> The major hurdle right now is actually pivoting LLMs from just generating code: integrating those tasks into workflows.
Funny, I thought that the major hurdle is improving accuracy and reliability, as it's always been. Engineering is necessary and useful, but it's a much simpler problem, which is why everyone is jumping on it.
As much as that’s true it’s clear a huge amount of people have accepted the current state and are working around it, successfully(in terms of ticking an executive’s checkbox) in a lot of cases. And it’s worth considering we’re seeing strong strides outside of model quality in the tooling and integration
It's ok friend, all I did is put acceptance criteria in a list so I can parse it and quickly track cross-references. The rest is just Elixir/Phoenix and some creative writing.
> your extreme atheists aren't much different from your extreme believers; they both have strong beliefs about things they can't prove, and for some reason want to go off on them.
You have a mistaken understanding of what atheism is. It is not a belief in anything, but an absence of belief in a deity.
> there are a whole lot of things that we all go around everyday "not believing."
Sure, and yet theism is part of 75% of the world population and influences everything from education to politics. It's perfectly reasonable to talk about atheism within appropriate settings.
>You have a mistaken understanding of what atheism is. It is not a belief in anything, but an absence of belief in a deity.
I consider that to also be a wrongly held position, because you'd need proof either way. So atheists are just making a bet.
I think agnostic is the most valid position as far as I am concerned, lacking proof of one or the other. I do not know. We can get into technicalities as well. What exactly do we mean by God? What if some religious God does exist but it's wrongly interpreted by believers? What if there's some highly technologically advanced entity that meets the criteria as far as the more primitive religious perspective is concerned? Do we have proof such thing exists? Do we have proof such entity cannot exist in our universe? I find both perspectives shortsighted.
Having certainty something that can be perceived as God by believers cannot exist in our universe is in the end a belief, with no proof.
> I consider that to also be a wrongly held position, because you'd need proof either way.
Proof that something doesn't exist? Ever heard of Russel's teapot?
The burden of proof is on the claimer.
> What exactly do we mean by God?
Absurd question. Pick up any holy book, or ask any believer. An atheist is simply a person who doesn't hold those beliefs.
A famous Dawkins quote is apt in this discussion:
> We are all atheists about most of the gods that humanity has ever believed in. Some of us just go one god further.
> Having certainty something that can be perceived as God by believers cannot exist in our universe is in the end a belief, with no proof.
Again, you're mistaking what atheism is. It's not being certain that a deity cannot exist—it's not having any reason to think that it does.
People who claim certainty in either direction are equally delusional. The problem is when a belief crosses into realms of reality, defines the identity and culture of people, and influences the rest of society. Based on history and personal experience, theists are far more prone to this than atheists.
The word seems to be used both ways, despite what anyone might like: either as a person who doesn't believe in a god, or as a person who believes there is no god. It's a subtle difference.
Again, I'm more talking about "atheism in effect" than I am "textbook definition."
I'd say, sure, it's reasonable to discuss, but it's often overblown. The older I get, the less significance "atheism" feels like it has just about anywhere.
No shade to anyone, but I have very little interest in "internal beliefs that seem not to affect people much," which atheism seems to fall square into.
From what I understood, any color and material involved in high precision manufacturing requires careful design and thorough testing. They likely prioritize the brown color and material due to branding, so changing this to anything else requires redoing large parts of the pipeline.
You have to redo injection moulds anyway as they have a limited life. And you can do a lot with materials too, some materials simply shrink more than others as they cool down.
I'm glad companies like Noctua exist that put so much thought and care into their products. I don't even mind being advertisted to when that's the case.
I'm exhausted by these shiny vibe coded projects that overpromise and underdeliver.
Knowledge comes from doing the hard work, not from being spoon fed information. All these fancy graphs represent a tentative mental model produced as a result of research and learning. Everyone's model is different based on their own experience and focus, so trying to present it as a unique map will more than likely not be conducive to understanding at all. Besides the fact that it will almost certainly miss important details or be hallucinated.
HN users: stop upvoting and promoting this garbage. HN mods: please give us tools to label and filter this content.
There's no such thing, but all content publishing web sites should at the very least provide tools for users to self-moderate, which this forum heavily relies on anyway.
Now that the internet is flooded by machine-generated content, which is often published and promoted autonomously as well, all content should be scanned and labeled with a value that indicates the likelihood of it being machine-generated and published.
I'm thinking of JSON fields like `machine_gen_probability` and `machine_pub_probability` returned by the API. Then the frontend should expose settings to show these labels next to each post and comment, and filtering rules to decide what should be done with content above a certain value (hide, adjust feed rank, etc.). Some people might even want to boost this content, for whatever reason, so making the system flexible would be smart.
The scoring system of course won't be perfect, but I figure that a company like YC should know a few talented individuals that could do a solid job of implementing this. They've certainly profited from investing in companies that cause this problem.
But... considering HN is merely a promotional tool for YC that runs on limited resources as it is, I wouldn't hold my breath that such a system would ever be implemented. So all we're going to get are changes to "guidelines", and hope that the system won't be abused. Which is laughably naive in this day and age. So this forum will most likely be overrun by the noise, and end up with minimal participation from reasonable humans, as is happening and will continue to happen on most online platforms.
The idea that a tool intended to replace all human cognitive work is the next level of abstraction is so fundamentally flawed, that I'm not sure it's made in good faith anymore. The most charitable interpretation I can think of is that it's a coping mechanism for being made redundant.
Nevermind the fact that these tools are nowhere near as capable as their marketing suggests. Once companies and society start hitting the brick wall of inevitable consequences of the current hype cycle, there will be a great crash, followed by industry correction. Only then will actually useful applications of this technology surface, of which there are plenty. We've seen how this plays out a few times before already.
> So this is proof of the models actually getting stronger (previous generations of LLMs were unable to solve this one).
No, it's not.
While I don't dispute that new models may perform better at certain tasks, the fact that someone was able to use them to solve a novel problem is not proof of this.
LLM output is nondeterministic. Given the same prompt, the same LLM will generate different output, especially when it involves a large number of output tokens, as in this case. One of those attempts might produce a correct output, but this is not certain, and is difficult if not impossible for a human not expert in the domain to determine this, as shown in this thread.
This is one of a number of such results achieved only in the last few months with only the last crop of models. They have undoubtedly gotten better in this domain. Saying anything else is just denial. You can run these same problems on GPT-4 or 5 all you want, you'll get nowhere. In fact people did, and you're hearing about it now because it's these crop of models that are getting meaningful results.
As others have pointed out, a key part of the prompt used here may have been "don't search the internet" as it would most likely have defaulted to starting off with existing approaches to that problem...
Funny, I thought that the major hurdle is improving accuracy and reliability, as it's always been. Engineering is necessary and useful, but it's a much simpler problem, which is why everyone is jumping on it.
reply