It's just another way of saying, you can't have AI submit things on its own and you can't "blame" AI for anything you submit, you're expected to understand and be able to explain anything you're putting forward, if there is an issue with your submission, doesn't matter at all if you go back to AI to address it, so long as all issues are properly addressed.

The code is either of sufficient quality, or it's not, doesn't matter whether you or AI wrote it, it's your name on it, you're responsible for it and its your reputation on the line.

The sentence is maybe ambiguous, but it sounds like it's saying to not use the AI at all until you understand what's happening in the area that needs to change.

I don't agree with this as you can spend quite a bit of time digging around to first find the correct place which needs to be updated and then more time reading the code (likely in multiple files or places) to build up a mental model of what's going on.

I typically have more than one AI conversation on the go at any given time, so I'll always start by asking it to make a plan to solve a problem and while it's busy exploring to make a plan I review some other AI conversation's work in the meantime, when the AI's plan is ready I review it.

If at this point there is anything that is unclear to me about its approach to implement the plan, or I don't already have familiarity with the part of the system it's proposing to update, I start questioning the AI on how the system works. I will scrutinize the AI's plan until I'm absolutely sure that it's going to implement something that is sound from a system design perspective and as a senior software developer with architecture level experience I am capable of making such judgments.

I'm not sure what it might be like for comparitively inexperienced developers, but I would expect that AI's capacity for infinite patience allows them to ask as many questions as they need to get a clear mental model of what's happening, so if I were mentoring such a person in this age of AI I would drill into them on how to interact with AIs to ensure they have this clarity before the AI starts implementing anything.

The difference between vibe coding and highly accelerated agentic development is having the expertise and judgment to ensure you're building properly sound software.

What’s the issue they have with South Africa?

She remembered her late mother used to talk a lot of her as she managed to become a lawyer despite being an Indian in apartheid South Africa.

Because adding to a blocklist is generally more effort than it’s worth to me, I have an email signature which simply says “Please stop emailing me”, nothing else, no “hello”, no sign off, just an (appropriately) completely cold one liner response to them.

Not memorable enough that I’d burn any bridges in my future, but also uses as little of my time as possible while they get a completely unambiguous message of what I want them to do.

I then use my mail provider’s report spam button.

This was used to reward people who had actually paid for the movie with being forced to endure obnoxious and long “you wouldn’t steal a car” anti-piracy appeals before they could watch the main feature.

An interesting effect of them doing this is that it made the pirated version of the product objectively better than the legal version. Great thinking by movie studios.

In her defense, I had my birthday party just a few days before and it was an unusual year in that school had started by (maybe even on) my birthday.

I remember feeling a bit disappointed on the day, but certainly not distraught or traumatised. I also never found myself randomly thinking about the incident after except for when my mom mentions it or from stories like this.

My point is that it depends on the person on how big a deal it was. Also, perhaps the parents do feel guilty about it, hence sometimes sharing it as a “funny story”, but I wouldn’t know.

My own 6 year old child seems to be neurodivergent (I suspect I am too) and is struggling a bit with pre-school, even now at a bridging school.

We first had to confront it about 2 years ago when the highly regarded mainstream school she was at said she would have to repeat the year.

At least for me as a parent, the possibility that my child won’t grow up to be a self sufficient is terrifying, so it was quite a bit to come to terms with.

Fortunately we weren't in denial and she’s getting exceptional assistance, but some parents definitely find themselves in denial when confronted with this situation.

Main domain has SPF set up correctly as per MS docs.

We have the CNAME set up for a like em1234 sub-domain as per SendGrid’s docs which their docs say should cover SPF even if the emails we send have a from address of our main domain (eg noreply@example.com), this is apparently because receiving servers are supposed to do SPF checks against the replyTo address which sendgrid does populate with an address on the subdomain.

This has worked fine for years, Gmail for example is happy and looking at mails from us says everything passes.

However, we recently onboarded a large corporate whose server was blocking the SendGrid emails because it was checking the SPF against the from header rather than the replyTo.

Only way to let the email through was to either add SendGrid SPF records to our main domain, or change the from address of the SendGrid emails, neither of which was 100% ideal.

Not going to try tell the client they’ve configured their server wrong, but have they?

Firstly I find it’s an excellent way for me to get very good clarity about what will be built and whether it’s going to be done in a sensible way.

Very often I don’t really know what the work will need to look like until I’ve explored the problem with the LLM towards first making the plan.

Without a plan I find myself having to do the initial understanding through code review of its generated code which is much harder than reviewing a plan, and then I invariably need the LLM to fix up what it did which is much slower when it’s doing code than working on a plan, never mind the next review I need to do.

And when the plan is good enough, I clear the context before telling it do it, which I’ve found vastly improves the quality of the LLM output.

AC charger is plenty good enough and charges at about 2kW.

Last filled up my tank on July 11, since then have done about 3000km and, checking the car app now, the tank is 87% full.

Not everyone has the same typical daily commute distance, but I chose our PHEV based on ours, BYD has PHEVs with more battery and of course full EVs.

Amount of battery capacity significantly affects cost so for us the ~USD 35K for a SUV made sense. We also got it just after the Iran war started figuring it made even more sense in terms of reducing petrol costs.

Seriously though, speaking as a South African living basically next to Sandton, there are plenty of traffic lights near to me which cause complete chaos when they are out. They will cause significant delay for the entire working day on certain routes, e.g. anywhere on Winnie Mandela (formerly William Nicol) Drive.

There is however one intersection (Stirling and Balmoral Ave) on the way to my kid's school on a non-major road which definitely works better when its traffic light is out, and this is at any time of day in my experience.

Some traffic lights definitely should rather be mini traffic circles, e.g. on Peter Place by the one office park and had the exact same frustration for the one on Grosvenor Road near to Bryanpark.

As the Johannesburg Road Agency (like all City of Johannesburg Municipal companies) is so incredibly incompetent, our traffic lights are incredibly unreliable, I shudder to think how much it sucks out of our economic output all the people sitting in traffic longer than is needed simply due to an out of order traffic light.

And the local police also don't help with regular roadblocks while at the same time being terrible at law enforcement (yes, you'd think roadblocks are a sign of law enforcement, in practice though they're more for soliciting bribes while screwing up traffic).

It's that they made no effort to bring over profiles and their associated watchlists, "continue watching" is my most used section of all the streaming apps (which they love moving every time I relaunch the app, but that's a different topic) and they didn't even apologize for essentially throwing away years' worth of tracking what the last episode I watched was.

On the bright side the "new" app (which I assume is actually the same as used in the US) seems to not have the following annoying issues of the old one:

- It was impossible to enable auto play on a "kids" profile. As my kids can't use a remote yet I had their profile set to "adult".

- An episode would never be marked as "fully" watched, but this wasn't just a mild annoyance, it would "break" autoplay when re-watching a series because when it started auto playing a next episode which was previously watched, it would "continue" from where you had watched upto (minus about 30 seconds). This results in it cycling through the last 30 seconds of every episode of a show. The easiest way around this I found was to make a new profile every now and again for the kids.

Of course any guest is free to write in their review “can’t make temperature lower than 18°!” but the number of people choosing not to book for that reason in any given year is likely 0 or very close to it.

There are a lot of people who don’t “get” thermostats and their intuitive model of ACs is like that of getting a bath to a particular temperature by turning a faucet to full hot or cold to get what they want.

And if these people are out of the room when it reaches the actual temperature they want (and would change it), then the AC is left possibly for hours working full time, burning maximum energy and related wear and tear by trying to hopelessly achieve an unattainable temperature.