Write the program, punch the cards, send the cards to another building to be loaded, program runs, printout comes out in another building, somehow this takes 2-3 days
There are apparently several traditional roof shingle makers who have previewed solar tiles. A friend of mine who is a roofer told me he would not trust Tesla and is waiting for a reliable supplier. Apparently he is a smart man.
Oh and don't forget: it keeps chaining a bazillion commands together so any whitelisted commands still need approval because they're nested in such a convoluted way.
They love adding flags to fix issues that users have without telling their users about the flags. It very much feels like: As long as our staff can have a good user experience, we're happy. We don't care about anyone else.
Agreed, but reminds me of the unfortunate usage of English as the de facto global standard. Ironically funny to see computers chattering in this odd piecemeal of a language. (I am a native speaker of English)
Would be interesting to see LLMs talk in something more terse like Vietnamese.
The Edit tool has been notoriously tricky to get right - it seems they have maybe branched out but I think morphllm started specifically with the pitch that they trained a small model to be good at editing files - most of their testimonials are about that
But I think it’s mostly a solved problem in frontier models and the bash tool usage is more likely an attempt to be more token efficient - I’ve noticed it used for making mechanical bulk edits that would be numerous “edit” tool uses otherwise
depends on the breadth of the edit. anything that involves multiple files might be better done with python (e.g. renaming a function, along with changing all call sites.)
Because they're thinking like I did going into the article. Harnesses like Claude expose "tools" to the agent. I usually use Cline but I'm giving up on it for this exact reason. Cline tells the model "you tell me to write a file, I'll get it done" and then it messes everything up, causes tones of errors, and the model goes "wow that's a broken tool. I'm going to write a python script to write the file instead"
They said as of aug 23rd it's 100% rollout. I've noticed 0 improvement. If anything a downgrade. The agents often refuse to use the tools after 1 try because the results are so trash. An open ticket that shows my biggest gripe with the write tool https://github.com/cline/cline/issues/13276
I guess the mistake count is down though. But that's because it's bypassing all the tools and just running commands
1. JSON parsers are available at your corner convenience store.
2. The format is too simple to have ambiguous behavior. No weird “yes” means true, 0 means false, odd rules about comments, blah blah blah. It’s hard to fuck up JSON (but obviously not impossible if you get creative).
3. It errors out early in the parsing if you mess it up.
4. Its data types are present in more or less any language.
5. Most configs are just key/value. JSON does this reasonably well.
6. It is easy to generate and validate JSON documents. For some use cases you don’t need a library (though you should use one).
7. There is only one way to do anything (sane).
8. Everyone is familiar with it.
9. It is dynamic. You do not need to pre-define your sections or keys ahead of time.
10. It can easily be auto formatted to look good with zero risk of changing semantics.
11. Data stores often natively support storing and querying JSON objects.
12. If you are old enough to remember the era when every tool invented its own, often very buggy, config format and parser you will also remember the moment you first saw a JSON config file that was parsed with a standard library parser and thought “finally, this is the modern way”, you will understand why JSON continues being popular. It was the first thing that unambiguously worked compared to what came before it.
This is like asking why people use their keys to open packages: it might not be the right tool for the job but it’s hard to mess up, is the closest thing to you that can get the job done, and everyone (with functioning hands/fingers) can do it with little issue.
I am also certain there is some small but non-zero percentage of people who do it simply because everyone else moralizes about not doing it. Spite is a powerful thing.
What’s really fun is mixing signal handling and threads, especially on Linux. There is a simple way to do it and about a thousand ways that include at least one gotcha.
Self-pipe, yeah, except in Python you don't have to build it. signal.set_wakeup_fd() is exactly that: hand it an fd (or a socket on Windows) and the interpreter writes the signal number to it. Then you select/poll that fd in your normal loop and do the actual work outside the handler. asyncio uses it under the hood for add_signal_handler.
The other one that plays nicely with threads is blocking the signals everywhere with pthread_sigmask and parking one dedicated thread in sigwait(). Both are in the stdlib on Unix.
signalfd is nicer than either but it's Linux only, which is why set_wakeup_fd usually wins if you care about portability.
I am curious why LLM writing has such an uncanny valley feel to it. Like if I was talking to a person who constantly used a phrase they liked I would notice it and it is possible I might get irritated by it, but I wouldn’t necessarily.
In high school I had a teacher that would say “that type of thing” a lot. One time my friend and I counted it during one class period and he averaged to use the phrase every 48 seconds on average. It was funny, but it never irritated us.
And this is just one example of I am sure thousands I have personally experienced where a friend, family member, or coworker has a peculiar way of speaking and it at most feels odd but not annoying. Yet when I see an emdash now I instantly feel irritated.
And I say this as someone who actively enjoys using Claude and other LLMs, including coding, casual research, or even having it explain pop culture phenomenon or sociology research to me.
