spupe

From my experience with similar projects, it doesn't get any simpler than creating a virtual environment, running requirements.txt and using a simple function to get what you want. Did you have a problem when you tried running that? Colab in this case is just abstracting that part for the user.

So their proposal is to leverage large language models to propose automated, reasonable changes to a codebase that is constantly checked for "fitness" in a particular task.

I've always been fascinated with evolutionary mechanisms in programming, this is a clever idea to make things more interesting.

Ok, thank you for the clarification.

I see your point, although "as good as Google" qualifies as "enterprise-level" in my book.

The ICOs were the OG NFT's. The whole crypto ecosystem of today is definitely wilder though.

I think they will have a very hard time recruiting in Germany. There is certainly no patience for this permanent crunch, anti-union attitude that seems characteristic of Tesla in the US.

Their Colab examples work pretty much like that.

Can you elaborate further? I am not familiar with the field, but their benchmarks here seem to show quality similar to Google: https://github.com/snakers4/silero-models/wiki/Quality-Bench...

The only trick I can see being played is that Google was benchmarked on September 2020, so likely has already improved and they don't want to show that. Is CommonVoice a better standard to use when comparing these tools?

My energy usage estimate is also based on daily use. And even if we extrapolate it to a 10x increase, 55000 GwH, and it could be available now, we would not even achieve the same levels that nuclear already does just in the US, today. Every discussion we have on debating batteries as alternative to nuclear must tackle this obvious issue.

I feel sad when I see replies like yours, because honestly I don't think you understood my point at all. If you combine all the other things you mention here, all of them together, right now, are incapable of addressing even 1% of our energy use, let alone be the backbone of a grid to manage intermittence. The cost analysis you describe is imaginary, because again, those technologies simply don't exist at scale. There is not a single country in the world that operates with any of your strategies, but most developed economies regularly use nuclear in their mix.

We could address climate change with nuclear today, or we could wait until these technologies are mature and cheap, which may come too late.

Yeah, it's a great idea. But the problem is the specifics. For example, the electricity storage you describe, do you have any idea of its cost or feasibility? For example, all the battery capacity in the world right now is about 1000 GwH, and optimistic estimates project a 5x growth by 2030 [1]. Even if we take the 2030 figure, 5500 GwH, that number would be enough to power... Less than 500 thousand American people, if my math is correct? [2] Meanwhile, nuclear is currently responsible for 20% of electricity generation in the US, and could be even more, there's no fundamental obstacle in the way except money.

[1] https://www.reuters.com/business/autos-transportation/global...

[2] https://data.worldbank.org/indicator/EG.USE.ELEC.KH.PC?locat...

Most other energy sources are cheaper - at least upfront. For example, coal and gas are much cheaper, but they cause huge health and climate damage. Solar and wind on the other hand are clean, but they are intermittent, so in bad days you might have zero energy being produced from them.

We will always need a reliable source of energy that is not intermittent, and considering all the other alternatives nuclear is the best one for this task if hydropower is not feasible.

I had the opposite experience, I was disappointed to see how some famous cities (Copenhagen and Rio de Janeiro) look really bad with generic buildings. Maybe it's a US vs rest of the world thing, or am I missing any setting here?

This is an ad.

I was expecting a more objective evaluation of fact-checkers (ie, how often do they actually get things right or not). The reader is treated to a couple of anecdotes about egregious mistakes, and the implied conclusion is that no one is interested in improving fact-checking because it is merely a political tool. Yet the author dismisses the one objective study he cites about bias in fact-checking (from 2011 (!), which found the vast majority of statements classified as false by Politico were by Republicans [1]) because, well, who knows whether they were actually false or whether Politico is just biased.

It is not impossible to have objective fact-checking, and it would be very beneficial to start having that discussion right now. No one is perfect and it is difficult to handle your personal biases, but I could live without this attitude of "the other side also does it, therefore it's all 'political', so it's a wash".

[1] https://smartpolitics.lib.umn.edu/2011/02/10/selection-bias-...

Well, as the article itself describes, there are many decent materialistic explanations for this problem - I personally am fond of Daniel Dennett's interpretation of it.

Bitcoin at 12k in 2017? Really?

You should welcome all the nay saying here if you think Bitcoin is bound to go up again.

Could you elaborate on what phenomena are difficult to explain through materialism?

> By the way, please don't use the New York Times as a source for anything. It's both paywalled and notoriously biased.

