ceba

Karma: 10

ceba May 15, 2025, 12:41 AM
1 point
0
in reply to: rogersbacon’s comment on: Behold the Pale Child (escaping Moloch’s Mad Maze)
this is an introduction to a much larger body of thought.
Oh! In light of that, my criticism is diminished. I’ll edit my comment to reflect this.
I really want to believe that the hopes you express are well founded. But, I also want to believe what’s true. LW taught me to be (cutious/skeptical) when this happens, that’s why I was critical. I hope that my comment was useful as feedback, if not as critique.

ceba May 14, 2025, 6:32 AM
3 points
0
on: Behold the Pale Child (escaping Moloch’s Mad Maze)
Edit: the author has clarified that their post is meant as an introduction to a larger work, so it would make sense for them to introduce their view alongside others’ without justifying why just yet.
This was interesting. I enjoyed reading it. However,
Let us accept some version of Alexander’s eschaton: the final level of the Reality Game is beaten by compelling or convincing a “god” (either a machine god or an actual divine being) to transform the cosmos into an everlasting paradisal garden. How will we win—through work (coercion or argument) or through līlā, through rational intelligence or creative mêtis?
This paragraph confuses me. First you’re saying:
“ok, let’s think about Alexander’s ‘gardener’ solution to the problem”,
then you ask the question:
“how do we solve the problem on our own?” (meaning without the ‘gardener’)
These belong to different branches of discussion (from one view.../on the other hand...), but the second part is presented as though it’s salient to the first. It scans as if the author forgot what he was thinking partway through, and picked up on a different thread.
Am I being excessively pedantic? well,
this stuck out to me, because usually when different views are presented in a text (from one view.../on the other hand...), the author must decide in favour of one (or none) of the those presented, and explain why. This, the author has not done. Which is strange, given the apparent effort (and I believe, evident skill) this essay represents.
The above paragraph marks the part where the author is finally offering their own opinion, having explored those of others. Except it begins in just the place where we expected them to explain why they disagree with Alexander. Which should be the most interesting part, because this is Lesswrong, the place with
1. unusually strong norms around how people should form their beliefs, and
2. unusually high frequency of people who disagree with the Author’s conclusion vehemently.
Why? why do you think this, and not that?

ceba May 12, 2025, 8:11 AM
2 points
1
in reply to: Yates’s comment on: A Proposal for Recognizing Nonstandard Intelligences
A certain type of person tends to respond with antipathy to wide appeals to empathy. I imagine they dislike the subtext that it’s shameful that they don’t feel a certain way. Something Yaroslav Granowski said fit this pattern:
To me it feels like you appeal to the pity and guilt of the priveliged
To me, it didn’t seem like the post bade for my guilt at all, but he seemed to perceive that quite strongly, to the point of engaging it’s author.
I don’t think there’s anything wrong with people like Yaroslav feeling the way they do. Yes, it feels provocative to engage with. But you don’t gain anything for raging at them.
There are many, many people who share the values you promote here on Lesswrong (me included). There are also those who don’t. Posting will expose us to both. This is one place where we (try to) accept this, despite the discomfort it brings.
I do believe there is an audience receptive to your sentiment on Lesswrong. It just might take a few tries. I wish you the best of luck.

ceba May 11, 2025, 7:13 AM
1 point
0
in reply to: Yaroslav Granowski’s comment on: A Proposal for Recognizing Nonstandard Intelligences
I didn’t intend to badger. Just curious is all.

ceba May 11, 2025, 6:53 AM
4 points
3
in reply to: Yaroslav Granowski’s comment on: A Proposal for Recognizing Nonstandard Intelligences
Why shouldn’t we be curious about others, when they’re talking across a cognitive gap? Isn’t curiosity something people here value?

ceba May 5, 2025, 9:25 AM
1 point
0
in reply to: leogao’s comment on: leogao’s Shortform
✨Portability!!✨
Hygeine? keeps flies off, less surface area/volume to get your grubby hands on.
Storage: Insulation to keep meat cool for longer. Also keeping germs bugs and dirt out would help this.
Convenient way to save smaller scraps of meat when you don’t have utensils for a stew.
Easier food makes better nourished people: so much easier to eat sandwhich, gyoza, sausage roll than almost any other food. Obviously I give very refined examples, but possibly true for more rough forms too.

ceba Apr 25, 2025, 6:33 PM
1 point
0
in reply to: Peter Salien’s comment on: Open Thread Spring 2025
I’d like to see your work, when it’s ready to be shared.

