prmph

> A few days ago I was looking at a file with a list of expenses and I thought: “It’d be cool if I could automatically sum all of them… wait a minute, let me throw some code at it!” Half a soda later, I had the first version of calc blocks (a block that evaluates simple operations and shows the result in its margin) working in the app and already useful.

A few more of these, and you have a blob of software that has no recognizable shape, and does "everything", with no guardrails and gates on anything, whereas meaning arises from these.

The problem is not knowing when to stop, when to avoid implementing something. The AI will not stop you.

The skill and knowledge and intuition for what make good software cannot be offloaded, no matter how many times people write articles asserting that the machines have got it all figured out.

I'm not sure what this is all about.

If anything, I enjoy programming more now with AI assistance, not vibe-coding. The depth and breadth of my work has increased greatly (yes, I still read and write a lot of code and documents, that the agents build on); I'm like a kid in a candy store.

Projects that would have seemed too daunting now seem feasible to tackle. I could even get started on implementing my dream language, or a knowledge management system for agents, or ... the possibilities are endless.

Now, my software engineering is for my own company, maybe that makes a difference. There is no one breathing down my neck to use more AI where it does not make sense, or to use AI this way or that. I am in the driver's seat, and I only delegate to AI what I need it for, and what it is good at. There is still much that is required of me that the AI cannot do, not anytime soon.

Are you asking me for session logs or what? Here are two examples from just this week:

1. I created an authz lib for the app that enforces both at runtime and with types. I wanted to extend it to check not only the base DTO type shapes, but also DTOs with related record DTO properties. The context that is checked a triple of Role (e.g, admin, etc), Action (e.g., read, modify, insert, delete), and Entity.

Claude Opus initially tried to implement this enhancement by adding a role generic parameter to the DTOs (which kind of duplicates my utility type that accepts a DTO type and authz context and returns the narrowed type) and also, weirdly, wanted the role narrowing to apply only to nested DTOs. I had to throw out all that work, and ask it to enhance the existing utility auth types, and make that the authz narrowing applies to both the main DTOs and their related sub-DTOs always, at the same time.

2. I wanted to create a messaging system for the web app I'm developing. For example, where certain users with specific roles create specific entity records, admins are supposed to an email or SMS (depending in factor they registered with) plus an in-app message. These messages, especially the emails, are not just text, but formatted. Claude Opus wanted to embed formatting in the code that creates and send the messages. But this code works better when split up into various responsibilities: the message factory, the messages template, an the actual sender.

So I had it separate out the templates, implemented as JSX. Now if I want to change the templates, it is simple for me change just that, rather than hunting in some monolithic code to do that.

How is AI anywhere close to making better architecture decisions? And how it is close to even writing better code?

People keep saying this, and again and again I find it be false; a single session with Claude Opus is all it takes to debunk this, and yet people take it as given.

I have to wonder, have people drunk the vibe-coding cool-aid so much so that they are asserting the superiority of code they have never reviewed?

Thanks for the explanation. I agree the -loos suffix is indeed (amusingly) fitting, if that's what it means. Specu-loss?

Yes, cookies are not the most healthy thing, but there are better and worse ones. It is not just even about it containing some sugar. Why replace the spices that made a traditional cookie special, with more sugar? What is distinctive about sugar?

But that is my point: I specifically chose examples where people have an actual point to be concerned, and yet the amount of hand wringing over these issues pale in comparison to the concerns about mosquitoes.

Aren't they supposed to be full of sugar? Each time I've been tempted to pick them up when shopping, one look at the ingredients and nutritional profile is enough to dissuade me. I guess trying them once in a while won't kill me, though.

Apparently, they are a brand of speculoos cookies, and this Wikipedia synopsis is eye opening:

> speculaas [speːkyˈlaːs] ⓘ refers to the more traditional variant, made with a mix of various spices, which is nowadays most popular in the Netherlands; speculoos [speːkyˈloːs] ⓘ refers to a Belgian variant that omits the traditional spices in favour of caramelized sugar for flavour, and which is nowadays the most common variant in Belgium and worldwide

So the Belgians enshittified speculaas to create speculoos, pumping it full of sugar, and presented it to the world as something special.

Modern design is not synonymous to dis-empowering design.

