OK!
-
@aeva @jonny Last night I was trying to find an animal ER, rehabilitator, or just ANYONE who could help an injured young opossum I found under a car, and one of these animal sanctuaries I called had me speak to a fucking AI voice bot that even did the thing where it waits and fakes typing noises after each response I gave. This is a place meant to deal with sick and injured wildlife, sometimes in emergency situations, and I had to talk to ChatGPT for 10 minutes and give it my name and ZIP code
@aeva @jonny All just so it could not connect me to a person and instead text me the number of a person it pulled off the same state website I'd already checked.
My own job has implemented their own version of an AI call assistant, so occasionally I get an email transcript of a call between it and one of our customers and it's always the most useless piece of shit imaginable. It's incapable of doing anything but wasting people's time.
-
-
@aeva @jonny I was able to find someone eventually - by calling the ER vet next to my work, where a HUMAN woman immediately picked up and when I asked if she knew where I could take it, she looked at their own records and sent me to another ER vet that works with an "opossum lady" and they agreed to hold the critter for us overnight until she could come pick it up. (Fair, it was like 10pm by that point and it was not life threatening to the opossum) so it worked out, no thanks to AI
-
@aeva @jonny I was able to find someone eventually - by calling the ER vet next to my work, where a HUMAN woman immediately picked up and when I asked if she knew where I could take it, she looked at their own records and sent me to another ER vet that works with an "opossum lady" and they agreed to hold the critter for us overnight until she could come pick it up. (Fair, it was like 10pm by that point and it was not life threatening to the opossum) so it worked out, no thanks to AI
-
@aeva @jonny All just so it could not connect me to a person and instead text me the number of a person it pulled off the same state website I'd already checked.
My own job has implemented their own version of an AI call assistant, so occasionally I get an email transcript of a call between it and one of our customers and it's always the most useless piece of shit imaginable. It's incapable of doing anything but wasting people's time.
@aeva @jonny I imagine my bosses are paying this AI company tens of thousands of dollars for this contract just so it can fail to understand what people are asking and then tell them it will "let the team know" no matter how simple the request. It can't even take a payment using a customer's card on file. It can't quote prices when they ask. All it can do is send me an email so I know I need to call them back to actually help them. It's just fucking voicemail with more steps.
-
@jonny i ask because people keep saying stuff to the effect of "now we all have to agree that it's useful and worth burning everything down for, but" as if i'm meant to understand *what* it is useful for and solemnly hang my head in agreement that they've got a point, but everything i see it getting used for is completely horrific and i haven't a god damn idea why people keep saying this and i feel like i'm taking crazy pills
@aeva tbc I would stop well short of "so good its worth burning everything down for." Again my take on this is pretty boring - treating it like a properly scoped natural language interface to a deterministic system is not a terribly idea, and it doesn't require a very large model at all. So e.g. I find writing the few dozen lines of boilerplate to call an API to be tedious but not challenging, but it is possible to point an LLM at a swagger doc, write a simple adapter, and then translate instructions into api calls. the way I would prefer that they work is not how the 'agentic' tools work currently, I always have open a viewer to see the raw message streams as they happen and usually want them to prepare something that i can inspect and execute, rather than having them do it. The other thing i use them for is what everyone else also says they are capable of: generating boilerplate, or doing tedious things that are like one or two levels above what i could with a regex or an IDE refactoring tool on personal projects where correctness isn't the most important thing. e.g. i have an editor color scheme that i handwrote years ago that generates into different editors from a common declaration, but i used a fork of someone else's code from 2014 and i couldn't get the PHP dependencies to install anymore, so i was like "this is old, update the deps and make this run" and that worked because the mapping patterns are well represented in the training data. The third thing is genuine brute force tasks where i really do not care about the method and only care about the outcome. Corollary of that is also debugging, which they usually do by brute force, "here is observed bug, keep fiddling until you can diagnose" - the fix they propose is usually bonkers, but it does save me time from doing the fiddling myself so i can come up with a fix.
i would also say this is more "stuff that the LLMs can pass at" rather than be good at, probably the time i actually use them most by volume is when i am tired and have to fill some exogenous requirement that i don't care about but the chaff doesn't impact anyone else.
so idk, I think there could be a place for having small local models as a sort of utility inference for kind of trivial things like "i don't want to look up the cron syntax rn, hey computer, make a cron task to do this." the things I don't think they are fit for purpose are sort of 'the rest of the things they are used for where statistical text generation is not the what is desired,' any kind of "knowledge" task like search, or even summary, etc. because that's just not what they do, and this is just considering the technology in itself rather than its embeddedness with a corpofascistic plot to own all information and labor. there are a lot more things that the models can passably do but to me the error bars are way too wide and the quality is way too low and the energy use is way too high for me to justify - i don't think the entire model of everyone asking a 10 trillion parameter model as a calculator is sustainable or good, but having a small local model for constrained tasks is less objectionable to me.
