sanitycheck

I have a Samsung tablet in a non-flammable container waiting to be disposed of after being mostly plugged in for 3 years then suddenly swelling quite alarmingly. YMMV.

All the normie podcasts now falsely advertising VPNs as panaceæ for every possible security problem are cashing in too.

1 Is easy enough for trivial tasks but in a complex (typically horrible) production codebase nearly all the work is investigation and debugging. However good the initial prompt is, soon the context becomes flooded with log output and code and the LLM goes off the rails quite quickly.

Doing 2 well is the AI babysitting mentioned in the article. Of course you can stop it every minute and tell it to do something else, then watch it like a hawk to make sure it does it right, then clear context when it ignores you and makes the mistake you told it not to make. But that is often then slower than just doing the work yourself to begin with, probably leading to the findings we've all seen that LLM use is actually reducing productivity.

I think living with crappy AI code is the price we currently have to pay for getting development done quicker. Maybe in a year it'll have improved enough that we can ask it to clean up all the mess it made.

(Possibly I just have higher standards than most, other humans can be quite bad too.)

Not sure I agree with 3.

I can have Claude write stuff that does basically the right thing and actually works but, as so many people say, it's like working with a junior dev. The only way I'm going to get code I'd want to present as my own is by sitting looking over their shoulder and telling them what to do constantly, and with a junior dev and with Claude (etc) that takes longer than just writing it myself.

AI-coded/assisted work should be flagged as such so everyone knows that it needs extra scrutiny. If we decided to let a horde of junior devs loose on a codebase we'd naturally pay more attention to PRs from new intern Jimmy than revered greybeard Bob and we shouldn't let Claude hide behind Bob's username.

I can't pick a favourite, but "Consciousness" (1999) definitely sticks in my mind as one of the most amusing due to the prickly debate.

That's the only way they managed to cover such big subjects in relative depth in such a short slot! All the guests are used to waffling for a full uninterrupted hour to captive audiences about tiny segments of each topic; without that pressure the show couldn't exist.

Eh, Claude is like a magical spaniel that can read and write very quickly, with early-stage alzheimers, on amphetamines.

Yes it knows a lot and can regurgitate things and create plausible code (if I have it run builds and fix errors every time it changes a file - which of course eats tokens) but having absolutely no understanding of how time or space works leads to 90% of its great ideas being nonsensical for UI tasks. Everything is needing very careful guidance and supervision otherwise it decides to do something different instead. For back end stuff, maybe it's better.

I'm on the fence regarding overall utility but $20/month could almost be worth it for a tool that can add a ton of debug logging in seconds, some months.

Then it'll be no different than when they bounce their bad ideas off their human subordinates.

There are jobs in which one may find oneself where doing them poorly is better for the world than doing them well.

I think you and your colleagues should sit back and take it easy, maybe have a few beers every lunchtime, install some video games on the company PCs, anything you can get away with. Don't get fired (because then you'll be replaced by keen new hires), just do the minimum acceptable and feel good about that karma you're accumulating as a brake on evil.

Another cat (formerly) in UK politics: https://en.wikipedia.org/wiki/Catmando

Is this a known thing? I've called myself 'immune to charisma' which seems at least related, and I've thought perhaps it's an autistic/aspie trait but have never come across any studies or articles mentioning it.

I sympathise with the author but I fear what he's doing here is adding one more item to that list of thoughts: "I'm in a trap"

There's a lot of stuff I started writing from personal experience but I've just deleted it because I sense what's being talked about in the article is probably classifiable as an eating disorder and I am in no way qualified to say anything about eating disorders. People who identify with the sentiments expressed in the article may want to consult a doctor.

Every update is worse than the last. I did the same, and I'm particularly loving turning bluetooth on or off requiring a swipe and three button presses instead of one. And I thought the quick settings regressed in 14 when they went from icons to huge buttons, but now they've gone for less-huge buttons with always-truncated text in.

At least with Windows it alternately gets much worse then a bit better again with each version.

Regarding the monstrosity in the link, yes it does make me 'feel things' - things I will refrain from typing out lest they be construed as threatening behaviour.

Be serious, devs doing only useful work would lead to absolutely massive layoffs.

"X and Y were in the garden, Y noticed the ripe tomatoes as they went into the greenhouse". Is X in the greenhouse?

I'm way woker than the average person but I have to admit encountering a singular 'they' breaks my concentration in a distracting way - there's definitely possible ambiguity.

