pdpi

To me the second response reads more like PR-speak, like somebody who's been over-coached on how to talk to customers.

I find it strange that they'd have an AI-driven followup email days after the survey. If you're automating it, why the delay? "Delay the response so it seems more human" is way too cynical.

It's worth noting that OP is talking about Ocado, too. I've been a client of theirs for over ten years, and the number one reason why I keep buying from them is that their delivery drivers never feel super-hurried. They have lower quotas than drivers for other services, and they always have time for a little bit of banter. Compare that e.g. to Amazon Fresh drivers, who'll always just leave the groceries at the door and knock. They're already gone by the time I open the door.

I find it unlikely that they'd automate this sort of second-level customer support response while not squeezing the everliving shit out of their drivers.

And they’re not alone — mercurial is hg and the silver searcher is ag.

Nvidia doesn't really do B2C. They don't sell graphics cards to consumers, they sell GPUs to OEMs who build cards around them.

The "top 20 of 2026" table in the article puts Amazon at $776B, but only 3.75% of that ($29B) attributed to B2C revenue. I don't know whether that number is correct, but AWS is clearly a massive chunk of non-B2C revenue for Amazon.

> It's clear he brought a gravitas to the role that inspired a team of young/new actors around him

That's a lesson everybody on this site should learn — Every senior is a leader, even if they don't think they are. People (mostly the more junior sorts) look up to you as a role model whether you (or they) know it or not, and that's very real power that you must learn how to wield.

I've given negative reviews to technically excellent candidates in the past because just about everybody in the org would look up to them if they got hired, and they would be a terrible role model.

The issue with AI-produced text is rarely bad grammar these days (if anything, that's a marker of human-written text now!), but rather stiff-sounding constructions that humans don't actually use. In this case — semicolons are grammatically correct, but very few people actually use them them (same for em-dashes like what I just used, but you can pry them from my cold, dead fingers).

The same, I expect?

A judge might be a little more lenient with procedural mistakes to a pro se litigant who's trying to comply with the spirit of the proceedings, but smartarses won't get any such luck. At best they'll get a sternly worded warning the first time they get "creative".

I don’t know anything about the Brazilian legal system. In England and Wales, or in the US, the lawyers are officers of the court, which means they have a duty of candour, and a duty to put the integrity of the legal system above the interests of their clients.

That paragraph was never intended to be read by the judge or opposing counsel, it’s a deliberate attempt at messing with the tools they use to analyse the filing. The fact that it’s white text on a white background makes it an incontrovertible attempt at deceit.

We can discuss how wise it is to use AI to analyse legal documents for you, but that doesn’t change how outright deceit should be handled by the court.

A better framing is that security has largely been built on theory that most targets aren't worth the effort to target, so security spending is focused on preventing/mitigating untargeted spray-and-pray attacks.

It's effectively shareware. What was old is new again.

The 2020 Touch Bar 13" was 1.56cm x 30.4 cm x 21.2cm and 1.4kg, so this is the same weight, and thinner while having a slightly bigger footprint.

Yeah, that's what I was getting at. The £700/£800 difference between the iMac and the Mac mini doesn't buy you as much as you'd hope. It's just not very good value at all.

If I were buying a display for work (which is to say, mostly programming), my choice would probably be the £600 BenQ RD280UG — a 28" 3840x2560 display, which is 4k vertically enlarged to a 3:2 ratio.

Lowest spec config costs £1529 vs £899 for the equivalent Mac mini, but the mini gives you Ethernet and 2 cores. Highest spec goes up to £3099 vs £2299 for a matching mini.

£700/£800 buys you a 24" display (with an unspecified refresh rate, so presumably 60Hz), a keyboard, and a mouse. I loved my 27" iMac to pieces, but it's really hard to justify buying this thing today.

One thing that helped me understand it was to reframe it as a network application, where you have a client running on the CPU and a server running on the GPU (the shaders).

A bunch of the complexity comes from having to write both client and server, and the communication protocol between them, then some more comes from the client being responsible for spinning up the server. As with every network service, client/server traffic is the slowest part of the architecture, so a bunch of complexity then comes from having to minimise the amount of network calls (or draw calls, in this case).

Then you hit the part where all of this is performance critical (soft-realtime requirements), so the whole thing is geared towards “writing your own userspace IP stack in C++” rather than “writing a flask application in Python”.

Funnily enough, my bird watching does in fact involve a screen — the electronic viewfinder on my camera[0].

Jokes aside, there is a very definite difference in how my brain "feels" when I'm walking around, camera in hand, looking for things to photograph. It's like I'm in a different "cognitive mode", for the lack of a better term. It's not just about being present and in the moment, but about seeing the world around you from a different perspective.

0. https://drive.google.com/file/d/1tnMlazV0g1syoRPLALJ6576YSYQ...

That model of development is what led to Eric Raymond’s seminal The Cathedral and the Bazaar essay.

“Cathedral” refers to development behind closed doors, where you just get a release and a source dump to go with it.

“Bazaar” refers to development done in the open, where you have constant access to the bleeding edge.

At the time of the essay, GCC and Emacs were ESR’s examples of Cathedral-style development, and Linux was the premiere example of the Bazaar style.

The Cathedral style has become a lot less common over the years, but it’s still how e.g. SQLite operates, I think? E.g. Google definitely operates like that with AOSP, but that's arguably more a fruit of them being maliciously compliant with their Open Source promises.