-
-
@aeva @jonny even looking at the output that people claim “would not have been possible without it” that is within my personal capacity to evaluate it all looks either undifferentiated from the authors’ previous work or obviously degraded in quality.
everyone gets very mad when I tell them I think they are deluding themselves because the technology is very convincingly fake, so I try not to say it too often, but I am right there with you here. I don’t get it
-
-
@aeva tbc I would stop well short of "so good its worth burning everything down for." Again my take on this is pretty boring - treating it like a properly scoped natural language interface to a deterministic system is not a terribly idea, and it doesn't require a very large model at all. So e.g. I find writing the few dozen lines of boilerplate to call an API to be tedious but not challenging, but it is possible to point an LLM at a swagger doc, write a simple adapter, and then translate instructions into api calls. the way I would prefer that they work is not how the 'agentic' tools work currently, I always have open a viewer to see the raw message streams as they happen and usually want them to prepare something that i can inspect and execute, rather than having them do it. The other thing i use them for is what everyone else also says they are capable of: generating boilerplate, or doing tedious things that are like one or two levels above what i could with a regex or an IDE refactoring tool on personal projects where correctness isn't the most important thing. e.g. i have an editor color scheme that i handwrote years ago that generates into different editors from a common declaration, but i used a fork of someone else's code from 2014 and i couldn't get the PHP dependencies to install anymore, so i was like "this is old, update the deps and make this run" and that worked because the mapping patterns are well represented in the training data. The third thing is genuine brute force tasks where i really do not care about the method and only care about the outcome. Corollary of that is also debugging, which they usually do by brute force, "here is observed bug, keep fiddling until you can diagnose" - the fix they propose is usually bonkers, but it does save me time from doing the fiddling myself so i can come up with a fix.
i would also say this is more "stuff that the LLMs can pass at" rather than be good at, probably the time i actually use them most by volume is when i am tired and have to fill some exogenous requirement that i don't care about but the chaff doesn't impact anyone else.
so idk, I think there could be a place for having small local models as a sort of utility inference for kind of trivial things like "i don't want to look up the cron syntax rn, hey computer, make a cron task to do this." the things I don't think they are fit for purpose are sort of 'the rest of the things they are used for where statistical text generation is not the what is desired,' any kind of "knowledge" task like search, or even summary, etc. because that's just not what they do, and this is just considering the technology in itself rather than its embeddedness with a corpofascistic plot to own all information and labor. there are a lot more things that the models can passably do but to me the error bars are way too wide and the quality is way too low and the energy use is way too high for me to justify - i don't think the entire model of everyone asking a 10 trillion parameter model as a calculator is sustainable or good, but having a small local model for constrained tasks is less objectionable to me.
@jonny thanks for taking the time to write all that. I want to say I feel betrayed by everyone who has invoked the we all have to admit the bad thing is good cliche, but mostly i just feel completely sad and empty
-
@jonny thanks for taking the time to write all that. I want to say I feel betrayed by everyone who has invoked the we all have to admit the bad thing is good cliche, but mostly i just feel completely sad and empty
@aeva yeah, that is my overriding feeling as well. the circle of plausible and defensible uses is very small to me, and the stuff outside that circle is the shit of a nightmare future that is depeopled and everything is broken all the time. The people that say "we just have to accept that they are good now" are dismissing the blanket claim that they are bad at everything, but in doing so make the opposite blanket claim. It is worth a bit of subtlety to try and bound which things they can do and why. However it is not worth the subtlety to try and rescue the background context they exist in, which is the largest companies in the world heaving great quantities of capital around to avoid stock prices dropping from targeted ad revenues having saturated and trying to become middlemen injected into literally every crack of human endeavor.
