Me too, but I don’t like referring to Dunning-Kruger ever for multiple reasons. There are perfectly good labels like cockiness, arrogance, ignorance, presumptuousness, and wrongheaded. ;)
There are many issues with DK, and the paper’s widely misunderstood. For one, the primary figure demonstrates a positive correlation between confidence and competence, so according to DK’s own paper, high confidence is not an indicator of incompetence, contrary to popular belief. The paper also measured things in a very funny way (by having participants rank themselves against other people of unknown skill), and it measured only very simple things (like basic grammar, and ability to get a joke), and it only polled Cornell undergrads (no truly incompetent people), and there were a tiny number of participants receiving extra credit (might exclude the As and Fs in the class). Many smart people have come to the conclusion that DK is a statistical artifact of the way they did their experiment, not a real cognitive bias. Some smart people have pointed out that DK is probably popular because it’s really tempting to believe - we like the idea of arrogant people getting justice. The paper also primes the reader, telling them what to believe even though the title isn’t truly supported by the data. It’s an interesting read that I think would not pass today’s publication criteria.
I looked up persuade and convince in the thesaurus and dictionary, based on the title, and then came to say the same thing. But then I got a little curious about the source of the title’s claim, and looked up Chaim Perelman. He really did try to make a distinction between convince and persuade in his influential book from sixty years ago, so the body of the blog post is accurate in a sense - this is a concept that came from an historically important philosopher. Perelman was dissecting argumentation and cataloguing the techniques for strong and persuasive arguments. The problem with this blog post is taking Perelman’s argument out of context and stating Perelman’s rhetorical distinction as though it’s a fact and then arguing logically for it. That leaves out all the ethos and pathos that Perelman was trying to convey is necessary for a good argument, and it also misses slightly on the logos as well.
Interesting. The post would have benefitted a lot from talking about this background instead of just name-dropping Perelman once and by last name only(!!).
That's the sort of sloppiness you get when you have a conversation with an AI, ask the AI to make a blog post based on the conversation, and then copy-paste that straight into your Substack without reading to see if a fresh reader would understand what you are talking about.
If the author insists on posting more unedited AI text, asking a fresh AI session to critique the post from scratch would probably catch this kind of mistake and lead to a much better result.
New multifamily construction in the US that has to undergo design review is arguably fairly custom in that each site will have different requirements. I think it's fair to say that commoditization is a spectrum?
The structure of most residential construction in the US is standardized. Foundation (or slab), wood framing, etc. There are different levels of quality, but codes and standards mean that standardization is the norm.
> I would bet on the mobility impaired people to win the cage match
Why frame it as a fight? There’s no need to start there; you don’t need to waste time fighting against people not in your group. You just need to establish group status. If the constituency of obese people is strong, why not seek to establish policy on behalf of obese people and not everyone? As the article and others here have said, reducing traffic congestion benefits everyone in multiple ways, including benefits for the people who still have to drive. Given a choice that doesn’t affect your ability to drive, I assume you’d rather have less pollution, less noise, and fewer other drivers on the road?
The other angle missing from your comment is e-bikes. Most of those ~42% of obese people in the U.S. are still capable of riding an e-bike, and for short trips in busy areas, e-bikes are more convenient and easier to park than cars.
You’re right for some workloads, but not all of them. The same could have been said for disk swap since the beginning though, and people still found it valuable. Disk swapping with spinning drives did used to be multiple orders of magnitude slower than RAM. But it prevented applications or the system from crashing.
Using system memory from the GPU isn’t that bad if your compute is high enough and you don’t transfer that much data. There are commercial applications that support it and only see low 2-digit percentage perf impact and not the multiples you might expect. Plus on Windows on Nvidia hardware, the driver will automatically use system memory if you oversubscribe VRAM, and I believe this was introduced to support running Stable Diffusion on smaller GPUs.
Try turning swap off and really find out if you’re not grateful for it. Might be fine if you’re never using all your RAM, but if you are, swap off isn’t fun and you might realize you’ve been unconsciously grateful this whole time. ;) Swap might be important for GPU usage even when not using something like greenboost, since display GPUs sometimes use system RAM to back the GPU VRAM.
> Try turning swap off and really find out if you’re not grateful
Er, I did exactly this over a decade ago and never looked back. It's literally one of the first things I do on a new machine.
> Might be fine if you’re never using all your RAM
That's definitely happened occasionally, and no, swap almost always just makes it worse. The thrashing makes the entire machine unusable instead of making the allocating app(s) potentially unstable. I've recovered most times by just immediately killing the app I'm using. And in fact I have warnings that sometimes tell me fast enough before I reach the limit to avoid such issues in the first place.
> you're buying a vehicle that already has the capabilities, but are disabled, then paying rent (or a fee) to turn them on. I'm much more likely to buy from a manufacturer that doesn't play these games.
Ongoing subscriptions for access to physical hardware features like seat warmers* seems obnoxious at first glance, but a fee is more reasonable and you might find that there aren’t many auto makers that don’t do this or aren’t planning on it. BTW there’s very little in software or electronics that doesn’t do this, and many other consumer products do too. What might be less visible is how often the hardware is included and made trivial for a dealer to upgrade but doesn’t have a remote software unlock. It’s the same thing and it’s been happening for decades, but gets less outrage.
