jprival

I think Scheme works well for the kind of conceptual overview the course is trying to provide. I think there is something to the argument that Scheme syntax is not ideal for readability of larger programs, but I would wager that the bigger reason some students find SICP confusing is the same reason it blows others’ minds - the whole approach is at a higher level of abstraction than most “intro to programming” classes.

> Paracetamol is an anticholinergic isn’t it?

No, per https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/j.15...

Can’t say whether it’s good for your brain otherwise.

because of his sort of ascetic impulses in other domains one doesn’t imagine him as particularly hard-living, but this 2002 ilxor thread (bumped for his death) suggests that he had his first heart attack way back: https://www.ilxor.com/ILX/ThreadSelectedControllerServlet?bo...

edit: viral pericarditis, actually, before it was a hot topic

It’s not a particularly fair accounting of the framing of the article, which also profiles people who are working precisely on making the technology more practical and portable and ends on a hopeful note. For the time being it’s a high maintenance way to keep people alive, though, so the ethical dilemmas of resource allocation are real.

To look is to train one’s eyes on, or to scan for, something. To see is to perceive it.

So one can look without seeing, and one can also see something without intentionally looking.

OECD definition

> any chemical with at least a perfluorinated methyl group (–CF3) or a perfluorinated methylene group (–CF2–) is a PFAS

I don’t think that includes any of those, and Xanax is not fluorinated at all. You might have been thinking of Prozac, which I believe does meet those criteria.

It is a broad category with competing definitions, some of which are even broader, though, yes.

Also sometimes they do make really painful and arbitrary UI/UX changes - don’t get me started on the iOS podcast app. Personally I’m all for preserving keyboard shortcuts because that’s something I really internalize but it would be hard to argue that they are consistent taking any kind of broader view.

> enter doesn't open the folder

That’s a little like complaining that Alt-F4 is unintuitive, isn’t it? It’s historical continuity. On the other hand, they changed Command-N…

> there is only a back button and not a go up directory button

Can’t argue with that one, though.

> This is despite the arrival of a boom in funding due to the generative model bubble.

Does the boom in funding for AI companies make up for the overall drop in funding from late 10s-early 20s heights? That would not be my guess offhand, which makes playing up the contrast between “plenty of funding” (for one specific thing) and “job cuts” feel a little overstated.

It does complicate predictions, though.

Reminds me tangentially of the study that found that strong chess players were much worse at recalling randomized positions than realistic ones: https://link.springer.com/content/pdf/10.3758/BF03200937.pdf

I’m wary of inviting the too-easy “it’s just like how it works for people!” comparison, but the implied context and history of a game state seems to be important in processing it.

The appeal of prompting a whole song into existence kind of baffles me for the same reasons you are suggesting but there’s room for fun and creativity in working with samples and given the substantial cottage industry that royalty-free samples already are this was pretty much the first application I was expecting to see in the audio GenAI realm.

>Five years ago people were in agreement that the turing test was a really good test

The limitations of the Turing test have been discussed for much longer than that. Ironically its dependence on human judgement is arguably a weak point.

“Functional style” doesn’t offer the same guarantees, sure, but I wouldn’t underestimate the structural and readability benefits it can offer for certain kinds of tasks.

It’s a pretty good bet that quaternary ammonium compounds are used to wipe many surfaces wherever you are. They are probably the most widely used non-chlorine disinfectants.

There’s a graph in this very article that answers that question. Heroin deaths have declined significantly as fentanyl has displaced heroin on the market, but fentanyl deaths are far higher than heroin deaths ever were, because it’s far easier to OD on.

Worth noting that glucose meters are OTC. Continuous monitoring solutions generally seem to require a prescription, though there are telehealth companies that will help you sort that out. According to the article what the FDA really objects to is “non-invasive” glucose monitoring devices of any sort, on the grounds of general lack of evidence of accuracy.

I recall certain intro CS courses being calibrated to be difficult enough to fit with the department’s (generally tough) grading guidelines without actually applying a strict curve to exam results after the fact, so as to encourage students to see it as a collaborative endeavor rather than a competitive one. That seems kind of ideal if you can make it work. But yes, it’s hardly unusual for grades to work a lot like stack ranking.