You still need to add to that the cost of the films/series themselves. (You can choose not to and sail the seas instead, of course, but then it's no longer an equivalent exchange).

With this exploit, what's stopping me from sending my own CSR for your domain?

Yeah, it's a mostly ceremonial role. But that's my point, really — it's not the title, it's the behaviour.

I would argue that, even though it's not my preference, there is real political value in the monarchy today still. E.g. Charles's speech to Congress[0] has a certain weight to it that wouldn't be there if Charles were an elected leader instead of a monarch.

0. https://www.youtube.com/watch?v=tCUZWnX8JIM

> even basic linear algebra is pretty crap compared to python

I wouldn't qualify that with "even". Numerical stuff like that is one of Python's most well-developed niches, and Python's numerical ecosystem is the benchmark everything else needs to be compared against.

I would go further than that.

A bunch of European countries (including all of Scandinavia) are constitutional monarchies, and they don't devolve into the sort of anti-democratic behaviour we're all seeing from Trump. As a UK resident, I would prefer if the UK were a republic on general principle, but I fail to see the monarchy causing any real large-scale harm to society (individuals like Andrew notwithstanding).

The real point they should be making is "no more despots". Some despots are kings, but some are styled presidents, such as Xi Jinping, Vladimir Putin, and Kim Jong Un.

Also, legitimately elected or appointed officials become despots all the time. Hitler had a plurality (but not outright majority) of votes in 1932. Salazar was an MP and minister of finance, and actually legitimately tried to resign from that role several times over many years before he was eventually nominated to the role of prime minister. Ferdinand Marcos won the Philippine presidential election in 1965 with some amount of intimidation, and the Philippines only formally became a military dictatorship under his rule in 1971.

In many ways, that's the scariest scenario — people who legitimately get elected to power, and then use that power to solidify their illegitimate rule.

> Linus Torvalds recently said he uses AI to develop Java GUIs because he doesn't know how to write Java GUIs

Given the quality of the output when I ask it to do things I do know how to do, I'm not particularly comfortable with the idea of using AI to do things I don't know how to do myself. It reeks of the Gell-Mann Amnesia Effect.

I have a slightly more nuanced take on it — I use Claude to write Ansible playbooks for my homelab because I know enough about server management to have a clear opinion on how I want things done and why, and enough Ansible that I understand the fundamentals and can recognise (but not write) idiomatic playbooks. That's the perfect storm of getting the LLM to do something I'm not good at, while still being able to judge its output.

Scala doesn't have an IO monad by default, but ZIO, CatsEffect, and probably others are fairly popular in at least some parts of the community. IMO it's kind of the worst of both worlds (all the complexity, none of the guarantees), but it's definitely there.

I'm struggling to see the point in this. Your roll example using Winnow. Not really that much (un)wrapping logic:

    use winnow::{Result, ascii::digit1, combinator::separated_pair, prelude::*};

    #[derive(Debug, PartialEq)]
    struct Roll {
        count: u32,
        sides: u32,
    }

    fn number(input: &mut &str) -> Result<u32> {
        digit1.try_map(str::parse).parse_next(input)
    }

    fn roll(input: &mut &str) -> Result<Roll> {
        let (count, sides) = separated_pair(number, 'd', number).parse_next(input)?;
        Ok(Roll { count, sides })
    }

    #[cfg(test)]
    mod test {
        use super::*;

        #[test]
        fn test_roll() {
            let input = &mut "3d6";
            assert_eq!(roll(input), Ok(Roll { count: 3, sides: 6 }));
        }
    }

This is, effectively, the game that branch predictors in CPUs are playing with your code. Except that, in that version of the game, you win when the CPU guesses correctly.

Writing by itself isn’t art, but poetry and literature are. And, by the same token, I don’t expect most works of literature to be exhibited in a museum — books are easily copied, so there’s little value to just having the book’s text in a museum.

Works like the Black Perl[0] are decidedly artistic in nature. Demos (in the demo scene sense) are an artistic endeavour. Competitions like the Obfuscated C Contest or the Underhanded C Contest are very much judged on aesthetics.

0. https://en.wikipedia.org/wiki/Black_Perl

I think it's sort of obvious that the limit to this general rule is when data dependencies between fors and ifs forbid you from pushing things further up/down.

A request body in a GET is not standard HTTP. Refusing to handle non-standard requests might be inconvenient, but it's an eminently defensible choice.

I understand why you'd want it, and that's why RFC10008 introduces QUERY as a GET-with-body alternative.

> Vx is the right language for the thing that must be correct and fast across ten kinds of silicon. It is not the right language for the thing you are still figuring out.

A bit tangential, but I really *really* appreciate them making this distinction, and wish more people did this. Way too many tools try to advertise themselves as all things for all people, catering to all use cases, and that helps nobody.

Can we please stop with the "AI = Slop" rhetoric? I'm fairly bearish about AI, and even I'm getting annoyed by it.

"Slop" just means something is low-quality and of poor craftsmanship, irrespective of who or what made it. If it's slop, you don't need to look for tells, you can just tell it's slop. Sure, AI is prone to specific types of mistakes, and humans are prone to other types of mistakes, and you can often tell whether a specific poorly-crafted UI is man-made or machine-made, but bad is bad is bad.