Also, we need to remember that corporations exist at the mercy of the state, having received a special dispensation (corporate charter) to exist. Those companies that are not a net benefit to society have no right to exist and ought to be dissolved.
Then when they unionize, all the software jobs will head to cheaper locales. Unless of course you’re suggesting that the U.S. use tariffs on labor to prevent that?
It's extortion. Sometimes the extorting party can act on its threats, sometimes they can't.
In regards to the auto industry, I'm not sure what you mean by "look where auto manufacturing investment in the US is". Most auto manufacturing jobs are still in the rust belt [1]. Most EV Production investment is in the rust belt [2].
If the goal is to have high paying jobs in the US, then yes the government should put penalties in place to encourage that. US cost of living (housing, real estate taxes, health care costs, college costs, etc) is way higher than many countries, especially those where jobs are being offshored to, especially India, so salaries have to be higher here.
So, do we let US companies invest US consumer derived revenue in the Indian economy, just to boost profits a bit, or do we protect good jobs at home instead, and have a virtuous circle where US profits get plowed back into the US economy?
Sure but when it comes to tech US currently (supposedly) has very high labor costs and weak labor laws. All other countries with a tiny number of exceptions have low to very low labor costs and more regulation.
I only see a push for AI to produce more developers. How could AI, as we know it today, even replace developers?
More developers isn't at odds with the previous comment. That is how you can more easily push the jobs to low cost areas! When 张三 in rural China is given his first computer he can jump right into being a programmer too. Thus you can give him the job instead of a high priced developer in America.
I want to build a thing. Before AI it would take 1 year and a team of five developers. Now with ai, it's gonna take you 6 months and 3 developers. Those two developers didn't get jobs because of AI.
You haven't even put developers out of job as you must remember that I also wanted to build a thing, but couldn't because you had all the available developers tied up. Now there are two freed up who can come work for me. There is no end to all the software we want to write.
That is the hope! Only time will tell if this tech is deflationary or inflationary though. You also want to build a thing, but do you have funding for it? in this economy?
> Only time will tell if this tech is deflationary or inflationary though.
Or both. That would be my bet. The industry in general will see a decline. The massive growth in developer numbers will place enormous supply-side pressure. But certain experts who remain supply constrained along with increasing demand for those special services amid the explosion of new software being written will make a killing.
I work for a US org that also hires internationally, including Europeans like me. It does location-based pay, so we're much cheaper than especially my Silicon Valley colleagues, yet somehow, we keep hiring there more than we are in Europe.
A lot of our “value chain” is bullshit. If and when China becomes twice as big an economy as the US, much of our “edge” in marketing, finance, and services isn’t going to mean squat. E.g. how much “GDP” will evaporate overnight when American universities no longer have the cachet that comes along with being the best universities in America?
No it isn't, productivity has to increase, that's why we constantly get rid of jobs that do not provide much value. People want more money, and the only way we get there is with more productivity (doing jobs that make more money).
> If and when China becomes twice as big an economy as the US, much of our “edge” in marketing, finance, and services isn’t going to mean squat.
China gets to 2X our GDP by doing what we basically did in the 90s, so you are definitely right! They will have their own marketing, financing and services. The only difference is that they won't need to outsource manufacturing to China (well, they are outsourcing it a bit now, but also investing tons in automation).
> E.g. how much “GDP” will evaporate overnight when American universities no longer have the cachet that comes along with being the best universities in America?
I don't know what you are ranting about, but I get the feeling that if I did know what you were saying here I would probably agree with you.
> No it isn't, productivity has to increase, that's why we constantly get rid of jobs that do not provide much value.
Our measures of “value” are wrong.
> I don't know what you are ranting about
My point is that a lot of what we think of as “higher value” activities are actually derivative of and downstream of our industrial supremacy. As China takes up that mantle, the higher value activities will go along with it. E.g., how long do we expect the US to do the cutting edge nuclear power and weapons research when China is the one building all the nuclear power plants?
I mean look at the path dependency that led to Silicon Valley. Why did the software revolution happen in the same place we were building the microchips?
We basically agree then: as China takes up higher value activities, they won't need those activities from the US anymore. Also, France is the cutting edge designer of nuclear power plants these days.
> Why did the software revolution happen in the same place we were building the microchips?
Hardware people becoming software people was extremely common back then, and still is today (EEs can make more coding than using their degree directly). Now we have the opposite problem (we don't have enough hardware people because software sucked all the oxygen out of the room) and China has less of it (although increasingly...they are repeating history as well). If anything, this just backs up my point in how higher value activities de-emphasize manufacturing (even super high end manufacturing as in semiconductors).