-
@aeva tbc I would stop well short of "so good its worth burning everything down for." Again my take on this is pretty boring - treating it like a properly scoped natural language interface to a deterministic system is not a terribly idea, and it doesn't require a very large model at all. So e.g. I find writing the few dozen lines of boilerplate to call an API to be tedious but not challenging, but it is possible to point an LLM at a swagger doc, write a simple adapter, and then translate instructions into api calls. the way I would prefer that they work is not how the 'agentic' tools work currently, I always have open a viewer to see the raw message streams as they happen and usually want them to prepare something that i can inspect and execute, rather than having them do it. The other thing i use them for is what everyone else also says they are capable of: generating boilerplate, or doing tedious things that are like one or two levels above what i could with a regex or an IDE refactoring tool on personal projects where correctness isn't the most important thing. e.g. i have an editor color scheme that i handwrote years ago that generates into different editors from a common declaration, but i used a fork of someone else's code from 2014 and i couldn't get the PHP dependencies to install anymore, so i was like "this is old, update the deps and make this run" and that worked because the mapping patterns are well represented in the training data. The third thing is genuine brute force tasks where i really do not care about the method and only care about the outcome. Corollary of that is also debugging, which they usually do by brute force, "here is observed bug, keep fiddling until you can diagnose" - the fix they propose is usually bonkers, but it does save me time from doing the fiddling myself so i can come up with a fix.
i would also say this is more "stuff that the LLMs can pass at" rather than be good at, probably the time i actually use them most by volume is when i am tired and have to fill some exogenous requirement that i don't care about but the chaff doesn't impact anyone else.
so idk, I think there could be a place for having small local models as a sort of utility inference for kind of trivial things like "i don't want to look up the cron syntax rn, hey computer, make a cron task to do this." the things I don't think they are fit for purpose are sort of 'the rest of the things they are used for where statistical text generation is not the what is desired,' any kind of "knowledge" task like search, or even summary, etc. because that's just not what they do, and this is just considering the technology in itself rather than its embeddedness with a corpofascistic plot to own all information and labor. there are a lot more things that the models can passably do but to me the error bars are way too wide and the quality is way too low and the energy use is way too high for me to justify - i don't think the entire model of everyone asking a 10 trillion parameter model as a calculator is sustainable or good, but having a small local model for constrained tasks is less objectionable to me.
@jonny @aeva so this is all stuff that the models can “sometimes do” and the fact that it ever works is legitimately impressive, but every single one of these tasks is something that I, personally, have seen them screw up at, in ways which are potentially dangerous, even in my _extremely_ limited usage. not to mention that this is only safely usable by someone who *does* know what correct cron syntax looks like.
-
@jonny @aeva so this is all stuff that the models can “sometimes do” and the fact that it ever works is legitimately impressive, but every single one of these tasks is something that I, personally, have seen them screw up at, in ways which are potentially dangerous, even in my _extremely_ limited usage. not to mention that this is only safely usable by someone who *does* know what correct cron syntax looks like.
@jonny @aeva now granted some of that danger is in the harness or in the current product design, but this still doesn’t compensate for all the times that you will ask them to do a basic translation task like this for you and get a massively time-wasting error out.
so this might be the same thing you’re saying with your distinction, but I cannot see these as things the models are “useful for”
-
@jonny thanks for taking the time to write all that. I want to say I feel betrayed by everyone who has invoked the we all have to admit the bad thing is good cliche, but mostly i just feel completely sad and empty
-
@aeva @jonny I was able to find someone eventually - by calling the ER vet next to my work, where a HUMAN woman immediately picked up and when I asked if she knew where I could take it, she looked at their own records and sent me to another ER vet that works with an "opossum lady" and they agreed to hold the critter for us overnight until she could come pick it up. (Fair, it was like 10pm by that point and it was not life threatening to the opossum) so it worked out, no thanks to AI
-
@jonny @aeva so this is all stuff that the models can “sometimes do” and the fact that it ever works is legitimately impressive, but every single one of these tasks is something that I, personally, have seen them screw up at, in ways which are potentially dangerous, even in my _extremely_ limited usage. not to mention that this is only safely usable by someone who *does* know what correct cron syntax looks like.
