ryanklee

The saltines is basically inappropriate.

The problem is with the person, not the tool.

And I didn't suggest how to do so. I implied that there is an issue with the quality of their coworkers, not ChatGPT.

That sounds to me less of an issue with ChatGPT, and more with having colleagues that don't understand how to engage in reasonable discussion or evaluate information correctly.

What matters is whether the suggestions are actually good, not where they came from.

I don't understand what you're pushing back against, but I'm not standing there.

I assumed it was in reference to llama.cpp. It's a weak assumption, though.

It's an LLM technology that allows certain models to run on CPUs rather than big beefy GPUs. Makes running locally viable for consumers.

> The most charitable explanation

We need to stop being charitable to known malignant actors like Musk. We should assume the worst and those with leverage should oust him from public life.

Again, I acknowledge that citations have value and are important and desirable as a feature of how we communicate.

But citations did not begin their rise a standard until the 1900s. The idea that knowledge production was not valuable until that point is absurd.

Further, citations do not prevent statements from being incorrect. They are sometimes used to lend credence to bullshit.

We live all at once in a world full of citations and of epistemic chaos.

Do I want LLMs to cite sources? Absolutely. But it's not the fundamental path to value people seem to sometimes think.

Imagine being an actual woman and having to deal with this bullshit all the fucking time. I have no idea how the hell they do it. Must be low-level torture.

I didn't ignore it. But we do not require citations for every statement. Statements have value, citations notwithstanding.

I don't have a position on that, and there's not one implied by my comment, which is why I am unsure why the commenter assumes one.

This is not correct. You can get an LLM to improve reasoning through iteration and interrogation. By changing the content in its context window you can evolve a conversation quite nicely and get elaborated explanations, reversals of opinions, etc.

There is a ton of arrogance everywhere on this subject. Anyone who doesn't see how complicated and nuanced and difficult figuring out what the hell is going on here really needs to sit down and think a little more carefully. There's absolutely no room for out-of-hand dismissal. We are all babes in the woods right now. We need each other, like always.

> cannot reason anybody out of something that they want to believe.

Especially when your own reasons are not the slam dunk you think they are.

I totally fail to see how you got that from my comment.

We shouldn't need his alleged drug use as an excuse to eject Musk from public life and positions of power and responsibility.

His abhorrent and disgusting behavior should be enough on its own.

Please spell this out for me.

I'm a very tech savvy individual, and as a parent who has spent hours and hours navigating the murk of disparate parental control systems, it is clear to me that what you say is only true if you commit to preventing your child from having access to anything at all.

Worst of all, children have friends!

In other words, you must lock them in a closet and deprive them of access to sunlight itself to accomplish what you say.

It absolutely is useful and one of the most significant points in this comment section.

People are applying standards to LLMs that don't exist elsewhere. They don't exist elsewhere because they absolutely can't exist elsewhere.

It's not a technical short-coming of LLMs that they can't produce citations in every instance. Rather, it's a property of information, representation, and knowledge itself. Much of it floats far above the otherwise load bearing pillars of citations.

People contend with this constantly in every arena of life, and we have come up with very elaborate ways to offset the difficulties caused by it.

And we get along very well despite it all.

I'll also leave you with this: citations are just pointers to more sources of information. Not some ground truth. It's just another tranche that requires evaluations.

Lastly, your comment is an unfortunate bit of low-level snark and probably would have been better left unsaid.

What are the citations actually required for though?

Another way to ask this is: what value remains without them?

I'll add this as well: humans produced valuable knowledge for thousands of years without the use or standard of citations.

To be clear, I think citations are highly valuable and desirable and I very much want LLMs to cite when appropriate. However, I think the necessity of this is overstated.

Edit: what you said of experts can be said of LLMs as well.

> I never understood how many of the other kids were starting to smoke

Human behavior is not generally a result of reasoning.

Based on what?

That is not correct.

You can interrogate an LLM. You can evaluate its responses in the course of interrogation. You can judge whether those responses are coherent internally and congruent with other sources of information. LLMs can also offer sources and citation for RAG operations.

There is no self-sufficient source of truth in this world. Every information agent must contend with the inevitability of its own production of error and excursions into fallibility.

This does not mean those agents are engaging in bullshit whenever they leave the domain of explicit self-verification.

I'm not referring to hallucinations.

I'm referring to novel relationships drawn between datum in the corpus that are a result of training and inference.

This is apparent in something as simple as a summarization.

We don't only reason from facts. We also reason from value.

Further, reasoning that rests on facts that does not cite facts still has massive utility. (See people, all day long.)

Citations are useful, but not required.

> Nearly everything published

What a bald exaggeration

I don't think that's necessarily the case.

I get the strong impression that even people who have much experience using LLMs have astoundingly little insight into what they are actually witnessing. This is often paired with astoundingly little insight into what's actually going on in their own cognitive processes.

Somehow, it's still not clear to most people that LLMs and even vector databases create knowledge that wasn't in the original data.

In fact, that's most of what they do! Isolated, non-novel direct quotation is the exception, not the rule.

Synthetic knowledge is not bullshit.

Synthetic knowledge refers to propositions or truths that are not just based on the meanings of the words or concepts involved, but also on the state of the world or some form of experience or observation. This is in contrast to analytic knowledge, which is true solely based on the meanings of the words or concepts involved, regardless of the state of the world.

The additional context doesn't do any work.