FWIW I love modern design, and I hate all these too clever skeuomorphs. Modern design is honest to the digital medium. I find its beauty lies, not in the quirky ways it tries to mimic real textures and objects, but by how it provides usable controls and information by applying basic design elements like spacing, alignment, color, typography, discovery, layered info density, etc. well.

None of these requires skeuomorphs. It is like the difference between thinking a website is visually impressive and stunning because it uses a nice background image. That would not impress me. What would impress me are stuff like: does it have an effective information architecture, as reflected in it main menu and general link structure? Does it use a tasteful and usable color scheme? Does it have an interesting and still usable visual grouping? Etc.

I always find it kind of odd that whenever any thing about reducing mosquitoes come out, you have a ton of people all echoing some variation of: "but what about all the possible effects"?

I don't know, what are the possible effects of antibiotics? What are the possible effects of uprooting weeds, which some organisms may depend on? What are the possible harmful effects of putting out wildfires?

For every action that someone might think of as beneficial, there are all sorts of "harmful" effects that can be posited. In the examples above, these harms are not even just theoretical, and yet people do not constantly wring their hands worrying about wildfire fighting, weed control, or the proper use of antibiotics.

That this seems to happen only with mosquitoes seems quite weird to me, especially when it occurs even with efforts to reduce mosquitoes that spread malaria and other diseases. People are oddly very worried about some nebulous harm caused by reducing real and serious harm now.

I mean, there are unknown risks to having children, to eating, to having mosquitoes around, to pretty much everything.

It is not meaningful to allow every conceivable unknown unknown (to channel Rumsfeld) suck out any sense of agency.

> If the fence line is on the state land, which it almost certainly is barring a surveying error when it was put up, it is entirely possible that if he fixed it, he would actually get in trouble with the state.

Oh, the joys of living in a place with actual rule of law.

Here in my 3rd world country, people are free to annex land demarcated for streets that adjoin their property, without any real consequences, as long as it is not meant to be a major street.

The average white collar or knowledge worker in my 3rd world country can usually pay for mid-tier private medical care, including specialists, without any problem.

Of course the top tier private hospital and specialists can be pretty expensive, but is not that expensive to avoid the public/government health system.

I'd be rich if I got a dollar every time dentists have tried to push a root canal as the solution for pretty much every dental problem I have had in recent years.Stood my ground, and got a crown without a root canal, and I am fine.

Same for optometrist/opticians. Always pushing me to get glasses. I know my vision would be better with glasses, but my vision has remained pretty good without them even into middle age; no need to fix what is not truly broken

I usually just try to give them a detailed picture of my symptoms, as far as I can, and listen to them. When it seems like they are getting ready to wrap up, I simply bring up my concerns, if they have not been addressed. The attitude should be one of seeking their reassurance that they have adequately considered more serous possibilities.

It does not always work, but a lot of doctors in my experience take some time to kindly explain why they think the other possibilities I'm concerned about are unlikely, and sometimes my questions actually prompt them to add some additional investigations.

But the thing is that paying out of pocket would be very expensive in the "developed" countries.

The advantage of the less developed countries is that you can get pretty good private medical care for a reasonably price. The quality may not match the very best care available in the rich countries, but in many cases, it is adequate.

You are very right.

But, might be more complete to say that complex numbers are two-dimensional numbers used to model a lot of different things. So simply saying they are two dimensional numbers only part of the explanation. To fully understand them, you need to consider the notion of a 2-d number fits the application area, and what it means in context.

For example, complex numbers can provide an elegant way to represent rotations, oscillations, waves, etc. making certain calculations much easier.

But complex numbers are also used to extend the real number system in a way that ensures every non-constant polynomial has a root. In that context, they aren't merely a convenient tool for representing two-dimensional phenomena; they help complete the number system itself in a mathematically precise sense.

And then there is quantum mechanics, where complex numbers play a fundamental role in the mathematical description of reality itself.

Oh, but they can be considered numbers.

I guess what counts as a number is arguable. But -1 is a number, and yet it has two parts: the natural number, and the sign. Sqrt(2) is a number, and yet it cannot be written down exactly. It requires an algebraic representation.

Generally, I think we use the word "number" for mathematical objects that have some notions we associate with the natural number (which are what we most intuitively understand as numbers), like ordering, summation, a notion of zero, etc. Complex numbers do not have all of these, but they have enough that we consider them numbers.