More likely they didn't care and nobody wrote an automated test for it because that would be hard, no human testers are employed (because who even does that now?), and only two users got all the way through the labyrinthine process to report it as an issue so managers triaged the bug as wontfix.

I think this is industry standard practice in 2025, right?

Contracting will do it. (Caveats apply - understand IR35, keep 6+ months of money in the bank, be able to temporarily relocate, YMMV)

It's very surprising to me that this needs to be said - but maybe that's because I've mostly worked in, let's say, 'resource constrained' environments. You focus on the stuff that keeps the money coming, or people start losing their jobs.

I do all that, and don't use any Facebook products. It's not so hard, and has the benefit (to some) of never being invited to a WhatsApp group.

Sadly we can explain most bad things that happen by understanding that most people are f*cking idiots.

Non-Chinese iCloud accounts seem to work fine in China without a VPN. (A VPN is needed for Google's services, though.)

Apple cares about profits, and only about users to the extent that it can profit from them. The touted security and privacy aspects of their service are useful for marketing, but they are not primarily why people buy Apple devices.

They already operate in China, allowing the Chinese government to access all "Chinese" user data. Non-Chinese Apple accounts also work perfectly well in China, which leads me to believe that the extent of data sharing is greater than has been disclosed (or dissidents could just use a foreign phone to avoid surveillance). I strongly suspect the CIA/FBI/DHS has similar access.

(Microsoft stuff also works fine in China, Google stuff does not. Draw your own conclusions.)

I think it's somewhat likely they'll just backdoor it and not tell the public, they can see this is just the way the world is headed.

If you want to message someone secretly, use an audited open source solution - don't rely on a megacorp to look out for your interests.

Any evidence the reproduction rate is higher among tech workers, or even among the wealthy in general? From what I've gathered in the past autistic people are more likely than average to not reproduce.

If the US government is "chucking all their secret data" into OpenAI servers/models, frankly they deserve everything they get for that level of stupidity.

This was my reaction upon seeing the question in the post title, too.

I've chosen not to have kids (my childhood experience of other kids being one contributing factor) but if I did I would not want them to be anywhere near a UK state school.

Also, there's no such thing as sculpture productivity because all sculptors are doing is removing material.

My first thought was similar, though followed quickly by "...but it's Apple, so what's the catch?" The relevant extra things to know are that the SSD is soldered, there are no slots for extra SSDs, and choosing a sensible (1TB) drive is >4X the price of buying similar storage at retail. Still a no from me, then.

(The only thing I do often that's CPU-limited is compiling, being faster at that saves me maybe a few minutes in a full working day; I don't care. I am frequently limited by RAM and I really hate shuffling things around to make space on drives.)

Seriously. What does "screen time" even mean if it doesn't include the 9 hours a day I spend staring at 8 square feet of monitors and the ~45 mins the TV is on during meals?

If they mean "phones" say "phones". If they mean "social media" say "social media". Oh but then all the bullshit about stuff being close to one's face wouldn't make "sense" (I wear glasses, I must be constantly terrorising myself.)

It may be a very slight overreaction but my first thought is that everybody in the entire faculty that produced these findings should be re-employed doing something actually useful like grocery deliveries.

I'm not sure I'd even call "local debt" technical debt in ordinary circumstances - realistically there's always going to be mess somewhere, and encapsulating it away where it can't hurt anyone is normal. If it probably never needs to change unless requirements change (in which case any other implementation would also need to) it's fine.

Perhaps if 24 minion instances constitute an actual problem (rather than just inelegance) their example for it is actually foundational debt related having a "minion" being the simplest primitive that would do the job when maybe something lighter could have existed.

I've done the same, and my experiences also match those in the article.

Using a LLM as an autocomplete (typescript, kotlin) I found it'd be right about 10% of the time, very obviously wrong ~30% of the time, but mostly (~60%) *plausible*. Plausible is the worst, because if I don't spot the error immediately it'll come back to bite me in the coming hours/weeks and it takes me much longer to fully comprehend existing code than it does to write new code myself.

Asking questions fares better, but I never have easy questions and for difficult ones it'll usually be wrong. Like, if I'm asking an LLM a question I'm usually at a point where I've read the docs and there's something I don't understand - so it's not that surprising that the LLM won't understand it either. Occasionally pointing out the flaw will get a better answer, but more often it just leads to a circle of flawed answers. I feel like people who say this is some sort of revolution are perhaps people who weren't very good at finding information themselves.

The only places where this stuff seems to be an unqualified success so far is for jobs where the quality of the output is not very important.