OK, I read the article, wondered a bit, then started to read the responses (first aligning with the one "April Fools' in December?"). I mean, what's the point of asking people to jump from airplanes on the ground, write a study on it, and waste everybody's time
But then, reading other responses, and taking into account the name of the journal, and the fact that it did get published, I realized it's a very good example of a flawed study, where everybody see a flaw because it's evident, whereas in other studies a flaw of the same magnitude might be quite difficult to see (I could give quite a few examples from my field).
They are free to be. When joining the EU, Poland accepted to join the monetary union at some point, but there is no date set, so in theory it can be infinite. And with its current strategy to drastically modernize and increase its military force, there is no doubt it will not met the financial criteria anytime soon so nobody is worried about that.
Poland will not adopt it in foreseeable future because it makes no sense: the Polish currency (zloty) acts as a kind of bumper in difficult times. It's very simple: something bad happens like a crisis or a war, investors immediately sell assets deemed as risky (such as zloty), zloty goes down by x%, and automatically Polish products and services become more attractive by x%. This all happens by itself. So for the Polish economy it's a no brainer.
For people themselves, they're not so keen either as they know well adopting euro was abused in other countries to increase prices (and not just round them up a bit). So basically nobody wants it, even if there are a few real benefits of joining, such as potentially lower mortgage rates which are among the highest in the EU).
> Now I have to rework my site to not use "prompt engineering" and have a Take™ on "context engineering". Because of a couple tweets + a blog reverberating over 2-3 days.
The difference here is that your example shows a trivial statement and a change period of 3 days, whereas what Carmack is doing is taking years.
Right. Nothing against Carmack. Grew up on the guy. I haven't looked into, at all, into any of the disputed stuff, and should actively proclaim I'm a yuge Carmack fanboy.
Which ones, exactly? I heard this phrase tossed around but on close examination it always turns out it something related to protecting the citizen. Which I believe, is a conscious choice on this side of the ocean.
The Hanoi Towers example demonstrates that SOTA RLMs struggle with tasks a pre-schooler solves.
The implication here is that they excel at things that occur very often and are bad at novelty. This is good for individuals (by using RLMs I can quickly learn about many other aspects of human body of knowledge in a way impossible/inefficient with traditional methods) but they are bad at innovation. Which, honestly, is not necessarily bad: we can offload lower-level tasks[0] to RLMs and pursue innovation as humans.
[0] Usual caveats apply: with time, the population of people actually good at these low-level tasks will diminish, just as we have very few Assembler programmers for Intel/AMD processors.
You might be right to some extend, but not entirely. For example, there have been almost no incidents in AWS where one customer would be able to access the data of another customer because of AWS fault. The cases so far like Superglue etc. were very limited and IMHO AWS security is quite solid.
So I would say there is a difference between AWS architects and engineers (although I know first hand that certain things are subobtimal, but...) and those of several other companies who have less customers but experienced successful attacks (or data loss). Even if you take Microsoft, there is huge difference in security posture between AWS and Azure (and I say this as a big fan of the so-called "private cloud" (previously know as just your own infra)).
And yet, we haven't found a good solution to that yet, and everybody uses smartphones controlled by Apple and Google, communicate via channels controlled by Meta and so on.
reply