@self_made_human's banner p

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users  
joined 2022 September 05 05:31:00 UTC

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


				

User ID: 454

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users   joined 2022 September 05 05:31:00 UTC

					

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


					

User ID: 454

I don't care about the moral worth of non-human animals, if they didn't want me to eat them, they should have been less tasty.

On a more serious note, I have no innate preference for cruelty, I simply do not care. If lab-grown meat (or even meat substitutes) tasted just like meat, and were cheaper, I'd eat them with equanimity.

I have a cousin in the UK who is a vegan, initially to get laid (his ex was vegan), but apparently the moral draw remained. He's stuck fast to it, even if his current soon to be fiancé is merely vegetarian. He's not preachy, when we meet, he makes sure to look for mutually acceptable options, and I have no issue with his lifestyle. I can see it makes his life significantly harder, but that's his choice. I introduced him to an Indian friend of mine in Edinburgh, who started lecturing him on nutritional deficits. I pointed out that he looked perfectly healthy to me, and if there were the kinds of serious issues he was positing, the man would have been dead by now. Each to their own, and me to a plate of bacon rashers please.

On an intellectual level? I'm well aware that people are using LLMs for smut. I just never looked into the... specifics of it. I regret to say that I think the I read all of the Evangelion stuff, I got off the train when the beastiality station was the next stop. Or is it even beastiality if it's a dragon fucking a pony? Only Satan knows.

If I want titillation, I'm opening PornHub. Text isn't my jam, but if it's yours then I'm not one to actually judge. From a purely technical standpoint, it all came across as competent, probably way better than the average literotica writer.

Are you accusing me of being a worse version of Scott?

If you are, I can't fault you haha. I live in the shadow of giants, and try to grow in the space between their toes.

Very high salary, ironclad employment security, lifelong employment, clear delineation between work and rest, etc

God, I wish that were me. Or true of most British doctors. At least the WLB is better in psych.

That's a pretty thoughtful comment, and I think it makes eminent sense.

Hmm.. How would that work? If they're self-hosting, then are they executing all the server-side code themselves and also mirroring the content?

Thank you. That seems quite sensible

You can be a software dev in a small 100% male company in Eastern Europe that has a chat channel for sharing porn.

I suppose they need additional perks to make up for the low salary haha.

Not that I'm aware of. In the specific instances of Gwern, X, HN and a few others, I was able to track down the original link manually, or using Google search parameters.

Now that makes me wonder why more doctors aren't fat and/or drunk. Everything you've said about attorneys fits our bill. Maybe we're more health conscious (and I hope we are), maybe we run around more, or maybe we just sleep better at night from a clean conscience.

I'm grateful for Substack's ability to show me the sources of traffic. As I've previously mentioned, it keyed me in to the fact that Gwern himself had archived me. I was quite pleased to see a few people independently share my posts on HN too.

But, sometimes it raises more questions than it answers. Why are my posts being shared via the Steam forums? Why Slack? Who is using their company comms or a gaming website to talk about anything I've written?? I wish I knew, but it's a mystery that I'm unlikely to solve. I even saw Localhost:1881 in there, which I strongly suspect shouldn't be sending me any traffic.

After writing the above, I went back to my most popular article, the one about the effects of psilocybin on depression. Substack has 800 views from LinkedIn, 15 from Microsoft Office, 2 from the Brave browser (?).

I hope someone can tell me what the fuck is going on here.

a story where Shinji agrees to undergo conversion therapy in order to cure his homosexuality

Good news: I am very interested in your comment. I'm taking notes.

Bad news: A psychiatrist is very interested in your comment. And the notes are on NHS mental health branded stationery.

Truly, I have been ignorant of the depths of depravity, I am happy to concede that I'm using these models the wrong way now.

I genuinely can't see myself using any LLM in this manner, but I consider it an entirely legitimate and harmless use case. I'll go hang out with Voltaire when he says that he doesn't understand what would compel you to do this, but we'll defend your right to try. You have done a great service. We are all more informed, and perhaps slightly more damned, for your efforts.

Attacking a metaphor is not an argument.

I've explained why I think the parrot is a terrible metaphor above. And no, metaphors can vary greatly in how useful or pedagogical they are. Analyzing the fitness of a metaphor is a perfectly valid, and in this case essential, form of argument. Metaphors are not neutral decorations; they are cognitive tools that structure understanding and guide action.

A computer virus shares many properties with its biological counterpart, such as self-replication, transmission, damage to systems, the need for an "anti-virus". It is a good name, and nobody with a functional frontal lobe comes away thinking they need an N95 mask while browsing a porn site.

The idea of the Cloud at least conveys the message that the user doesn't have to worry about the geographical location of their data. Even so, the Cloud is just someone else's computer, and even AWS goes down on rare occasions. It is an okay metaphor.

The Parrot is awful. It offers no such explanatory power for the observed, spiky capability profile of LLMs. It does not explain why the model can write functional Python code (a task requiring logic and structure) but often produces insipid poetry (a task one might think is closer to mimicry). It does not explain why an LLM can synthesize a novel argument from disparate sources but fail to count the letters in a word. A user equipped only with the parrot model is left baffled by these outcomes. They have traded the mystery of a "fallible human" for the mystery of a "magical parrot".

I contend that as leaky generalizations go, the former is way better than the latter. An LLM has a cognitive or at least behavioral profile far closer to a human than it does to a parrot.

You brought up the analogy of "parroting" information, which I would assume involves simply reciting things back without understanding what they mean. That is not a good description of how the user can expect an LLM to behave.

On an object level, I strong disagree with your claims that LLMs don't "think" or don't have "minds". They clearly have a very non human form of cognition, but so does an octopus.

Laying that aside, from the perspective of an end-user, LLMs are better modeled as thinking minds.