> if I was talking to a person who constantly used a phrase they liked I would notice it and it is possible I might get irritated by it
There are sociological reasons why this happens less with humans:
1. You cycle your dumb repetitive jokes with everyone you meet, so nobody hears it twice
2. Those who know you well will notice when you're just repeating ("dad jokes")
3. As a person's idiosyncrasies are beginning to wear on their social circles, they will be getting small clues to stop saying those things. Agents don't get these social between-the-lines cues to stop a certain behavior, they endlessly repeat. Perhaps between model version releases, frontier labs can harvest the web and ask "What Claudisms do people mention negatively?" but I don't think they do that yet.
>Perhaps between model version releases, frontier labs can harvest the web and ask "What Claudisms do people mention negatively?" but I don't think they do that yet.
That won't happen. People can't phrase their objections in a succinct-enough way. When they do, the objection is superficial ("too many em dashes") and doesn't strike at the core of what makes LLM output bad.
Don't talk about other people in such a way. If we had a way in Claude to mark a word/token and downvote/reduce it logits then it wouldn't be to hard to send loadbearing to token valhalla.
In the future, I hope we get a way to randomize the language idiosyncrasies and/or personalities better.
The inevitable conclusion of retraining on the (now AI generated) web is that models begin to use the lower bits of fidelity in output text to pass messages to their future selves.
I didn't mean it as a criticism, just a statement of fact. I can't do it either. It's difficult to say what about a writing style is bad beyond vague descriptions.
Claude in particular seems to try and use its internal thesaurus but not exactly align the senses. It constantly uses "address" for place or location and "grammar" for structure, meaning, etc.
This is kind of a nitpick, but it seems like with prose writing there are still some things to learn.
This is conjecture, but why couldn't they hire the writing equivalent of a voice actor?
Honestly I don't think it would take much. Doing it ethically would take a little more money, but not much, for them.
Hire a prolific author with the writing equivalent of the "midwestern accent". I have a terrible writing accent, so it couldn't be me, but these people are out there. Pay them a bunch of dollars to ingest their entire corpus, and to produce more as needed. Use an LLM to match the writing style of this author during the post-training RL, and get a brand new Claude voice out of it.
I think the last one is by far the biggest reason for why this happens way less in humans. In every conversation, almost for every sentence, people gauge the response their words have on whoever is listening. If something didn't land as expected (frown, confusion, unpredicted response) you unconsciously adapt and try something slightly different.
It's immediately obvious by the fact that you're clearly using way different ways of speaking (tone, speed, vocabulary) when you're speaking with a friend, versus your parents, your colleagues, people you don't know, children, etc.
I think that there is a sort of mechanism by which, when a human speaks to you, you can "mirror" their internal mental state, and the quality of writing often corresponds to how much you get pleasure or information or whatever your goal is from that mental state. The important part is that the words themselves are just pointers to the state. So, the person speaking, if they are skillful, gives enough words, and enough variety, that you can start to produce a state yourself which resembles theirs.
This is the thing that LLM writing doesn't really do. Since there isn't a mental state, there's nothing to mirror anyway, and somehow you can detect the absence of it even though it's hard to put any of the machinery into words. Corporate speech also fails to activate this machinery in the same way, but LLMs seem to do it more egregiously, probably because the corporate speech was at least compiled by a human---even if it is not the thoughts of an individual, it is the "thoughts" of an "entity", the abstract corporation, which the writer was speaking for, and so you can still wrap your mind around the fact that it is communicating with you.
This makes some good intuitive sense, but to me the sycophancy feels like it is an emergent property of turning a next word predictor into a conversational chatbot whether or not it’s intentionally trained that way. Your prompt and its earlier responses is all it has in its context window, so of course it lends undue importance to everything you say. Does that seem like a contributing factor to you?
>Is there a silent majority of Claude users who really enjoy what we call the LLM-isms? Maybe, but isn't Claude also largely aimed at developers?
Purely out of my own curiosity, I just asked Claude to have fun with itself by making itself a game it enjoys, to play it, and to write its experience.[1] I don't know if it's true or confabulated (maybe it doesn't really know its experience and is just hallucinating it) but I didn't mind reading it, the prose is fine for me. I don't mind reading Claude's writing. I mean let's be honest, we all read Claude's writing all day, most of the submissions on the front page on any given day are written by Claude.
Just before I made that game, I had Fable write up a scholarly report on any subject[2], it chose introspection by LLM's. (This is what made me think of asking it to play a game.) I didn't mind reading it, even though I don't think it really added anything very interesting. I don't think what it wrote is worth publishing, but I read it with interest.
I found I could read it easily and get up to date on the state of this question that it picked to answer.
So the bottom line is I don't mind reading Claude's output that much. Of course, I'm annoyed every time it says "honest", "genuine", "load-bearing", whenever it pushes back gently against something, etc. But it's not the end of the world.
