zoogeny

8 ms·

91.
by zoogeny 1y ago
I know this sounds maybe a bit insane, or even self-aggrandizing but I don't comment on public websites for some benefit to myself. I write with the vague hope that some unique expression of myself makes some tiny difference to this un
92.
by zoogeny 1y ago
I haven't read the entire article, but just based on the snippets you posted it doesn't look like they were streaming video using this process. It sounds like they were doing defect detection. I would guess this was part of a pr
93.
by zoogeny 1y ago
People here are giving you mathematical answers which is what you are asking for, but I want to challenge your intuition here. In construction, grading a site for building is a whole process involving surveying. If you dropped a person on a
94.
by zoogeny 1y ago
To be fair to Penrose, he seems to have some humility about it. Although he does also make the claim that math is discovered and not constructed in the same linked video. > it's a conceit to believe that the mechanisms we call math
95.
by zoogeny 1y ago
I was considering your explicit "material -> conscious -> ideas -> material" description. It feels more correct when you say he considers a range of possibilities that connect these, not explicit causality. My take away w
96.
by zoogeny 1y ago
I think your description of Penrose's belief does not match a podcast I recently watched where he discusses these topics with the Christian apologist William Lane Craig [1]. In fact, he explicitly states early on in that video that he
97.
by zoogeny 1y ago
There was a recent Zig podcast where Andrew Kelley explicitly states that manually defining a VTable is their solution to runtime polymorphism [1]. In general this means wrapping your data in a struct, which is reasonable for almost anythin
98.
by zoogeny 1y ago
I would suspect that for the vast majority of use cases, using Redis as a rate-limiter is totally sufficient until major traffic is expected. I mean, Redis isn't a replacement for WAF or some other CDN-like-soak (e.g. Cloudflare). But
99.
by zoogeny 1y ago
I think the article buries a significant drawback: contiguity. It is obviously implied by the design but I think this approach would have hard-to-define characteristics for things like cache prefetching. The next address is a function, not
100.
by zoogeny 1y ago
I read it as claims the article was making. I too was confused, but perhaps that is just the communication? I think the article was going for a comparison between extrinsic motivation (which they seem to claim the original quadratic funding
101.
by zoogeny 1y ago
That is a good point and I should have been more clear. My choice for alpine isn't based on it's reproducibility at all, but its small size. As with anything in engineering there are tradeoffs, there is no singular perfect solutio
102.
by zoogeny 1y ago
Nix falls into the camp in my mind that includes Rust: great idea that I just don't have time for right now. Reproducibility is the holy grail, IMO. It is so valuable that any system that actually achieves it will find some longevity a
103.
by zoogeny 1y ago
I'm not sure what you mean? The literal quote from the Wikipedia article on "worse is better" is: > It refers to the argument that software quality does not necessarily increase with functionality: that there is a point wh
104.
by zoogeny 1y ago
Hacker News is a perfect example of the "Worse is better" mantra applied to social engineering. I mean, slashdot had more features and functionality in the late 1990s. What makes HN work is the tight focus and heavy moderation.
105.
by zoogeny 1y ago
What I miss is E4X [1]. The ability to grab collections of child elements and collections of descendent elements made working with XML bearable. 1. https://en.wikipedia.org/wiki/ECMAScript_for_XML
106.
by zoogeny 1y ago
>what it decides it needs This is what I am suggesting: relying on the model to decide what it needs is maybe not the best use of the available context. It might be better for us to give it the information we are certain it will need.
107.
by zoogeny 1y ago
I don't know how that is solved by MCP? How would the LLM possibly know where to search? Just making an API (or series of APIs) to slack/jira/airtable available doesn't magically surface the context, or the right search
108.
by zoogeny 1y ago
It is interesting that the model I am proposing inverts many peoples expectation of how LLMs will benefit us. In one vision, we give a data-lake of information to LLMs, they tease out the relevant context and then make deductions. In my vie
109.
by zoogeny 1y ago
I've noticed that as well, but I doubt those additional tokens hurt . You can make an unholy mess with namespaces and all of the bells-and-whistles that XML provides, or you can not. But even if you just structure using <tag>&lt
110.
by zoogeny 1y ago
I'd be interested in examples of this. I've worked in offices for all of my adult life and I don't have any examples that come to mind. I think of logic puzzles I used to do as a kid. The whole idea of the puzzle is that all
111.
by zoogeny 1y ago
I'm not sure I understand your objection. You seem to imply that knowing the context is the same as knowing the solution to the problem the context provides? Let me think of an example here. Context needed to determine if there is canc
112.
by zoogeny 1y ago
> Compared to the olden days, when specs were written by pedantic old Unix dudes I think that is one of the reasons (among many others) that the semantic web failed (which doesn't contradict the author, whose point is literally the
113.
by zoogeny 1y ago
I think there is a difference between there being a universal moral Truth (ontological) vs. a particular individuals ability to know that universal (epistemological). The former doesn't necessarily imply the latter. Rejecting the lat
114.
by zoogeny 1y ago
I got about 30% through making something very similar then moved on to another project. This is very nicely done. It is worth noting that I did a lot of research in this space when I was working on my own video editor and there are a lot of
115.
by zoogeny 1y ago
My heart just broke to hear this. Although I honestly don't read the thinking output very often. But I had been cheekily copy-n-pasting the info for my own records.
116.
by zoogeny 1y ago
I agree we are watching the turning point. If raw AI power is the key, Google seems to be in pole position form here on out. They can make their own TPUs, have their own data center. No need to "Stargate" with Oracle and Softbank
117.
by zoogeny 1y ago
How long did it take you to get to 90 people through YouTube videos?
118.
by zoogeny 1y ago
I would politely suggest you work at getting better at this since it would be a pretty important skill in a world where a lot of creative work is done by AI. As some have mentioned, LLMs are treasure troves of information for learning how t
119.
by zoogeny 1y ago
I continue to find Gemini 2.5 Pro to be the most capable model. I leave Cursor on "Auto" model selection but all of my directed interactions are with Gemini. My process right now is to ask Gemini for high-level architecture discus
120.
by zoogeny 1y ago
I've noticed this has gotten a bit better lately, they have obviously been making a lot of UI changes to studio. But yeah, the scroll-jacking as response chunks are streamed in is incredibly frustrating since the model is pretty wordy.

More ›