I could give you a number of other credible sources to verify that hospitals were in fact exceptionally strained at several points last year, a verifiable and true statement that you do not believe in. You are welcome to provide a source for your claims, I'm sure it would come from a very good, unbiased source.

> The jury went home a year ago dude, you just can't accept what they found. There is no correlation between severity of countermeasures and outcomes (when looking at countries with reliable data). This has been examined six ways from Sunday and the results are always the same - no correlation

You are wrong again. To just focus on one example, here is a paper published in Science last year, suggesting various distancing measures produced significant reductions in transmission of COVID [1]. I would advise you to follow your own advice and have a look at the evidence, because in this short discussion you were proven factually wrong twice already.

[1] https://www.science.org/doi/10.1126/science.abd9338

> You're still scrabbling about to try and hide this with cherry picking. There's no scientifically valid reason to only compare Sweden with Norway, or Portugal, or Italy. That's why the blog post shows it relative to all European countries (and they could have included the USA, Canada etc too).

I'm sorry but this is my main point, and it is crystal clear. I'm not sure why it's eluding you. No, you cannot compare highly disparate countries like that. Are you willing to compare China and Vietnam with Sweden too? Or maybe we should say that Norway, which did impose a lockdown, was the model to be followed in Europe? Cherry-picking is taking data you know is not appropriate and using it, I'm arguing for the opposite here. There is of course subjectivity involved, but you absolutely cannot claim that an imagined "default" (ie, all European countries) is unbiased or automatically valid. The best comparison for Sweden for most metrics, not only COVID, is among Scandinavian countries.

> Hospitals are always approaching 100% utilization because ICU capacity is flexible (it's partly a designation related to staffing levels), and having unused capacity is wasteful. Again, this is a talking point that was spun to try and justify the catastrophic policies, but they aren't valid. Hospitals weren't actually overloaded and especially not in Sweden which didn't bother with most of the mitigations.

You are absolutely wrong in your claims about hospitals. A non-exhaustive list of countries where the healthcare system neared collapse at rates much different than baseline: [1][2][3][4]. Your point about Sweden is literally what I wrote before, which is that Sweden did not face this problem, and therefore never had the same issues as other countries that issued lockdowns.

The jury is still out about the effects of the lockdowns. In order to find out, we have to be intellectually honest, which in my opinion is the opposite of what the author of this piece is doing.

[1] https://www.nytimes.com/2020/03/12/world/europe/12italy-coro... [2] https://www.nytimes.com/2021/02/02/world/europe/portugal-cor... [3] https://www.thelocal.es/20200329/where-in-spain-hospitals-ar... [4] https://www.nytimes.com/2021/03/27/world/americas/virus-braz...

It's not as simple as that. First of all, why pick only 2018, but not 2019, or in fact an average of previous years? Second problem is that 2020 was an exceptional year. People contacted each other less, so we expect them to die less from many other causes (traffic accidents, other transmissible diseases, etc).

> 2014 travel between NYC and Chicago is only barely faster than it was over 80 years ago.

Your issue is twofold: you are focusing solely on one metric to measure efficiency, arguably the worst one (speed), and you then further narrow the scope by saying efficiency gains must come from one dimension of the business ("aeronautical" advances).

There are obvious reasons for speed to not increase further, namely the speed of sound as you already mentioned, so this is a non-starter. We are simply not trying to make commercial planes faster anymore because that would be useless.

Regarding the second point, I think it is remarkable that you present data that directly contradicts your hypothesis and fail to realize that. If we are starting more flights, carrying more people and using the same fuel, the precise definition of it is an efficiency increase. To achieve that we had improvements in booking, avionics, crew training, business models (Ryanair and other low-costs), etc. The fact that we could make a few flights at incredible cost 80 years ago does not mean that we routinely achieving far safer, cheaper and frequent flights at the same speed today is stagnation, rather the opposite in my view.

> Comparing with the "peers" you see fit is just as much cherrypicking as comparing with other European countries.

I disagree. If you imagine almost any variable - wealth, education, weather, socialization patterns -, Sweden is vastly more different from Italy than from Finland. Any sensible analysis has to take into account what a valid comparison really is.

> It's not misleading. Sweden's "peers" are all rich western countries, because there's nothing different about the Nordics that a virus would care about.

Sweden's GDP per capita is more than double of Portugal and 40% bigger than Italy, two countries prominently noted in his analysis. There are two huge factors here other than wealth that are vastly different between Scandinavian countries and the rest of Europe: socialization patterns and weather, with both tending to make the virus way less likely to spread in the North.