I see a lot of conflation of “companies are shifting their product focus to AI” - which everybody knows everybody is doing - and “companies are replacing staff with AI.” To be fair the latter aligns nicely with the “bloat-shedding” rationale for layoffs, which also feels like a very real trend, but so many of these articles put companies like Google “pivoting to [selling] AI” in the same breath.

I thought “Bard” was an Asimov reference: https://en.m.wikipedia.org/wiki/Someday_(short_story)

(on top of the more obvious references)

About 420 ppm, an increase of about 120 ppm in the past 100 years? This is a strange question since it is obvious that it’s heavily “looked into” and easy to look up, and simultaneously hard to figure out what rhetorical point is being made in asking people to cite it out of context.

Yeah the Power Macintosh line, as the name suggests, introduced PowerPC in 1994. There were Performa versions of those as well. Jobs’ return as CEO roughly coincided with the beginning of the G3 era (calling them that instead of PowerPC 740 is already pretty Jobs-ian).

“rips” -> “kilorips” etc.

> What are the benefits these chemicals have provided

They have a lot of properties that can be useful in materials - hydrophobicity, lipophobicity, chemical resistance, low friction, etc.

They are certainly used in some niches (like medical devices, protective clothing, advanced manufacturing) that people would agree are Important, but a significant percentage of production goes into, and a significant amount of contamination comes out of, stuff like stain-proofing couches and making food packaging grease resistant. I’m under the impression that a lot of environmental contamination likely also comes from their use in firefighting foams - a useful application to be sure, but there are some things it’s not a great idea to spray straight into the environment.

> Are they chemically bonded to something that reduces bio-interactivity?

A huge part of the story of PFAS is that they are quite resistant to (permanent) chemical bonding, which has often lead to them being thought of as “inert” in a sense that is conflated with safety. But this sense of inertness is a bit of a red herring when it comes to biology (plenty of things interact with receptors without covalently bonding) and meanwhile makes them highly persistent both within organisms and in the environment.

“PFAS” as a literal chemical category feels pretty broad. The ones best established to be harmful are fluorosurfactants like the whole list explicitly targeted here. But broadness is also kind of the point because there are tons of chemicals in use that are likely to have similar accumulative properties that have never really been studied for health effects, and which can’t really be assumed to be harmless if the others aren’t.

I’m just used to seeing annual flu death counts that look more like this: https://www.cdc.gov/flu/about/burden/index.html. My first guess would be that it’s an issue of “confirmed” vs. “estimated” - and I’m sure actual covid-related deaths can similarly be estimated to exceed confirmed these days - but I don’t know exactly what the methodology is for each.

it’s been about 3x as bad as flu in January: https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

edit: which is higher than some recent annual flu totals they have, but considerably lower than estimates of annual flu deaths I’ve seen elsewhere? I feel like I’m missing some context for interpreting these numbers.

Recalls Dick Van Dyke (still alive, by the way) being a VFX and 3D modeling hobbyist in the Video Toaster/LightWave Amiga days.

Surely Apple recognizes “*nix-sphere web software developers who want a nice laptop” as one of their target professional markets in the OSX era, too. That’s sold a hell of a lot of MacBooks. It’s developers of software for their own platforms that Apple has long had a tense relationship with.

The original iMac hardware specs were also pretty reasonable for when it came out, no?

[2] doesn’t actually seem to call out California by name, though. It would have been a little unfair if he had, since California’s most famous public university was teaching intro CS with Scheme and SICP at the time.

I don’t think the autonomous vehicle example and the radiology example are really analogous. I’ve agreed for a long time with the premise that an “almost self-driving” car isn’t much good - certainly at the consumer level the whole value proposition is that you don’t have to pay attention. But that’s a result of it being a real-time application constrained by physics. Medical diagnosis is a more deliberative process that already incorporates inputs from multiple tools.

My first ever programming book, a well-intentioned gift from my parents, was “Sam’s Teach Yourself C++ in 10 Minutes” - meant of course as a quick reference but always conceptually funny to me because… it’s C++.