You can replace perceived value with actual value if you don't agree with the value judgement calls that were made, which is entirely reasonable.
"we don't have enough hardware people because software sucked all the oxygen out of the room"
How much is this exacerbated by the lack of domestic hardware manufacturing in the US for the hardware people? Seems like software boom starting in the late 1990's happened as China came online for hardware outsourcing. Not suggesting a causal relationship there, just complementary effects.
The same thing happens in China. If you want a good job, software will pay way more than hardware. Heck, you even see people from Taiwan doing software jobs in China because they at more than the hardware jobs in Taiwan that they could get.
That’s the biggest thing I took away from the whole Boeing corporate disaster
You need to maintain at least a minimum amount of internal competency in almost all areas
If you completely give away a capability to other countries (in this case, fishing knowledge and labour) it is much harder to bring back than just coughing up the money
Those sectors you let die might not matter right now, but they might matter later. And you might have to scale up fast.
We buy local brands of shoes that are in inr 300-2000 range and that solves like 70% of the shoes market in India. From shoes to skippers to formal shoes to ladies heels and such.
Then you gave INr 3000-8000 that are considered really really expensive.
Convert that to usd and you will see how much premium is being charged.
Perhaps. But does a subsidized industry retain competence, or retain incompetence? After all, if you're making a profit no matter what, what incentive is there to do well?
Many of the EU farming and fishing subsidies are to NOT produce anything.
That often depends on the structure of the subsidy.
"We will pay you 5 euros per kg of fish sold in supermarkets to consumers" is different from "We will pay you 500,000 euros a year to keep fishing".
There is a very reasonable argument in fisheries starting at least a century ago (and locally long before that), that we're looking at a partially renewable good - that it would be easy to cause an unsustainable population collapse with unrestricted harvesting, and so you should try and intervene in the market to sustain fish populations and stabilize harvests. Subsidies intended to do this are distinct from subsidies intended to keep fishermen employed fishing.
See also How Tech Loses Out over at Companies, Countries and Continents https://berthub.eu/articles/posts/how-tech-loses-out/, where the author asks, "In any organization, in any company, in any group, any country and even any continent, what level of technical capability, do we need to retain?"
Once you've outsourced everything except the management work, the organization forgets how to do the thing they're supposed to be managing.
> If you completely give away a capability to other countries (in this case, fishing knowledge and labour) it is much harder to bring back than just coughing up the money
I feel like money is overwhelmingly how we denominate value, effort, and agency in our society. Almost every time somebody says "You can't just throw money at the problem", they are arguing that we shouldn't even try that, contrary to all established reasoning about how society works.
There are diminishing returns to funding, but the people who use this expression are typically at a tiny fraction of where we would expect to hit them.
If you want to have a fishing industry because fish are your idealized heritage, then choose to subsidize it heavily either to continue to exist, and/or to expand it into waters and economies of scale where you can still fish profitably. Like the Japanese and the Chinese do, respectively.
You need to pay money to people who will put in effort and agency. You can't just throw money at random people and expect something useful to happen. Sometimes, the people who will make things happen if you throw money to them don't exist. Sometimes, you have to turn people into those people (which also costs money).
Money is something you give people so they can eat and stay warm while they do the thing you want. They still have to be doing the thing you want. Sometimes there's enough reputation and legal threats on the line that you can assume the person will do the thing just based on the fact they're taking money from you and not freaking out. Companies do things this way a lot - individuals not as much.
The abstraction is not the territory, and the idea that money denominates value is an abstraction... often we define "value" as "that for which money is exchanged", making the abstraction tautological, at no gain. This is often done by people who want to think the thing they're spending a lot of money on is very valuable, or want to make you think the thing you're spending a lot of money on is very valuable.
Also, continuing overfishing is a terrible long-term strategy. Sure, we will have the boats, fishermen and infrastructure around fishing, but that's of no help if the fish are gone.
You missed the "you can fix it with money though, and our friends are just the people to give the money to" step that comes after each problem statement.
Agreed, and the same goes for most strategic sectors: energy, agriculture, animal husbandry, semiconductors, communications, space, the infrastructure to support all these, education, etc.
It's only ~100 years since seafood was the primary protein source of most coastal regions. The rampant mismanagement of fish/shellfish stocks that put an end to that has had knock-on effects across our entire food supply, that continue to influence agricultural policy to this day
Which amid everything, maybe it is time to focus again on our own programming languages and OSes like in the cold war and export regulation days, it will suck for a while, but apparently it is how everything is going.
As a UK non-citizen resident, my observation is that our economy is pretty much involved with USA (and the global economy more generally).
