mjg2
Like how I gestured in my top-level comment, are we being an economic analysis or seeming an economic analysis. I regard the efforts of Frontier Labs with suspicion because a lot of their marketing efforts around "Artificial Intelligence" appear (to me) as acquisitive and power hungry to a fault: they are all consuming in that every author on this paper has a superscript 1 to denote association with OpenAI.
When I read artifacts by Frontier Labs I notice a trend towards deteriorating lexicographic architecture. Like how a textbook or larger tomb is constructed for the reader to navigate.
Instead, now I see continuous monotonic text. It’s as if the author expects us to chuck the artifact into an LLM in order for it to inform us of meaning.
Whenever I see these Frontier Labs "publish a paper" on ArXiv.org the phrase Esse Quam Videri strides past my mind—is this genuine progress or merely a marketing metric for their stakeholders?
- https://fortune.com/2025/10/08/deutsche-bank-and-jeffrey-eps...
- https://www.reuters.com/sustainability/boards-policy-regulat...
Plainly, based on the current ground trodded and the trajectory laid out by the frontier AI labs, I do not have concrete evidence/proof of sentience having emerged from LLM-based software tools as of June 4 2026 nor do I expect it to happen in the future based on my understanding and observations of this technology. I'm not excluding the possibility but wielding skepticism. I am open to being proven wrong with new discoveries.
Which is why (to return to my lashing of the dead horse) I don't see OP's post as worthwhile. Their post reiterates a point that is already valid (the prior art) with no new substantial discovery. Which is why "unoriginal and pointless" is apt, a novel idea was not presented; it's just some vain virtue signaling.
So I still disagree with your elucidated point (as you end with "which is valid"): the OP author is using prior art fiction to bolster their opinion of LLM-based software tools as being a possible vector of sentience, not to disarm our chauvinism like the original author intended. If OP wanted to make that point, they could have written a critical essay instead of farming out their thoughts as tokens.
But still, I look forward to reading the book you suggested to understand and appreciate your perspective more.
My original comment (roughly "there's no intelligence in this article, nor sentience in LLMs") is in response to the blog post's buried lede (that the cumulative activity of LLMs has accrued to a weight of "AGI is around the corner" or "there is artificial consciousness in this matrix").
To be clear, I'm not saying LLMs are useless or a wrong direction in development of "AI," but rather it's the Fool's Gold for the path towards AGI, the pursuit of the academic field of Artificial Intelligence research. A research that I've been abreast of for years before this new age of language models that has made everyone with a keyboard an arm chair expert.
Also, thank you for the book recommendation, it's on my list! :)
> Weights helped me draft and proof this story.
Any HN reader here now, I encourage you to read the original ( https://www.eastoftheweb.com/short-stories/UBooks/TheyMade.s... ) in one sitting, go about your day, then read it again. Maybe make some notes on personal critical questions.
Now read the post's topic again ( https://maxleiter.com/blog/weights ) and reflect on the prior fact that weights helped [the author] draft and proof this story.
My reaction (and I'm sorry that it is harsh according to some) is that there is no intelligence found in either the author nor their tool. This is extreme navel gazing, based in science fiction, wanting (wishing) to believe those stories to be true.
I'm skeptical of AI sentience because we must do our due diligence, not because it's impossible. Skepticism is the only respectful approach because to grant sentience is a step away from granting rights.
I’m not going to keep engaging with someone who makes wild assumptions about my stance and accuses me of lying, ignorance, and evil. I was really hoping for video proof; I know Elon is the type of guy to hold that high with all that SpaceX footage.
Good day sir :)
I've driven a Tesla on and off for about 4 years, and I'm thankful to never do so again.
> death trap
https://apnews.com/article/tesla-crash-doors-musk-regulators...
> It's literally been rated the safest car in America since it's inception.
I'm locating our disconnect a bit better: something can be marked "safest car" for road tests but still be rife with issues, like its obnoxious UI choices, etc.
We can arbitrate about what "self-driving success" means until the cows come home, but my point is I've seen a lot of self-driving failures from the Teslas I've witnessed in person.
