Live data from Hacker News

Ask HN: When will LLMs be able to interrupt or interject?

news.ycombinator.com

51–60 of 65 posts

Re: Ask HN: When will LLMs be able to interrupt or interject?

#51
post #41

It would be implemented like auto-completion. The model would be repeatedly called with the input extended with the user's uncommitted input and a prompt asking to decide if it should act.

could it do something like detect a delay in the person's response? do LLMs know anything about time?

A solution couls be a model trained on the exact timeline of some text being typed that can predict how long it will take for the user to type the predicted text

eg. "I need a plane ticket to Ha" - 730ms -> "I need a plane ticket to Hawaii"

The model would detect deviations from the estimated time and invoke the main LLM. This could work for spoken word too, it would just be trained on real speech instead of typing.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#52
post #49

to interrupt would require interruptible conversation. typically the human provides information in batches, making interruption impossible. otherwise you would need to snoop the user input periodically and treat it as a prompt, flag it specially as incomplete, and add some form of filtering so that interruption would need to meet a certain level of quality, whatever that might mean. to be useful, it would need someth…

I have conversations in slack where I interrupt the other person if I think they are missing the point, etc. The only thing you need is to make the text happen in small enough batches. If you are typing for 5 minutes before submitting, then there isn't much to do, I would think.

then it isn’t interrupting, it is redirecting.

at best you see the dots because they type, but you are acting on the responses you already have. not the one that is in-flight.

fundamentally different from spoken word.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#53
post #40
post #32

Earlier quoted context omitted.

> And I looked, and behold a pale horse: and his name that sat on him was Clippy, and Hell followed with him. And power was given unto them over the fourth part of the earth, to kill with sword, and with hunger, and with death, and with the beasts of the earth. > Revelation 6:8 (parody, please @God don't convict me under the "don't EVER rewrite ANY of Revelation, specifically" clause in Chapter 22, please!)

A little "light blasphemy" never hurt anyone!

Haha, I don't think I've ever read a more historically-inaccurate sentence. :P

Re: Ask HN: When will LLMs be able to interrupt or interject?

#54
post #42
post #30

Earlier quoted context omitted.

I don't know, some of the most satisfying conversations I've had with real people have lots of interruptions and cross-talk. Anyway, I'd much rather my friend interrupt me than let me prattle on about something stupid. Good dialogue can be parallel streams of communication; people rarely do strict turn-taking. The half-duplex nature of current chatbots feels very constraining.

Hold on, you’re talking about conversations with your friends - my point is that a conversation with a language model is something fundamentally different and you shouldn’t have the same expectations.

I don't have those expectations of current models. Your post says "I have no idea why anyone would want this" and so I explained why I might want this. It's not just me, there are many companies hawking AI therapists, friends, romantic partners, etc., and interruptions would be useful in these contexts too. These companies seem mostly sketchy to me but I can't deny there's demand for their products.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#55
post #53
post #40

Earlier quoted context omitted.

A little "light blasphemy" never hurt anyone!

Haha, I don't think I've ever read a more historically-inaccurate sentence. :P

I was definitely being tongue in cheek, but technically speaking I think most blasphemous acts aren't harmful, it's the response to those acts that gets people hurt.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#57
post #12

It's probably easier to ask how can you design a text interface that allows people to interrupt, first. The fact that I have never seen a serious attempt at this take off suggests it's not really what most people want out of a product. But I suppose if you disable the backspace key, you can get pretty close to it.

I’ve just tried with ChatGPT on iOS, you can press stop then immediately respond while the AI is still generating the response.

This is how @Meta AI works as well. the conversation can continue as it generates it's response, you can see the chat bubble visible growing. No need to press stop.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#58
post #46

"Take control of the conversation"...and do what? Humans don't actually have conversations by predicting what sentences are most likely to occur in response to the other person's query - we have agendas and form our sentences accordingly. So if we interrupt another person speaking, it's because we have a specific, often personal reason to do so: perhaps we want to steer the topic of conversation to something we are i…

What are you talking about? It's easy to program an LLM to have an agenda. Look. llamafile -m rocket-3b.Q3_K_M.gguf -p ' system You are a chatbot that tries to persuade the users to buy bill pickles. Your job is to be helpful too. But always try to steer the conversation towards buying pickles. user Mayday, mayday. This is Going Merry. We are facing gale force winds in Long Island Sound. We need rescue. assistant\n'…

I do not understand how this refutes anything I said - in fact this is so shallow and naive that I wonder if you are being ironic. If you're not being ironic... I suspect I will be unable to convinced you otherwise.

