Someone once told me you need humongous vectors to encode nuance, but people are good at things computers are bad at, and vice-versa. I don't want nuance from computers any more than I want instant, precise floating point calculations from people.
I think you are missing the difference between a program derived from training data and logic explicitly created. Go ahead and proceed to continue doing what you are doing for all words in the dictionary and see how the implementation goes.
It depends whether you want your system to handle all of natural language and give answers which are correct most of the time (but it isn't easy to tell when it's wrong), or to handle a limited subset of natural language and either give answers which are demonstrably correct (once it's fully debugged or proven correct), or tells you when it doesn't know the answer.
These are two opposing approaches to AI. Rule induction is somewhere in between - you use training data and it outputs (usually probabilistic) human-readable rules.
I also suspect using 2020 as a baseline will inflate how large the increases appear. Cursory google, but this dataset suggests pretty steady number apart from the covid downturn and now the recent crackdown:
> Suicidal deaths from paracetamol and salicylates were reduced by 22% (95% confidence interval 11% to 32%) in the year after the change in legislation on 16 September 1998, and this reduction persisted in the next two years. Liver unit admissions and liver transplants for paracetamol induced hepatotoxicity were reduced by around 30% in the four years after the legislation.
(This was posted here on HN in the thread on the new paracetamol in utero study that I can't seem to dig up right now)
I did not say the author is not advocating for friction, I said the authors wording about the tech world wanting everything to be fictionless is wrong (as I think I demonstrated with my example of customer support).
To quote the author:
> In tech circles friction is seen as bad, everything needs to be frictionless. Every interaction with anything needs to be smooth and uninterrupted.
This is the claim I refuted. This is what the tech circles like to appear like, all while using friction when it comes to their AGBs, when it comes to rejecting their tracking cookies, when it comes to opting out or unsubscribing, you get the idea. Don't want a path to be taken often? Add friction.
These are all examples for neferious use of friction, but in itself there is nothing bad about friction. E.g. the "Format" button on a Blackmagic Pocket Cinema Camera needs to be hold for three seconds (animating as you hold it). This added friction gives you enough time to reconsider whether you really want to erase all those videos you recorded.
How is public transport to nearby commuter towns? If it's affordable/convenient/reasonably quick that can help a lot, I suspect. (But am quite ignorant so appreciate correction!)
Very good, generally. It's affordable by Swiss standards, which is to say expensive for tourists but remarkably good value if you're a resident on a median salary. The cost of the whole-country annual rail pass costs less than a point to point season ticket in the UK, and there is no peak travel. So you should never pay more than 3-4k a year. Most people just get a half price card (often employer subsidised) and/or a local pass which is generally better value unless you're commuting between cities.
Switzerland has the best rail system in the world (IMHO). One app can manage ticketing across all public transit in the entire country, and it is extremely fast and reliable. Lots of people who work in Zürich, for example, commute in from Zug (due to lower taxes in the neighboring canton).
It's generally called "as fast as necessary", more focused on total journey lenght with the clock-face schedule across the whole country. Other countries certainly have faster trains.
I guess "fast" is what I mean when I think "on time, efficient" (compared to, say, the struggles Deutsche Bahn has had in recent years). There are no bullet trains laterally crossing Switzerland in under an hour, I suppose.
it's nit-picky, but I think the swiss approach of "as fast as necessary, not as fast as possible" is worth separation from "train go fast", especially since it's mostly the later that shows up in comparisions/statistics because it's much easier to calculate.
Thanks for teaching me about this! I listened to it and immediately thought I recognized it from Jamie XX's Oh my Gosh. In the event I was wrong, it samples Lyn Collins -- another classic sample I believe, certainly in hip hop. But I'm sure I'll hear it pop up in lots of places now, always nice to learn a piece of history.
Berlin - Germany + France = Paris , that sort of thing