I just looked up the latest trade and investments factsheet[1] and there are some interesting deets. If you're wondering about direct investment in the US as well as imports:
- Total UK imports from United States amounted to £111.5 billion in the four quarters to the end of Q3 2024 (a decrease of 5.1% or £5.9 billion in current prices, compared to the four quarters to the end of Q3 2023).
- In 2023, the outward stock of foreign direct investment (FDI) from the UK in United States was £494.1 billion accounting for 26.7% of the total UK outward FDI stock.
In addition to direct investment I would also count portfolio investment since we're sort of involved at an individual level through our workplace pensions (and/or personal), savings, stocks and shares ISAs, and so on. A preliminary report[2] foreign holdings of US securities as of June 2024 puts the UK as the top holder at over 3 trillion USD.
>Total UK imports from United States amounted to £111.5 billion in the four quarters to the end of Q3 2024 (a decrease of 5.1% or £5.9 billion in current prices, compared to the four quarters to the end of Q3 2023).
Also "Total UK exports to United States amounted to £182.6 billion in the four quarters to the end of Q3 2024 (a decrease
of 0.5% or £889 million in current prices, compared to the four quarters to the end of Q3 2023);"
Which is why the current government is being particularly cautious at the moment.
I do not speak latin, although I studied it for two years in high school and I'm a native speaker of a romance language, so my understanding of latin is pretty much basic to guesswork.
This is a really cool tool -- I often read latin texts with the original on one page and the translation on the other, just because I think it's interesting to see how they wrote/spoke at the time, but for the most part certain words or declinations throw me off guard. Inline literal translations really help there.
That being said, I noticed whilst reading some of the texts that the inline literal translations are still in latin, e.g. in "Part IV. I Some Barbarous Customs", most of the translated text is just latin. I guess OpenAI won't take all our jobs just yet!
I do have one suggestion for improvement though. Many of these texts have translations that are already in the public domain (older translations). It would be helpful to display the original Latin and a fluent English translation side by side, whilst still being able to toggle the literal translation on or off. This setup would make it easier to compare the original text with a fluent English translation, similar to the format used in some bilingual books.
Speaking of postgrest, it looks like the article links to `www.postgrest.org` which has been "hijacked"? The correct url should be https://postgrest.org
Sure but fear created by biosphere collapsing all around, massive human migrations, desertification and miscellaneous climate catastrophes are not things that mere terrorist groups can plot. :)
If you're going to treat logs as a db entry and afford them a schema, then by all means knock yourself out.
The main point he's trying to make is not just about the logs you wrote but about logs coming from other systems or services, e.g. monitoring kernel logs in your OS. As he rightly points out, one of the reasons is that logs are not an API.
> One reason this happens is that almost no one considers log messages to be an API, and so they feel free to change log messages at whim.
create table log (dt timestamp not null primary key, msg text not null)
is usually all you need to start sifting through the haystack. Problem with this approach is if you aren't diligent, you'll end up killing you app AND you're db.
We already established that log sifting at scale isn’t productive nor does it work for reasons outlined in the article. This was just an answer to how to do it in a smaller scale. Yeah, primary key is going to give you problems, a simple index shouldn’t but the issue is this, logs are not event sources - or shouldn’t be. The only time you should be looking is when your looking for stack trace and even then there’s better options like sentry.
To be fair a slightly better comparison would be MMS, but even then it doesn't compare. It's simply because data is cheap, the app is free, and the rich media features that came before iMessage/RCS existed (voice, photos/video, attachments, group chats, voip/video calling, etc), and because it's multiplatform (ios to android usage is near 50/50 or 60/40 depending on the country)
You do raise a good point re: tradeoffs in a healthy society. Mill anticipated this objection and addressed it directly. He didn't advocate for free speech without consequences but developed a harm principle specifically to establish what limits are acceptable. Acceptable limits on thought and speech should be based on demonstrable harm, rather than alleged offence, discomfort, or the current popular opinion or cultural disapproval.
He recognises the need to set some limits, yet also the dangers of who gets to set them. Historically, those with power to restrict speech have restricted truth as falsehood. The bar for restrictions should be very high, not because free speech should be absolute, but because the dangers of overzealous restrictions far outweigh the cost of permitting speech we might personally find objectionable. Even completely false opinions might have their value as they force defenders of truth to better articulate their position or reasoning, and prevents beliefs from becoming prejudiced, or platitudes.
I thought I'd share since it's relevant and there may be some younger readers here that might not have come across his work. I really recommend reading it, even if it's an LLM summary as an introduction (as seems to be the trend nowadays)
Edited to fix a few typos (typing from mobile)
1. https://www.econlib.org/library/Mill/mlLbty.html#book-reader