Yeah I’ve been thinking that LLM unit tests are basically snapshot tests. Just sorta ossify things in place. If they break, you just ¯\_(ツ)_/¯ and have the LLM fix them. It’s like they were never there!
So it’s just like the olden days of everybody ignoring tests, but we give anthropic a ton of cash
Anecdotally, coworkers are writing a LOT more unit tests. By which I mean NOBODY is writing unit tests, but Claude is generating a ton of em. We’re talkin 300 lines of unit tests for 20 line changes that are already covered by other, better kinds of tests. Huge JSON objects of test data that we already have generators for, etc
> but Claude is generating a ton of em. We’re talkin 300 lines of unit tests
Of course - it is setting itself up for more token consumption later: when a small change is needed, it may have to parse/adjust 10,000s or 100,000s lines of test data / code...
A sensible senior developer would recognise the shackle that huge brittle tests suits can become and correct course. But this is against Anthropic's business strategy
It’s a fascinating story similar to this one, but using “old” AI. The dude spent years learning genetics, and built a program called “medikanren” with Will Byrd. It’s p much a huge graph database of facts they extracted from academic papers, and query using logic programming. It’s a great story
Thanks for sharing (I am OP). Lots of really interesting computational problems in genetics. The original speech recognition models, HMMs, diffused to genetics ahead of many other fields with which we associate machine learning. Look forward to watching this tech talk.
I hate the History API especially pushState. Even with this limit of 100 times per 10 seconds it still pollutes my browsing history too much. I need to vibe code an extension that makes pushState/replaceState noops on all webpages.
Seems like you hate the abuse of the API more than the API itself. For Single Page Apps it makes sense to support the Back button by, based on merit, populating the history synthetically
(Lost my passwd to my throwaway so i had to create another, sorry)
Nah, just throwaway here. A few tech/work friends know of it, most of my non-tech friends know of my background but most them have crazier stories. And those folks dont really understand what I do for work or how much money I'm making. I'm too much of a dirt bag for the tech world and too much of a yuppie for my old punk friends. Its double-sided imposter syndrome.
I keep getting baited by these comments so this is the last one I'll respond to, lol.
Elixir is always been sort of a "typed dynamic language" due to how baked in pattern matching is. Any good Elixir developer has always been thinking about types anyway, it's almost impossible not to.
Also, as I kept forgetting to mention, there are no overloaded operators (`+` only works on numbers, for example... unfortunately it does work on both ints and floats but that's another story). The one pain point is that comparriason operators works across all types, but the compiler has already been warning against doing that for at least a year now.
Ok dang. Well. I guess I have no reason not to dive fully into the elixir now.
I’ve toyed around with it a handful of times and I really like it. I like the clojure-ey immutability and threading operators and such. And of course I’ve heard so much about the magic of the BEAM and the phoenix framework. But between typescript and clojure I’ve never felt like I needed anything else.
But if the type system is pretty good, that’s a huge plus over clojure in my book.
> Websites are broken by default. They used to be functional, fast, and accessible but ugly. Now they're slop, agentic, and on fire — but they get attention, and attention is the only metric left. Nobody's reading and you know it.
I’m upset if an LLM actually wrote this because this is p sick
It's AI. Telling a model to shitpost really isn't very interesting since that's pretty much half of everything on the internets and there was practically infinite training material.
Something that reads like an LLM wrote it is different from an LLM having written it to begin with. Something written by an LLM can be something that doesn't have the hallmarks of LLM all over it.
I was just saying that the original quote doesn't strike me as something that's an annoyingly good piece of LLM writing.
There's a lot of experimentation happening on how to get LLMs to write well, starting with half of what's been posted on Gwern's blog as of late.
If they are, I have not heard about it (which does not mean much, I check Java once a year). And if they really are then I'd give Java a serious look again because it's a mature ecosystem that was gimped by ancient runtime decisions for literal decades.
As of Java 24 (Java 25 being an LTS) I'd say they are equivalent. You can use a virtual thread just like you use a regular thread and there's basically no handicaps or gotchas. In Java 21, when they were released, there is a gotcha that the pretty normal use of the `synchronized` keyword would pin a "carrier thread" which ends up blocking all virtual threads from running on that carrier thread.
Pinning can still happen in some much more rare cases, same with go. For example, FFI.
The memory usage, performance, etc are all go like. You can spawn millions of virtual threads with hardly and memory requirements and without overburdening the OS with context switches. The JVM also enjoys faster GC performance with virtual threads.
Not the person you're replying to, but have you tried TypedClojure? I've always thought clojure-with-types would literally be the perfect language, but I also read TypedClojure is more of a research project than a real language that you should use in prod.
No sorry, of the things I've listed, I'd never seen nor heard of a project that uses Typed Clojure, nor probably inheritance via dervied/isa?.
For static (partial) typing, I instead use Malli schemas. I do this for every larger Clojure program I make, because there's always something that needs paranoia, or it's handy to generate example data.
So it’s just like the olden days of everybody ignoring tests, but we give anthropic a ton of cash
reply