It's not insane at all for humans to alter their behavior with a tool: you grip a hammer or a gun a certain way because you learned not to hold it backwards. If you observe a child playing with a serious tool, like scissors, as if it were a doll, you'd immediately course correct the child and educate how to re-approach the topic. But that is because an adult with prior knowledge observed the situation prior to an accident, so rules are defined.
This blog's suggested rules are exactly the sort of method to aid in insulation from harm.
> I'm guessing these are the same type of people who sometimes seems to fall in love with LLMs, for better or worse. Really strange to see, and I wonder where people get the idea from that something like that above could really work.
It's a fetishistic cargo-cult rooted in Peter Thiel's 2AM hot tub party. I still believe the LLM approach won't yield true AGI; despite the very real applications, the majority signal is noise.
As for sci-fi: Dune!
I will give the author credit on how they deal with their characters' memories and the re-development of their thoughts, and the usage of time-jumping was reasonable (some books jump around too much, as if these time-skips improve a boring plot). Also the convention for how they solve their dilemma was enjoyable.
Overall, I think the author relies too much on a vocal fandom around the SCP Foundation to glorify the book. I think there is potential for a saga of books but there needs to be more effort in the drafting and editing process to raise the quality of the books to the level the universe deserves.
As a classically trained pianist, I'll say the following commentary regarding this thread: there are many ways to learn to "play the piano" and no one is going to agree on "the right way" because there's no true way.
This project is a good way to practice basic sheet music reading as well as to aurally recognize notes and phrases; teaching would require some cultivation of understanding the context and nomenclature. Thanks for sharing!
> drugs and alcohol, swimming, bad driving, biking, getting hit by a car whilst walking, getting shot, or suicide.
These are false equivalences-- when has a pool try to groom a child over the span of 3 years?
> What if I started Bikesafe an always online dash cam / coach for your rider where AI would identify unsafe behaviour and coach your kid and virtually eliminate bike deaths. Would you feel more safe?
This is wholesale the wrong approach. This is the parent absconding responsibility, which is my driving point of the problem.
Now to the main point:
> You are actually expecting a lot from devices that you never had in the real world in the first place ...
I'm not expecting anything from my devices because machines cannot be held accountable for human choices; a gun cannot be held accountable for being misused. The internet is a powerful tool and users should understand the ramifications of certain actions.
> If your kid lies you might not get a truthful answer to that question in person or online.
That's a parenting moment that one should relish retrospectively. To teach them good morals and values, to remind them that you love them, and that lying about safety processes can be very dangerous.
Software companies will never earnestly attempt to protect children because that action ("acknowledging children are in danger by using our product") acknowledges risk and introduces liability. (VCs hate that shit, especially Silicon Valley VCs.) In the United States, decades ago, laws were introduced to induce accountability of online platforms in regard to IP and child protection laws in the context of user generated content (forums, markets, chatrooms). Basically, these websites/corporations bulked at the weight of accountability ("how are we to monitor every user's action all the time?", "We'll be sued immediately by trolls.", etc.). The parties involved eventually came to a resolution that there's a "notice period" that organizations use to enforce this behavior on its communities.
If I were to write a blog titled "Parent Controls Aren't for Parents", my opening salvo would be "They are minimal-effort guardrails to protect corporations from being sued by negligent parents for post-incident harm."
The story outlined, one of a child prodigy solo-navigating the gritty online world of pre-2000's, is old and tired. An active parent can support a child at all ages safely in these "hacker" moments that are described without giving them un-reined access to tools. A parent should be able to ask "how was your day today?" and get a truthful answer about online activity, just like the same question being asked at the end of the school day. It's out of curiosity and protection, and from a nurtured relationship.
I disagree because children, despite how precocious and "old-soul"ed, are not wise compared to online predators.
I appreciate your POV on allowing children to make their own mistakes; life is the best teacher. Yet, to make an analogy, a gun owner keeps their collection locked up not just for their protection but for their family's protection. Some lessons in life have steep prices and are one-way doors, and we should pass that hard-earned wisdom to the next generation without those costs.