thornewolf

This was well written. I'm glad that I powered through the initial hesitation I felt when I recognized the narrative style.

This doesn't matter. I don't intend to be rude, of course. I believe this doesn't matter at all.

Same issue

I think they forgot to include the benchmark time for "npm (cached)" inside the Binary Manifest Caching section. We have bun, bun (cached), npm. I think the summary statistics are also incorrect.

The gemini / gemma models all prefer to do json inside markdown blocks IME. I've adjusted my integrations correspondingly.

They realized that CLI is the much better interface for these kinds of tasks.

I'm actually surprised this went negative, I thought HN was gonna love this

everyone takes a side on everything. explicitly or implicitly. intentionally or unintentionally. due to the interconnected nature of the world, it is impossible to construct a true-neutral position. the appearance of neutrality becomes more difficult the more important an issue is to those you interact with.

so, it's not really well-defined to ask whether or not a coffee shop "should" be taking sides, it will take one in all situations. it is a (slightly) more tractable question to ask "how should a coffee shop take a side?" Given how difficult this question is to answer still, I like to answer it with, "doing good is generally aligned with the long-term profit-motive."

this is also why boycotts are actually useful! though collective action, you can align the profit-motive with "good" clearly. this makes it easier for a coffee shop to move in-line.

I'm sorry. Brain fart. Replace Tailwind V4/V5 with DaisyUI V4/V5.

After re-reading double brain fart.

I misread the comment as complaining about DaisyUI. I then responded as if they were complaining about DaisyUI but ALSO accidentally used "Tailwind" in my comment.

Preface:

I accidentally mixed up Tailwind and DaisyUI in my brain. The commenter above me is talking about Tailwind and my "Previous Comment" is responding by talking about DaisyUI but accidentally also using the word Tailwind.

For previous versions of DaisyUI my main complaint was that it looked childish. V5 fixed this. I misread the parent comment as if they were talking about this same issue. My bad.

Previous comment:

I'm not sure when you most recently used Tailwind, but V5 is a big improvement on V4. 4 definitely looked somewhat childish. 5 corrected most/all of this.

I haven't looked at OpenAI's ToS but try and track down a phrase called "indemnity clause". It's in some of Google's GCP ToS. TLDR it means "we (Google) will pay for ur lawsuit if something you do using our APIs get you sued"

Not legal advice, etc.

TFA is another iteration on the wonderful semantics debate about intelligence.

a rock is, a typewriter is, a computer is, a human is. all on some varying levels. TFA takes it as a given that "no thinking" happens. I take it as a given that thinking happens unless we prove it cannot be happening.

https://www.thornewolf.com/its-not-intelligent/

once again, both sides of this argument are making equally valid/invalid baseless claims that are all unfalsifiable. it is completely impossible to determine who is "metaphysically" correct. we can only judge outcomes

I argue that it's not that hilarious because the thinking is very tightly related. The very contemplations that lead to AI doomerism lead to pascals mugging.

One of my main gripes with AI doomerism is that it is downstream of being pascal's mugged into being a doomer

Self-help books are useful, not because they have "mind-shattering ideas", but because they collect a lot of self-consistent and reasonably coherent advice into the same place.

In this same way, the rationalist knowledge seeking strategies are not "mind-shattering" but simply reasonable. It presents a set of rules to follow to be more effective in the world around you.

The parts of rationalism that stretch past the half page of notes mainly concern all the downstream conclusions that pop up from this reasonable set of epistemological rules.

I attended Manifest to meet some rationalists. I consider myself a post-rat. I also heard some comments that Manifest was less Rationalist-y than LessOnline but w/o the additional commentary on "dummies" (not meant to imply you didn't hear it).

I did find some rationalists too far down their "epistemological rabbit hole" to successfully unwind in one or two conversations but nevertheless many clever people. I still need some time to make post-rats out of them, though.

Affirming that it was a positive experience. I'm glad to have attended.

> is qualitatively different from a robot

yes.

> To reduce a system to its inputs and outputs is fine if those are all that matter in a given context

we argue that this indeed is all that matters

> but in doing so you may fail to understand its internal mechanics

the internal mechanics are what we call "conscious" it is the grouping of internal mechanics into one unified concept, but we don't care exactly what they are.

> Those matter if you're trying to really understand the system, no?

since we cannot directly observe consciousness, we are forced to concede that we will never really "understand" it outside of observing its effects.

In the same way that a mechanical turk human and a robot can "play chess", a human and an LLM are "conscious". That is, consciousness is the ability to play chess, by some mechanism. The exact mechanism is irrelevant for the purposes of yes/no conscious.

We now enter a discussion on how much these two consciousnesses differ.

looks very fun excited to try it out

nooooo the side project ive put off for 3 years

This is a fair critique. While I am merely a "LLM wrapper", I should put the product's best foot forward and pay more attention to my showcase examples.

Thank you for the kind words! I am looking forward to creating a Show HN next week alongside a Product Hunt announcement. I appreciate any and all feedback. You can provide it through the website directly or through the email I have attached in my bio.

+1 to this flow. I use the exact same phrase "in extreme detail" as well haha. Additionally, I ask the model to describe what prompt it might write to produce some edit itself.

It is very sensitive to your input prompts. Minor differences will result in drastic quality differences.

The co-drawing is definitely not a fully fleshed-out product or anything but I think it is a great tech demo. What don't you like about it?

Model outputs look good-ish. I think they are neat. I updated my recent hack project https://lifestyle.photo to the new model. It's middling-to-good.

There are a lot of failure modes still but what I want is a very large cookbook showing what known-good workflows are. Since this is just so directly downstream of (limited) training data it might be that I am just prompting in a ever so slightly bad way.

To push back on this, I consider pandas/numpy so crucial to Python as a whole they are effectively stdlib to me. I wouldn't blink at this because it would happen sooner or later.

Unless is was absolutely critical the server have as small as a footprint as humanly possible and it was absolutely guaranteed there would never need to be included in the future of course. However, that first constraint is the main one.

GP comment is what has already happened "every 6 months" multiple times

He is less concerned that people can create an evil AI if they want to and more concerned that no person can keep an AI from being evil even if we tried.

We have no reason to believe that it is not reasoning. Since it looks like reasoning, the default position to be disproved is this is reasoning.

I am willing to accept arguments that are not appeals to nature / human exceptionalism.

I am even willing to accept a complete uncertainty over the whole situation since it is difficult to analyze. The silliest position, though, is a gnostic "no reasoning here" position.

Ignore the noncompetes. Never get sued. If you do, everyone else is on your side.

I have a similar experience. "A conviction to remember". I was a bit older though and playing with my toys.