> Epidemiologists and their models claimed such thresholds would be reached but those models were wrong about everything. It never actually happened.

Italy's healthcare system was overwhelmed early on in the epidemic, causing a ton of deaths, and hospitals all over the world approached 100% ICU capacity at several points in the last two years. I don't know where you pulled your numbers from, but they are wrong, and so is your conclusion.

>> "It's simply incorrect to use the number of deaths as the single measure of effectiveness of lockdowns"

> Reducing that one number was the entire justification for lockdowns! Now it's "simply incorrect" to observe that there is no link between them. Absurd.

It's basic logic: if we had to use lockdowns in the most extreme cases where hospitals were reaching capacity, then these countries would have experienced a higher death rate regardless of what they did. The question is whether the lockdowns had any effect on easing the burdens in the hospitals or not, but you cannot compare Portugal with Sweden and say "oh hey, more people died in Portugal and they did a lockdown, therefore lockdowns are bad/useless".

Very misleading analysis here. First of all, the obvious: Sweden should not be compared with an average European country, but with its peers. It is richer and has a much different weather and population density compared to all the big countries which the author incessantly compares it to. And in fact, if you compare Sweden with its Scandinavian peers, it fared much worse. In the only part where he addresses it, he says this:

> Although the determination of pro-mandate fanatics to exclusively compare Sweden to other Scandinavian countries is nonsensical, the 2021 excess mortality rates show Sweden with lower numbers than both Finland and Denmark.

Well, when a bigger proportion of your vulnerable population died in 2020, it's not surprising that in the subsequent year you see a decrease.

The author also seems confused about the causes for the lockdowns in all the other countries. Unlike Sweden, most of them reached a threshold where further spread of the disease would mean hospitals exceed capacity. It's simply incorrect to use the number of deaths as the single measure of effectiveness of lockdowns, as the true counterfactual would be countries that did no lockdowns and also had their healthcare system collapse.

You wrote above:

> Gordon's point is that the best commercial-grade service ... didn't move much in the seventy-eight years* from 1936--2014. It's been nearly another decade and that story's largely the same.

The fact that total passenger miles has been largely flat since 2000 (is that even accurate? [1]) has no bearing on that claim. Furthermore, the number of passengers has basically tripled in the last 20 years alone [2], at the cost of some comfort certainly, but this also must count as efficiency gains. I mean, if we ignore an order of magnitude reduction in cost and increase in number of passengers, yeah, I guess the industry is the same since the Wright Brothers.

[1] https://www.statista.com/statistics/185744/us-passenger-mile... [2] https://www.statista.com/statistics/564717/airline-industry-...

Flights were five times more expensive in the period you refer to [1]. I would bet that the number of fliers increased almost proportionally. A significant part of this cost reduction is efficiency gains, with better, lighter airplanes that consume less fuel. Airplanes are also safer today. To call all that "no appreciable movement" is unfair.

[1] https://www.fastcompany.com/3022215/what-it-was-really-like-...

There were several other mistakes there, but the most annoying to me was this hand-waving between correlation and causation. Sure, crypto boomed during a period of intense QE. There were also unprecedented cash payments for everyone and a big supply shock. I think the author's analysis of the relationship between inflation and BTC was equally dumb, simply noticing an inverse correlation.

I'm totally not a crypto apologist btw, just pointing out inaccuracies.

I'm usually on the fence between gross incompetence and straight out deception as the cause for these proposals, but in this case I'm convinced of the latter.

It's so primitive that laws can be proposed without any proper numeric estimate of their possible impact. How many Europeans would be ok with a draconian surveillance mechanism if they knew we would be preventing only a handful of abuses?

> It seems that globally, the success in these countries is bought from weakening the currencies in poorer countries.

Perhaps, but the period since 1914 saw the rise of many peripherical countries, notably China, who also have a tight control over their monetary policy, and use it all the time for their economic plans. For your counterfactual to make sense, you would have to present an example of a country that developed using the gold standard.

> But then again, are the central banks even following practices based on economics, or did they become politicized and corrupted, because of their central position in the economy? We see blatant monetary corruption happening in autocracies, but is it happening in western economies too, but it's just less obvious?

There is corruption and inefficiency at every level, but I fail to see how that would change by going back to the gold standard or cryptocurrencies for that matter.

I am familiar with the Austrian line of argument, and I'm sure in your years of study you have come across many criticisms of it, so I won't bother, but let's just say I don't think this perspective has brought us tangible benefits at all. Meanwhile, the imperfect system we have has shown itself to be at least robust to major hiccups recently.