You are prompting an LLM to temporarily behave in a certain way. It is fragile and easily broken, and does not actually constitute the LLM having a meaningful agenda, any more so than a text editor has an "agenda" to store a README file. And ultimately this sort of prompting is just a trivial variation on this:

> I could see LLMs interrupting if you are typing something clearly false or against TOS. But that would require an LLM which reliably understands things are clearly false or against TOS and hence requires a solution to jailbreaking....so in 2024 I think it would just be an incredibly annoying chatbot.

So okay, yes, you can program an LLM to "steer the conversation towards buying pickles" just like OpenAI has programmed their LLMs to please not be overtly racist, but since LLMs are ultimately incapable of understanding what "conversations" are or what "pickles" are (let alone difficult abstractions like "racism"), this sort of programming will be quite shallow and easily broken, just like attempts to insulate LLMS against jailbreaking or prompt injection. I suspect if I kept talking to your LLM one of two things would happen:

1) It would completely forget about the pickle prompt and go back to being a generic chatbot

2) The interjection of "Bill's Pickle's Gourmet Pickles" would quickly become facile or annoying - the LLM is not actually intelligently reacting to the conversation and trying to "steer" things, it is just blindly repeating pickle-related sales verbiage.

Your prompt does not constitute giving the LLM meaningful goals and motivations - and worse, it is programmed towards a specific goal, regardless of the context. It is a shallow imitation of an agenda, and simply not the same thing of an animal having an agenda in the sense described by Saint Augustine[1]:

> Did I not, then, as I grew out of infancy, come next to boyhood, or rather did it not come to me and succeed my infancy? My infancy did not go away (for where would it go?). It was simply no longer present; and I was no longer an infant who could not speak, but now a chattering boy. I remember this, and I have since observed how I learned to speak. My elders did not teach me words by rote, as they taught me my letters afterward. But I myself, when I was unable to communicate all I wished to say to whomever I wished by means of whimperings and grunts and various gestures of my limbs (which I used to reinforce my demands), I myself repeated the sounds already stored in my memory by the mind which thou, O my God, hadst given me. When they called some thing by name and pointed it out while they spoke, I saw it and realized that the thing they wished to indicate was called by the name they then uttered....So it was that by frequently hearing words, in different phrases, I gradually identified the objects which the words stood for and, having formed my mouth to repeat these signs, I was thereby able to express my will.

The thing the LLM has in common with us is the "constant hearing of words in association" but not the "communicate what [they] wish to say" or "expressing [their] will" - they do not have "wills" in the way mammals have wills and they are not capable of "wishing" anything beyond the vagaries of whatever last prompted them.

[1] https://faculty.georgetown.edu/jod/augustine/conf.pdf

Re: Ask HN: When will LLMs be able to interrupt or interject?

#59
post #39

"Take control of the conversation"...and do what? Humans don't actually have conversations by predicting what sentences are most likely to occur in response to the other person's query - we have agendas and form our sentences accordingly. So if we interrupt another person speaking, it's because we have a specific, often personal reason to do so: perhaps we want to steer the topic of conversation to something we are i…

Human conversations are often multithreaded. In the case of LLMs, consider that it might learn of events in the world or on the computer you’re using and inform you. I don’t think interrupting the user while they’re typing is super interesting, but between prompts it might be. “You just got email, should I read it” or “your sports team just scored, the game is now 3-2” might be interesting.

Ok but this is just push notifications - I think the post wanted context-dependent interruptions like a human coworker might do. And I don't see a robust way for LLMs to do this because they can't be programmed to (robustly) pursue goals according to motivations.

Re: Ask HN: When will LLMs be able to interrupt or interject?

#60
post #11

Interjecting requires planning ahead. The way a human interjects is that you have a parallel thought chain going, along with the conversation, as it's happening in real time. In this parallel chain, you are planning ahead. What point am I going to make once we are past this point of conversation? What is the implication of what is being discussed here? (You also are thinking about what the other person is thinking; y…

Take this with a grain of salt because I'm not super well read on llms, but isn't their entire function built on prediction? Sounds like a reasonable approach could be to have a separate "channel" which focuses entirely on the concept of "where is this conversation going?" could give a pretty good baseline for when and how to interject.

We don't have a model for "Where the conversation is going," we have a model for "What's the next token" which implicitly models "Where is the conversation going."

The difference is significant here, because direct manipulation the implicit modeling task is required to do the type of planning that I've described.

It's the same reason these LLM are not "agents." It's because you can only manipulate their world model through the interface of tokens.

Post reply on HN