Another way to think of this: most of these non-natural numbers arose as a way of being able to represent results of basic operations on natural numbers, that are not themselves natural numbers. These results could sometimes be operated on further to yield a natural number again. But if we did not invent new numbers for these, these result were not easily expressible, and led to gaps in our math frameworks. In fact, a lot of math progress was hampered by not recognizing these numbers, and progress followed once they were recognized. And the reason we call them numbers is that they arise from basic algebraic operations on natural numbers.

I don't know. Claude Opus 5.5 still makes rudimentary mistakes all the time as I use it.

More than that, it cannot go beyond a certain level. Like a dog that can only do dog-level actions. I've lost count of the number of times I step back, take a critical look at what it has been hacking away at for hours, which would seem OK, but can be replaced by a much more effective architecture.

For the foreseeable future, a human is still very much needed in the driver's seat.

OK, but CSS is not just styles; its in the name: Cascading Style Sheets. The cascading is the problem. Plus it trying to be a programming language.

A lot of these are CSS-related. Not sure why so much effort is still being devoted to piling kludges upon kludges on the train-wreck that is CSS.

CSS is one of the worst (web) technologies, breaking the very concept of composability, a core technique of complexity management. It also has all sorts of constructs trying to replicate the features of a proper programming language. Rampant use of CSS is an architectural smell.

And yet people will defend it to death. Every large project I have inherited has massive CSS files filled with hairy rules, utterly unmaintainable despite whatever people will tell you.

In my own projects now, large or small, I refuse to use CSS broadly. Except for a few global styles in a single small CSS file, I use inline styles for pretty much everything, combines with other code in a neatly composable components. This works very well.

Sorry for my strong opinions, but in my view the over-reliance on CSS and amount of work on it has held back the web from being what it really can be for rich apps. What we should be aiming for is to build in rich and very customizable components right into HTML with real-world sophistication. And allow websites to be installed as apps easily. These alone will do more for the web that all these CSS band-aids.

Sounds like none from the entire cast of characters in this affair were/are without blame.

Very good comment in general.

And this is very true, notwithstanding that I like Postgres a lot:

> SQL Server, which is still the easiest, most batteries-included and hands-off database to this day.

Indeed, HTML is like the assembly language of the web: clunky and very limiting to use directly.

That's why we have to resort to all kinds of libs that render and manipulate HTML. And from there comes the need to bundle large amounts of JS with pages, and almost all other web development "sins"

A real pity, elevating HTML to be actually directly useful for rich interactive UIs should have been a primary activity of the relevant working groups.

I deeply detest "wacky scrolling schemes where unusual things happen as you scroll down".

But I guess (some) people without much technical exposure might be impressed by them. My and my strong dislike of the UIs is based on aesthetics mostly, I don't even care to consider the technical horrors behind them.

Even with prions, there evidence that the body (and probably more so with some people than others) has some level of ability to resist, slow, or sometimes prevent a prion infection.

See these:

- The influence of PRNP polymorphisms on human prion disease susceptibility: an update [1]

- A naturally occurring variant of the human prion protein completely prevents prion disease | Nature [2]

[1] https://pubmed.ncbi.nlm.nih.gov/26022925/

[2] https://www.nature.com/articles/nature14510?utm_source=chatg...

Aren't ALL type system proof systems?

This kind of matches my experience. Not sure if I have Aphantasia, but my mental visual imagery is pretty sparse and not that well formed.

And yet, I can somehow imagine the look of a specific person or thing. I can't really see it as a vivid image, like actually seeing something in real life, but I have a sense of imagining it as visually distinct from any other image.

My audio imagination is top-notch though, I can replay songs and sounds in my head perfectly and vividly.

Interesting project.

One small suggestion though: try to always make the first sentence show fully in the card. Several times I was a bit miffed I had to actually click to find out what a card was about, when it would have been better for me to get the gist in advance to know if it was something I wanted to read about.

I really do not see any proper argument you are making here.

> their (almost certainly wrong) intuition

So because some people can't use some information well, we should deny that information to all?

> I also don't really see how checking world maps would be applicable outside the news in question being consumed as entertainment,

Is this even an argument? I do not consume news purely for entertainment, neither do those in my circle