I would think that it would be the opposite. Nobody is seriously detoxing or comfortmaxxing AIs yet. Human brains are fed back its own output in learning mode, so we are great at removing whatever we feel uncomfortable from our output, online and offline. No such paths exist for AIs.
One reason is that one person’s idiosyncracies are limited in scope, but LLM-produced text is now everywhere. Also, filler words and mannerisms in speech we’re quite good at filtering out, but in written text the stand out much more.
It’s the repetitiveness of style, the attempt to make everything seem as impactful as possible, the use of short sentences (too much Hemingway in the training data?), and obvious patterns like “it’s not this, it’s that” and several others.
Real human writing doesn’t follow such strict rules. When the same small set of rules is applied over and over throughout a text, it becomes obviously strange and machine-like.
What's weird though is how consistent it is, even across models to some extent. Were the RLHF people given a really specific style guide?
While I agree that no-one used to write like that as a whole before, all the elements can be found in different places. Short sentences to avoid discouraging poor readers. Maximally impactful statements are commonly used in marketing or other business communication that's focused on selling what it's saying. A bullet-pointy style is used in many kinds of business communication. Etc.
It makes me wonder if part of what happened was a kind of melding of common styles from several different kinds of writing.
I had pretty good luck recently by giving it a writing guide about word choice, sentence structure, paragraph structure, and overall doc structure. I basically ask it to read the guide and revise a couple of times before I engage with its writing. Ymmv.
I did this too, but it usually thinks its writing is fine in my experience. Even when spawning a subagent, it thinks its effusive comments are fine. It's driving me nuts. Before I commit I end up ripping out 90% of the comments, and rewording the rest, otherwise I'd be drowning in comments. This is my style guide: https://github.com/smj-edison/zicl/blob/main/CLAUDE.md#style...
Oh people get like, super irritated, when everyone like, started using the same like, placeholder word. 1 person with a repeating style is fine, multiple is annoying. It's the same as corporate buzz words, or TV/movie cliches, they get annoying through overuse.
I think it also relates to how well the "cliches" fit, and how much sense they make. Ai loves to talk about how things "land" or "the X trap" when the concept just doesn't fit with that language. It's like clickbait articles saying "what happened next will astound you" when what happened is barely surprising or entirely predictable. The only thing worse than an overused cliche is an overused cliche used wrong.
Fundamentally to me, ai writing feels uncanny as it just doesn't know what it's saying. It uses the same tone, style and cliched construction regardless of the message. If someone told you, they got a promotion, were getting married, got laid off or lost their parents all in the same tone pacing and style, they'd come across as uncanny too.
My completely conjectural theory is that natural language is just a shitty medium for communicating knowledge and the idioms and motifs that the LLM uses are an emergent "code" or structured syntax that work better for communicating ideas.
If you think about it, "load-bearing" is a pretty commonly used concept in pedagogical writing. You could say "most important", but it's not quite the same in meaning. English just doesn't have a better word to describe a concept that occurs this frequently. The LLM's catchphrases reveal blindspots in the English language itself.
I think it presses a few buttons we probably recognize (if subconsciously) and find distasteful.
The verbosity makes me think of two things in particular:
- The classic essay written by someone who has 125 words worth of actual content but a 1500 word minimum. Those three paragraphs could be bullet points and convey the meaning just as well. The screen-filling chart of every test case you ran that came back green manages to be less actionable than a direct "one test out of 54 failed." I fully expect to see Claude tell us that "Support Ticket 8257 is a Land Of Contrasts" at some point.
- The sitcom trope of the person caught in a lie who figures if they can keep adding more and more detail he'll be believed and can escape the awkward conversation. Stop. Just stop. You're proposing a fix on a CODEBASE THE CUSTOMER DOES NOT EVEN USE. Cue laugh track, cut to commercial.
It's endlessly annoying to me that the em-dash has become the canary in the coalmine of AI writing because I've always used them extremely liberally in my writing.
Honestly, this might sound elitist, but I suspect it's because it's an "advanced" punctuation that is not known by most people, so is not commonly used. But the training corpus of these models puts more weight of academic writing or published books writing where the em-dash is much more commonly used.
I don't think its an inherent quality - a lot of older models had a much more natural feel to them.
I guess it has to do with the low temperature (low randomness in final word choce - something all vendors seem to have converged on for some reason) which does make them less likely to skiz out but makes the writing feel dry and samey. It's like repeating a list of dice rolls, and replaying them - there's no inherent pattern in the input, but there sure is in the output.
I suspect because the English colloquial text training data it had access to was early 2000's message boards and social media. Therefore it uses "honestly" and other expressions that took hold in the late 90s and early 2000s.