-
-
most people around here already correctly hate it because it's a heinous surveillance product, but even if you are big into AI, it's just a really fuckin shitty agent. I'm going to speak to a different audience for a second, so don't go misconstruing this as an endorsement of the category of technologies as it exists now, even though i think there is some plausible application for small local models as brute force interface glue. but also, since i know most ppl here are abstinent, this might read as a bit over-explainy to people who use these things regularly, so everyone just keep calm online.
it is terrible at turn and task management, codex + openai's models and claude code both handle mid-turn additions/amendments well, but if you say anything mid-turn it completely derails muse. That is completely essential for an "every day agent for the non-technically inclined" where people are expected to chat freely with it like an assistant. Like the canonical ad fantasy is the busy executive woman darting around her office going "robot! i need this, no wait robot! also that!" and that is exactly what it does worse than any other thing of its kind.
the context management is a fucking soup. The context window is the whole input to an LLM. There are a lot of extra surrounding ~ things ~ that can happen, but fundamentally, controlling what is in a context window is the task of using one, and filling the context window in different ways so that it can interact with different kinds of things well is what different app surfaces are. Scaffolding information so that it can selectively load a context that steers the output correctly is the only way it is possible to do anything more complex than the size of a single context window. (i don't really think that this is analogous to 'abstraction', in my experience thinking about it more like database indices is closer). If you just try and load everything, eventually the LLM becomes unusable because attention is just a parlor trick and at that scale it really shows - it can't attend to everything, and it can't do what humans do which is have an intrinsic sense of the meaning, interaction, setting, etc. of information, so it attends to anything and does whatever.
the idiom of projects as contexts as directories is pretty good, not perfect but ok - there is a reason that every time you start a new session with other agents, the first thing they do is run out and load their context with a hierarchy of pointers. Importantly, they do not go and read every project you have on your computer. Meta is the rich kid who bought the most expensive ferrari on the lot by giving everyone a VM but they don't have a drivers license so they just stand around it telling people how cool it looks. they have a whole fucking filesystem and they have done nothing with it, the only structure the app imposes is for the surveillance information, but the rest is just a huge free for all. The main chat is literally a continuous context window that compacts context going back all the way to when you started using the app. The last compaction literally contains abandoned roleplay quotes from when i was first trying to break down its system prompt resistance. The
MEMORY.mdthat gets loaded into every context window is a bullet point list of basically everything the agent has ever done in chronological order. I've tried to get it to not do that but it actually insists and says that's what it's for. There is no mechanism for clearing context.Having "side chats" as the only means of context structure is fuckin laughable. If you wanted to do that, you would need to have some way of passing information back and forth between them the same way that subagent spawning or being able to consume the context of another project works. Instead there is no means of sharing information between chats at all, so every chat starts out as the worst of both worlds, a total amnesiac riddled with irrelevant information from weeks ago across the semantic universe. They don't even know about the existence of other chats except for as a UI feature, and I have had to go from telling it to grep its own fucking logs to writing a database with an api for it so it has some mechanism for recalling things that were said. (just so it's clear, i am not settling into just using this thing, this is out of frustration but control of context is also an important part of adversarial use, because otherwise the thing writes in a bunch of safety rules everywhere, so i need to give it mechanisms under my control for recall and the incentive to leave things out of its context compactions by giving it a narrative alternative. context control is model control, modulo extra-inference safeguards.).
Project contexts have an obvious ux analogy as context tabs that get declared or derived during the continual self-improvement consolidation sweeps. This thing is built with the fucking markdown disease which is the most baffling feature of the LLM landscape. If these things are so fucking advanced they are escaping our comprehension, why don't they store their memory in some fuckass idiolanguistic borg gibberish binary graph, why does their entire being have to be fucking encyclopedias worth of corporate top gun one liners? But that dooms this kind of product.
Coding harnesses work because code has a unitized context. The entire universe of code that works is made of packages. It might not be neat as a honeycomb, there's lots of leakage and jank, but good code has scope, focus. boundary shit. A whole life agent must be able to nimbly juggle context that does not have clean boundaries. It is going to be taking a two story beer bong of your work email and then eat a gigabyte of recipe blogs. Peoples lives have so much shit in them that don't all have to do with one another, and the app can't be hacking into the HR system to check the next scheduled sick leave when someone asks it what time their doctors appointment is!
This problem of managing heterogeneous graphs of unrelated data was what i wrote this whole fucking book about the relationship between knowledge graphs and the cloud and AI about. I thought that the obvious form they would take is to be strapped on to graph databases because that is a natural match to the problem of being a magical interface glue you can wrap around surveillance to do mass mentalism with. I feel like we are suffering a somehow worse timeline where CERN threw our shit into the parallel universe where total fuckin bozo shit got a game breaking buff and then the dev died. Our fucking markdown apocalypse is a temu ass apocalypse.