The "fallible but knowledgeable intern" or "simulation engine" metaphor is superior not because it is more technically precise (though it is), but because it is more instrumentally useful. It correctly implies the user's optimal strategy: that performance is contingent on the quality of the instructions (prompting), the provided background materials (context), and a final review of the output (verification). This model correctly guides the user to iterate on their prompts, to provide examples, and to treat the output as a draft. The parrot model, in contrast, suggests the underlying process is fundamentally random mimicry, which offers no clear path to improvement besides "pull the lever again". It encourages users to conceptualize the LLM as a tool incapable of generalization, which is to ignore its single most important property. Replacing a user's anthropomorphism with a model that is descriptively false and predictively useless is not a pedagogical victory. It is swapping one error for another, and not even for a less severe one to boot.

I still don't see why that applies, and I'm being earnest here. What about the "stochastic parrot" framing keys the average person into the fact that they're good at code and bad at poetry? That is more to do with mode collapse and the downsides of RLHF than it is to do with lacking "consciousness". Like, even on this forum, we have no shortage of users who are great at coding but can't write a poem to save their lives, what does that say about their consciousness? Are parrots known to be good at Ruby-on-rails but to fail at poetry?

My explanation of temperature is, at the very least, meant as a high level explainer. It doesn't come up in normal conversation or when I'm introducing someone to LLMs. Context windows? They're so large now that it's not something that is worth mentioning except in passing.

My point is that the parrot metaphor adds nothing. It is, at best, irrelevant, when it comes to all the additional explainers you need to give to normies.

I would, if I felt like I gained anything out of it. As it is, the previous thread has only contributed to early hair loss. You still haven't noted any of the clear and correct objections I've raised, and I'm tired of asking.

Thanks for verifying the answer! If there's a takeaway here, I'm not sure why you're ?paying for Grok 4. Grok 3 was genuinely impressive, and somewhat noticeably better than the competition at launch. Not the case here I'm afraid.

With that aside, I'm not sure how other people see LLMs tackling problems of this complexity and then claim they're not reasoning. It bemuses me.

I tried using o3, but it correctly noted that the file you mentioned isn't available, and its web browsing tool failed when trying to use the website.

I can't do anything about the missing document, but I did manually copy and paste most of the website. This is its answer:

https://chatgpt.com/s/t_6892b68c0c3081919777d514df3ba8c2

Have you tried simply asking it not to search the web? The models usually comply when asked. If they don't, it should be evident from the UI.

I would consider myself an LLM evangelist, and have introduced quite a few not-particularly tech savvy people to them, with good results.

I've never been tempted to call them stochastic parrots. The term harms more than it helps. My usual shortcut is to tell people to act as if they're talking to a human, a knowledgeable but fallible one, and they should double check anything of real consequence. This is a far more relevant description of the kind of capabilities they possess than any mention of a "parrot".

Why am I surprised? People make Redstone computers for fun. I guess this all just takes a very different mindset haha.

Here I was hoping you'd cheer up after 5 months away. C'mon, all the nurses in your area can't be hambeasts.

What on earth is going on in that screenshot? I know Minecraft mod packs can get wild, but that's new.

Will he change his mind if someone presents a chess-focused LLM with a high ELO score?

We just had a certain somebody on this forum bring up current LLMs being bad at chess as an example. He even got an AAQC for it (one of the most bemusing ones every awarded).

I do not recall him acknowledging my point where I pointed out that I GPT 3.5 Turbo played at ~1800 ELO, and that the decline was likely because AI engineers made the eminently sensible realization that just about nobody is stupid enough to use LLMs to play chess. If they do, then the LLMs know how to use Stockfish, in the same way they can operate a calculator.

This isn't a great comment, from a moderation perspective, but I'm now going to talk as a general user.

Freddie is seriously mentally ill. He's disclosed his struggle with bipolar disorder (and I seem to recall he had psychotic breaks). However, the views he espoused aren't exclusive to the mentally ill. God knows I would like to pin a diagnosis on someone like Gary Marcus, but being pugnacious and immune to evidence isn't actually in the DSM.

As far as I can tell, he does have 'aesthetic' reasons to object to transhumanist/rationalist viewpoints. He is woefully under-informed about modern AI, but so are most people. It is easy to demolish his arguments without needing to resort to ad hominem attacks. If he starts accusing OAI of wire-tapping his house (like Altman's sister), then this facet of his personality would be worth raising.

Thank you, I doubt I'll have time to venture that way again since I'm leaving soon, but I'll bookmark those for another trip!

Norf vs Souf is a fractal, recursive pattern, it seems, that applies even to Indian immigrants in Britain.

His behavior makes no damn sense to me, regardless of where we're at. We lived in the same city in India, even if he didn't grow up there. If this is a behavioral difference that generalizes to Scotland vs England, that's news to me.

Was it one of these trains that split in two? I travelled to Canterbury by one once, and even the natives were mildly confused.

I don't think so. Our original train was supposed to get us to Dover directly. This one took a parallel route, stopping at Ramsgate. Wasn't too hard to go to Dover from there, thankfully.

The next time I heard such curses from my wife was a week ago, when we were hiking in 35-degree heat.

Jesus. I would have turned back. Anything north of 27° rules out a hike for me. Even though the sun made a surprise appearance on this trip, it actually made the whole experience way more pleasant.

You can actually get a good explanation by asking an LLM (preferably a good one) Something like Deep Research will at least give you the hooks to get into. I'm not saying this just because I'm lazy, but because I expect you to get a good answer out of it.

rather than affirming whatever feelings you might have without challenge

I've heard plenty of horror stories about therapists who do, in fact, behave in this way. I have no strong opinions on whether they're representative, but they're certainly not rare.