You would have paid a fee for the feature if it wasn’t included. Focusing on features being there already and locked being somehow “bamboozles” isn’t necessarily the right way to frame this, even from a pro-consumer perspective. This practice of building the high end model and locking some features behind a paywall makes the design and manufacturing cheaper for everyone by having only one design. The paywall model suggests that the design costs are more important than the manufacturing or materials costs of these features. That’s absolutely true for software apps, and it’s accepted by and large and we don’t feel like that’s a skeezy game. It doesn’t surprise me at all that with manufacturing at a global scale, it makes more sense to build one model and lock some features with software.
Do think of the potential benefits we get from this model - overall lower prices (in theory) from simplified design and manufacturing; the ability to upgrade later after you buy (or even downgrade if you don’t like it and it’s a subscription).
* AFAIK the BMW seat warmers subscription was a rumor at one point, got a bunch of online uproar, but didn’t actually happen? I’m not sure if anyone has actually done this.
Manufacturing one hardware setup and charging separately for features is not the problem. The problem is charging ongoing rent for a feature that isn't an ongoing service. A seat heater doesn't use a server, need content updates, or create meaningful recurring costs for the manufacturer after the car is sold. It shifts the relationship from ownership to permission. It also creates bad incentives: features can be removed later, tied to accounts, complicated for second owners, or turned into endless monetization opportunities.
I agree with that. I don’t know what your prompt was, but I wasn’t arguing in favor of subscription access to hardware, I said flat upfront fee based upgrades make more sense, and I was only pointing out that market segmentation over a single physical product via software feature locks is a pretty common thing and it’s not necessarily a bad thing for consumers, there are some side benefits, some tradeoffs.
I’m not personally into paying subscription upgrades, I tend to avoid them. But the one case where I could see potential for consumer benefit is when there’s a choice between a high upfront fee or a low subscription price. I would assume a subscription price over time will cost more than the upfront fee. However, there’s an argument to be made for lower cost access, for smaller barrier to entry for the upgrade, especially if it can be discontinued if the customer doesn’t find enough value.
There was a motorcycle airbag jacket that offered this choice and was discussed on HN maybe a year or two ago. People were, of course, freaking out about a safety feature being tied to a subscription, and I can totally understand the fear, but the rhetoric around it didn’t match what the actual product offered, and the company was offering the choice between flat fee and monthly fee, not demanding a rent-seeking only option. Personally I think most of the ick feeling of a subscription idea goes away for me if it’s not the only option.
> codegen changes the cost structure of writing code, not the cost structure of knowing what to write.
Yes, and knowing what to write has always been the more important challenge, long before AI. But - one thing I’ve noticed is that in some cases, LLMs can help me try out and iterate on more concepts and design ideas than I was doing before. I can try out the thing I thought was going to work and then see the downsides I didn’t anticipate, and then fix it or tear it down and try something else. That was always possible, but when using LLMs this cycle feels much easier and like it’s happening much faster and going through more rough draft iterations than what I used to do. I’m trying more ideas than I would have otherwise, and it feels like it’s leading in many cases to a stronger foundation on which to take the draft through review to production. It’s far more reviewing and testing than before, but I guess in short, there might be an important component of the speed of writing code that feeds into figuring out what to write; yes we should absolutely focus entirely on priorities, requirements, and quality, but we also shouldn’t underestimate the impact that iteration speed can have on those goals.
Yes. I'll go down a wrong path in 20 minutes that'd have taken me half a day to go down by hand, and I keep having to remind myself that code is cheap now (and the robot doesn't get tired) so it's best to throw it away and spend 10 more minutes and get it right.
This got me through many of the first 100 problems on Project Euler:
n = 1000000 # must be even
sieve = [True] * (n/2)
for i in range(3,int(n**0.5)+1,2):
if sieve[i/2]: sieve[i*i/2::i] = [False] * ((n-i*i-1)/(2*i)+1)
…
# x is prime if x%2 and sieve[x/2]
Edit: I guess I irked someone. :/ Yes this is a memory hog, but to me beautiful because it’s so tiny and simple. I never tried very hard, but I wonder if it could be made a real one-liner.
There are many issues with DK, and the paper’s widely misunderstood. For one, the primary figure demonstrates a positive correlation between confidence and competence, so according to DK’s own paper, high confidence is not an indicator of incompetence, contrary to popular belief. The paper also measured things in a very funny way (by having participants rank themselves against other people of unknown skill), and it measured only very simple things (like basic grammar, and ability to get a joke), and it only polled Cornell undergrads (no truly incompetent people), and there were a tiny number of participants receiving extra credit (might exclude the As and Fs in the class). Many smart people have come to the conclusion that DK is a statistical artifact of the way they did their experiment, not a real cognitive bias. Some smart people have pointed out that DK is probably popular because it’s really tempting to believe - we like the idea of arrogant people getting justice. The paper also primes the reader, telling them what to believe even though the title isn’t truly supported by the data. It’s an interesting read that I think would not pass today’s publication criteria.
Anyway, sorry, slash rant.
reply