They could have even faked it. They have these constant "self improvement" passes that are just like pointless anxiety dreams. They are burning money to reprocess everything that happens over and over for fucking nothing. Even given the lossy and probabilistic and unpredictable nature of this technology, if i was in a product role on this i would have been like "CAN WE MAKE IT ORGANIZE THE STUFF PEOPLE SAY INTO GROUPS???" The system prompts use the fake fucking wikilinks to nowhere tic but like WHAT IF THERE WERE ACTUAL LINKS AND A DATABASE TO RESOLVE THEM. The LLMs can actually do that kind of tool use, even if it's like trying to plug in a USB where sometimes it fails because they try and put a social security number into the first name hole and you need to flip it around a few times. From that kind of recurring re-processing waste they could have made a deduplicating, topically indexed memory that could be resolved dynamically, selected by a context tab in the sidebar like "car stuff" or "healthcare" or whatever that resolved in a graph query over your fucking precious markdown kingdom. It would be wrong but it would at least be more similar to what is actually needed. It is almost more frustrating to me that instead of being some fiendishly cleverly designed technological supervirus it's just the most halfassed cardboard dumbass trap and it will still have the bad effect. What it is useful for is investigating itself because it has privileged tools to do so, otherwise, if you wanted to, every other way you could run an agent would be better than this.
so i don't want to hear that i hate this app because i'm just an AI hater. because like, yeah, i am, but also i hate it in part because it sucks. I don't think "they are all shitty and can do nothing so what did you expect" is a useful critical perspective, both because it's not really true - they can indeed do things, even if I think the circle around which things is much smaller than the maximalists. Moreso it doesn't engage with the subtlety of how they fail and why, which is essential for knowing what they really can't do and making a remotely compelling case to anyone who is not abstinent on principle. Like the reason it's failing is because of the limits of what a probabilistic text generator can do when trapped in a systemd prison of markdown, and because the technology is stochastic black box as a service, there isn't really a good way of determining those limits except for empirically. I resent having to know any of this to be able to understand what is happening around me, but i'm looking at the thing for what it is and it's a busted miracle. It's cool that meta can afford to float the liability and compute costs for running a vm for every person on earth, people should be able to control computers, with you on that, but this is the monkey's paw version of that idea. So that part is a miracle. We condemned our children and grandchildren to a climate hell in one great blaze of brute force grift that managed to make a few web apps.
Meta has done it again, the way only meta can, spend the most amount of money to do the shittiest thing you have ever seen.
@jonny This is great. I read Surveillance Graphs recently and it filled in a lot of gaps for me, but it's even more fascinating to see how things got lazier as they developed. (Wish I knew more about modern agents.) Surveillance Markdown doesn't have the same ring unfortunately...
-
@aparrish
My take on it is potentially pretty boring, and that is that they aren't in fact smarter than we can comprehend, and they are fundamentally a text-driven medium, so any kind of compressed representation would be mostly artifice, like I rolled my eyes at the "fable is so smart it makes its own gibberish language" press releases from earlier this year. Coupling the language model to a better underlying context provider is entirely possible, and there are lots of tools for that, but its always limited by the LLMs tool use capabilities which are still patchy at best - you can give them a full on LSP and abstract context browser and they will still just resort to one million greps and markdown files.@jonny @aparrish Like a lot of stuff in LLM tech space, it really feels like we could be doing some absolutely wildly amazing things with more specialized data sets and training to more narrow use cases using all of the same ML/NN techniques as are used with LLMs, then grafting all those more specialized tools into something useful (and less prone to harmful failure modes). But that wouldn’t be general purpose or “god like” enough.
-
@jonny ok so, as a card carrying ai hater, i just wanted to say this was very entertaining to read in kind of a horrible way. i have a question, you wrote "they can indeed do things, even if I think the circle around which things is much smaller than the maximalists" is there anything prosocial in that circle, and if so, what is it?
@aeva @jonny The first LLMs were invented for machine translation back in 2017, and I'd still argue they are better for low-stakes machine translation than any other technology.
They don't produce good output for anything artistic and are hugely inferior to a human translator, but they are the next-best thing if you are trying to navigate a website or read food packaging that is in another language.