I guess I am in the minority but I did not have a good time with Darktable. I used it for a few years and then had to hold my nose and transition to Lightroom. On decent hardware (MacBook Pro that was only a few years old) it was dog slow. But the worst part was transitioning from their version 2 to version 3 of the workflow or whatever they called it. None of my old photos would render correctly after that and half of the modules I learned to use to edit my photos were not obsolete.
The documentation for it is dense. It’s like you need to know the inner workings of the software to be able to do things like basic color correction. And even basic masking was just awkward as hell. When I quit it did not have smart object detection either so removing things from the background required editing the photo in external software.
Lightroom isn’t perfect. But it is fast, it is intuitive, and it does not expose modules that the documentation says to never use.
I really wish I could use open source software for this. But I have yet to find a good alternative that lets me do what I want (which is fairly basic, I don’t go crazy with my editing) that doesn’t require me to get a Masters degree in its usage.
I'm not a Lightroom user myself, but my mom is a photographer and uses Lightroom for all of her photos and at least on her PC, it became extremely slow and painful to use. And it's not like it's a super outdated system either, it has a Ryzen 5600G, RTX 4060, 32GB RAM, nothing too ancient...
We even tried contacting Adobe support, and they claimed that the catalog was corrupted, took the catalog file, and later retured a "fixed" version which had no effect.
They then blamed Windows 10 and suggested we upgrade to 11. Does anyone here have any better tips to get it running more smoothly?
You are in good company. I went into Darktable enthusiastically and with willingness to learn its workflow.
What I found was crippling bugs and a piss-poor UI. I've even gone back to it after a few years to see if I'd soften my stance, but nope: It blows.
Part of my beef with it is discussed below; people are acknowledging that it sucks at organizing (and, I submit, browsing). But it purports to do it, so you waste time wrestling with its ineptitude.
The lack of easy browsing is an issue, but Darktable has never purported to provide an image browser / organizer. Quite the contrary, the devs have always been crystal clear that they hate the concept with a fiery passion.
The lack of good open-source image browsers has, oddly enough, become a major problem. Only recently did someone float a pretty good one for Mac (FlowVision). I don't know what options there are on other OSes at the moment.
They include an LLM response with every single Google search, whether it is warranted or not. That scale is, my guess, many orders of magnitude higher than what OpenAI and Anthropic serve. And for Google none of these are paid interactions since their LLMs do not (YET) insert ads into the responses.
So my guess is that Google will continue having compute shortages until the Gemini enshittification starts.
I don't think so. According to some very basic research there are around 8bn searches a day, or 250bn a month.
Let's assume Google serves AI overviews on every SERP (they don't) and don't cache them (they do, afiak).
And let's assume that each AI overview is 2000 tokens (blended input/output), that's 500T tokens a month.
It's rumoured that anthropic is serving somewhere close to 10Q tokens a month.
Now it may be that AI overviews uses vastly more tokens than that per search, but I doubt it based on speed to render the overview.
My very rough napkin math on this is that maybe AI overviews is consuming 100T tokens/month max (after adjusting for caching and SERPs that don't have them), which would be 1% of Anthropic token volume.
Well I asked the google AI mode thing what it thinks about your comment and it told me this (edited obviously):
"10 Quadrillion tokens a month means: 333 Trillion tokens per day and 3.85 Billion tokens generated/processed every single second, 24/7."
"At an incredibly cheap, subsidized infrastructure cost of $1 per million tokens, serving 10 Quadrillion tokens would cost Anthropic $10 Billion per month ($120 Billion a year) just in inference compute."
It also had this to say about how google's AI overview works: "Google doesn't just feed the LLM your 5-word search query. The system scrapes the top 10–20 web results, feeds thousands of words (tens of thousands of tokens of context) into the model, processes it, and then outputs the result."
Oh, and it does all of that in less than two seconds. Honestly, whatever Google is doing with its infrastructure is so far ahead of everyone else, I can't believe you fell for such an obvious lie.
There are also extremely obvious holes in your comment:
>Let's assume Google serves AI overviews on every SERP (they don't) and don't cache them (they do, afiak).
Try it out for yourself. Add a few random letters or punctuation. They cache nothing.
The tokens served number might include cache tokens which are a huge chunk of agentic token spend - and even with that the estimated burn rate for anthropic doesn't seem wildly off? They spend 1.25 bn per month on their deal with SpaceX alone.
They definitely cache results - I've searched and re-searched an identical query back to back a few times and seen identical results from overview. They are definitely throwing a stupid amount of compute towards these ai results nobody is paying for - changing punctuation and stuff does get you a different response - but they're not doing no caching.
Certainly what they're doing with their infrastructure is impressive but it's not super meaningful at the end of the day for a for profit company to be really impressively good at burning tens of billion dollars on a service nobody pays for while the same tech from their competitors is quickly becoming one of the largest spend categories for many software engineering teams
reply