ceba Apr 25, 2025, 6:19 PM
1 point
0
in reply to: Davey Morse’s comment on: Open Thread Spring 2025
What environmental selection pressures are there on AGI? That’s too vague, isn’t it? (What’s the environment?) How do you narrow this down to where the questions you’re asking are interesting/reaearcheable?

ceba Apr 25, 2025, 5:57 PM
1 point
0
on: Open Thread Spring 2025
If your product couldn’t exist without our labour, then our labour is worth some fraction of that product’s value to you. If it’s worth something to you, you can then afford to pay some fraction of that fraction for that labour. If in fact you can’t afford to pay any fair price for that labour, then this means the value generated by that product is far less than the collective value of the labour required to create the product.
This is a fundamentally uncapitalist venture. If you see a nonsensical buisness model, you generally assume it’s a tax thing, or a scam. The efficient market doesn’t allow such products/buisness models to exist. So if it yet exists, coordinated effort is being made to keep it alive.
This product is intended to automate all human labour.
We are in effect subsidising the development of a product intended to end the livelihoods of all people who work for a living.
Discuss? Agree/disagree? Incorrect? Commie spam?

ceba Apr 14, 2025, 7:02 AM
2 points
0
on: The Era of the Dividual—are we falling apart?
If I compared ‘the dividual’ to ‘memes’ would that be a conceptual rounding error?

ceba Apr 12, 2025, 12:04 AM
1 point
0
on: Debunk the myth -Testing the generalized reasoning ability of LLM
Hello!
Required to walk at the speed of a park walk on flat ground, walk 10 meters continuously without falling VS Required to participate in the campus 100-meter race and win a medal
If this post, including the examples you’ve shown, is representative of the general writing style used, I suggest that such a style may be a confounding factor.
Even if it isn’t, or you don’t feel it is important, writing style is a signal, and the people reading your work will be very sensitive to that kind of signal. Even those who are genuinely curious might not spend as much time or effort reading your work than they otherwise might.
Given that your findings as presented here seem to contradict a popular narrative (it appears to be “debunking the myth”), less curious people will be on especially vigilant lookout for flaws they can use to dismiss your work entirely.

ceba Mar 20, 2025, 7:52 AM
2 points
0
in reply to: Knight Lee’s comment on: I changed my mind about orca intelligence
Proving their intelligence to who? Who would even care? How likely is it that any coordinated action would be taken to save starving orcas? If its highly likely, is this fact legible to them?
It might seem to them that however many intelligent animals we’ve aided, we might just have harmed as many (intentionally or otherwise), or more.
Have humans ever taken large scale coordinated action to help an animal population that wasn’t redressing some harm done to them by humans?

ceba Mar 13, 2025, 12:46 AM
1 point
0
in reply to: duryt’s comment on: “Dangers of AI and the End of Human Civilization” Yudkowsky on Lex Fridman
if I want to find the maxima of a function, it doesn’t matter if I use conjugate gradient descent or Newton’s method or interpolation methods or whatever, they will tend to find the same maxima assuming they are looking at the same function.
In general, those methods find local extrema. They don’t tell you how many there are, or where the next closest point is once you’ve found one of them. A loss landscape might have several local minima. Which one you find depends on where you start.
Why shouldn’t there be different minds that are at comparable minimum values, but not very close on the loss landscape?

ceba Mar 6, 2025, 12:11 PM
1 point
0
on: How to Make Superbabies
I like this. But I have questions. I hope for discussion, not answers.
Which people are most likely to get access to this tech first? is it likely that access will ever spread significantly beyond this group?
How does one raise a super child?
If they require different support from other children, whose resources are available to provide it?
Would there be some institution connected to the biotech company that would maintain a relationship with families or are they just selling a limited service?
What roles/responsibilities are the super adults expected to take on? In what sense are they compelled to take on any responsibility?
What broad-reaching effects might the existence of superbaby capabilities have?
do I have license to speculate?

ceba Nov 12, 2024, 6:00 AM
3 points
0
in reply to: Alexander Howell’s comment on: BIG-Bench Canary Contamination in GPT-4
Claude’s initial justification for not completing the string doesn’t make sense to me, but you didn’t acknowledge this. Is this because I’ve misunderstood Claude?
common knowledge: Claude completing the string would be evidence that text containing the canary was included in its training data. This is the Canary’s purpose.
Claude’s argument: I should be cautious with string’s content, if reproducing it makes it less effective in the future.
At some point, we have to take the (small!) risk and check. It can’t get less effective than this.