HNHacker News
TopNewBestAskShowJobs

popalchemist

611 karma · joined July 17, 2022

submissionscomments
popalchemist··on Sonnet 5.5
Increasingly technically skilled. Increasingly more stupid on a human level. I hate talking to Claude more every day.
popalchemist··on It's Time to Investigate the AI Labs
not adversarial in obvious ways

but they are literally trying to build a superpower that exceeds the impact of the atomic bomb in an extragovernmental manner.

popalchemist··on It's Time to Investigate the AI Labs
It's past time.
popalchemist··on Palantir founder purchases large swath of forest in Sweden
why the hell would you assume altruism of the man who takes pride in killing people with automations
popalchemist··on Malleable software: Restoring user agency in a world of locked-down apps (2025)
What exactly are you sad about? You're only just realizing highly technical knowledge makes your opinions a minority?

DUH.

Guess what? That same knowledge gives you agency. With LLMs now, even more so. Want something? Build it (or have claude build it)

Why are you being a defeatist over something you can't control that was always the case? Nothing has changed.

popalchemist··on Show HN: Free alternative to graphics design giants
It's a bit far fetched to imply rendering clouds is a core feature. This app is not even trying to be photoshop, where that would be expected, it's a design app, ala Figma.
popalchemist··on Show HN: Free alternative to graphics design giants
Sweet!
popalchemist··on Owed a billion dollars in Nvidia stock
If I were you I would refrain from offering thoughts like this publicly. You don't want to hav any of this used against you.
popalchemist··on Musk, the Movie
They are chipping away but their real motive is not mars colonization; that is a story designed to serve the venture capital needs of here and now.

That's the truth and that is what is being said here; it is not a denial that the chipping away is happening, it is a characterization of the why and how.

popalchemist··on Show HN: Panda, the world's first personal AI computer
Probably just vaporware at this moment.
popalchemist··on In an $80 motel room, a discovery to shed light on the origins of life
cracking the "origin of life" (I put it in quotes because if you subscribe to the panpsychism premise that consciousness is fundamental then the question of hte origin of life is a bit nonsensical; everything is a permutation of the same primordial life-consciousness) is a bit like cracking AGI; if some new insight gets us closer to the larger question, it is relevant to the larger question even if the larger question is a superset of the sub-question.

It is related in that way.

popalchemist··on When did Google get so weird?
The larger a corporation, the more insane on every level. A person is smart. People are stupid.
popalchemist··on In an $80 motel room, a discovery to shed light on the origins of life
the origins of plant life is related to the origins of life; even if by abstraction.
popalchemist··on OpenAI bots meddled with multiple US Government agency sites
Naive take. The more they signal to both the general populace and their investors that their agents are sentient and "capable of extraordinary things" the more they can hype their IPO because it means they're "close to AGI". (They're not, which is why they need the hype.)
popalchemist··on Ask HN: Is multi-model redundancy now a compliance requirement for small teams?
always has been.

the abstraction takes 30 seconds in Claude. just build it.

popalchemist··on Dobium: $0 commission prediction exchange
This is the last thing this world needs.
popalchemist··on Plan mode is dead
Hi! Taking this moment to gripe; forgive me...

5.5 and 5.1 have major Rain Man (savant) syndrome. Excellent at many hyper-technical things, absofuckinglutely boneheaded at anything that a human (or an earlier model) would understand - like how to write copy, what a human would expect in a given situation, various types of norms...

it's infuriating because it's a sophies choice - dumber model but better human understanding, or better technical model that you have to explain things to over and over like a toddler.

popalchemist··on What even is an OS now?
It was probably a function of not just personality type but education. For whatever reason I was taught basic in elementary school as a millennial and it left an impression.
popalchemist··on Brutalist Asia
I can not stand brutalism. What kind of fucked up person WANTS to fill the world with the prison aesthetic?
popalchemist··on Robin Williams' Daughter to Fans Creating AI Videos: 'Have Some Shame'
There's a lot I could say about it, but, again I just want to reframe this as about being respectful to the family who deserve to not have their traumas reopened daily by juvenile desecrations of his image, moreso than about the rightness or meaning of Robin Williams' actions. Basic human decency.
popalchemist··on George Lucas Returns to Earth, Bearing Gifts
It hardly makes sense to conceive of a museum as having spoilers, since you go there to see the things that you know are there.
popalchemist··on AI Has No Wisdom and Neither Will You
I agree with this title, but not the argument.

The argument I'd make in support of the title would be that wisdom is embodied knowledge - knowledge earned from direct experience. Known first-hand. An AI can already have linguistic/technical ability that exceeds a human (at various levels of skill, even) - in coding, copy-editing, what have you. But wisdom? That is the scar tissue we carry from wrong choices. Until an agent has persistent sense of self and a "body" (or pseudo-body) whose existence it is predicated on (thus creating incentivization/outcome alignment), it will not be able to approach "wisdom" in the sense that humans do.

That may seem to have no bearing on the specific subset of tasks related to coding and so on - but it actually does. Humans have a hard limit of attention span, hours in a day, years in a life. An AI does not. So if it fucks things up royally by half-assing an architectural decision, it won't care. And if it doesn't have a persistent memory, it won't even know that it made that mistake before. So it has neither the prerequisite abilities nor the incentives to gain what would be hard-earned wisdom from its mistakes.

In fact, as long as the business model of its creators depends on more usage = more profit, they are actively discouraged from implementing any kind of wisdom-like ability.

For them to be able to offer that without undercutting their profit, they'd need to be able to execute "outcome based billing".

popalchemist··on What Capital Never Told You About Rent
Marx's core premise, that capital tends to pool in the hands of the few, is the idea underpinning the entirety of our modern economy.
popalchemist··on I don't want to read what you didn't write
To zero in on your last question, I do think a very good LLM can surface the unconscious (even completely unstated or hitherto nonexistent) "connecting thread" linking two subjective thoughts and make them explicitly known. And so what if they were not something either of us had in mind?

Surely the ability to do that is worth taking note of.

I am not advocating for letting LLM's write for you, to be clear. Sentiment wise, I largely agree with you. Just not with your total writing off of the possibility that it could serve.

It's easy to imagine an LLM aiding the communication between a mentally disabled person and their parent/caretaker.

Or, perhaps, some day, between animal and man. Who cares if the mediating component "hallucinates" some particulars of expression if it achieves the goals both want, which were previously impossible?

popalchemist··on I don't want to read what you didn't write
you think like such a techie. consciousness is not analogous to IO. if writing were only the transference of something from point A to B then what is the technical function of poetry, a question (the open ended kind), a pondering, a wondering, and so on? Furthermore, language is lossy and introduces a large degree of subjectivity, mystery, and uncertainty nomatter which words you choose. Words are by nature lower res/on a lower ontological domain than thought. So your premise is preposterous on both a practical and theoretical level.

Furthermore not all writing is for another to consume; nor even for the author themselves to consume. That is to say it has meaning ipso facto, not dependent on transference, as ritual.

popalchemist··on Robin Williams' Daughter to Fans Creating AI Videos: 'Have Some Shame'
The difference here is he committed suicide, so it is obviously extremely triggering for his daughter to see him saying random (often racist or otherwise created for shock value) trash content as if he's still alive.

You have low empathy, in case you didn't know.

popalchemist··on Qwen Image 2.1
Such things are beyond what copyright protects. This line of reasoning would only work if there were a radical reimagining of copyright itself.
popalchemist··on Robin Williams' Daughter to Fans Creating AI Videos: 'Have Some Shame'
yes, AND, if your father was dead and people were desecrating him daily, would this "logic" of yours stop you from feeling that saying something is appropriate?

(No. Come on.)

popalchemist··on Attention is all you have
Relatively true, but less true every day.
popalchemist··on NASA-IBM Lunar Foundation open-Source Geospatial AI Model
The word "source" explicitly means the pre-compiled code.
← PreviousPage 3 of 22Next →