Wilduck

When did spacetimedb pivot from being a gamedev backend to being something for AI agents to use? Are they also trying to pivot to a IaaS model?

Glad the feedback was helpful! Feel free to email me at either email in my profile. I think your next steps laid out here are good ideas. In particular, fully documenting (with video or text) a handful of actual successful uses of your site / the programs seems like a great next step.

Your reply to this comment got flagged dead for some reason (not sure why, it seemed like a good faith comment) so I'm responding with a couple thoughts here.

Looking at your set of steps, I'm definitely getting lost around step 3 "The program then guides you one step at a time. Each step is a small module: a short assessment (a few structured questions), a task, a referral, a verification, and so on. It adapts to your answers rather than dumping a wall of text." Am I wrong that this flow is not built yet? Or is completely inaccessible to me?

Like, looking at the program page https://dohmh.codify.nyc/program/71776 I don't see how I can follow any of the steps. You say this is drawn from DOHMH pages, so I found the corresponding one for this program: https://www.nyc.gov/site/doh/health/health-topics/stay-healt.... I think this page is miles better than what you've built for a few reasons:

1. It comes with context in the form of paragraphs of explanation 2. It doesn't use undefined words like "program", "agent", "target". No where on that page is a string like "PersonalPrepTaskAgent". 3. It provides links to more resources, also written in nice human readable language

I don't want to be too hard on your project though. I've had to interact with the Pennsylvania DMV website a number of times, and trying to understand how to meet the requirements for a residential parking permit when I moved into my house was crazy because I needed some non-standard documentation due to having out-of-state plates. I would have loved a someone to break down the steps to me in very clear, understandable language. But I'm not sure your site is doing that right now.

As a last note, I wouldn't say you're "spamming" the comments, but I might recommend reaching out to potential users directly and asking them for harsh feedback. Posting into the void here clearly isn't getting you enough actionable feedback.

My man, you're posting this project all over the HN comments, but I think you need to step back and really think about what you're building and what the user experience is.

Your link to an example protocol: https://dohmh.codify.nyc/protocol/71776 makes basically zero sense to me. Similarly the program page. What do you expect someone clicking on these links to take away from them?

It's because the author is comparing that incident to how they were prepared with the bulk of their analysis in advance. Ready to publish as soon as it happened.

The downvotes are because the discussion isn't really about "which option do users prefer". It's actually about returning user choice. As the original comment said: "If I want a separate tab, I can Cmd+click and Browsers don't have the reverse option for opening in the same window."

So when you say "Nope!", you're being downvoted because you're implicitly saying "actually users don't deserve choice".

You don't, you just cheese the level by slamming into the wall just to the left at the start and skip most of the course.

This is one of a bunch of issues that have been popping up since a vibe coder took over the bulk of development on this project. There's a (probably also AI generated) list of a big portion of the issues here: https://github.com/benbjohnson/litestream/issues/1221. That proposal has been open for a few months, and it seems (from my POV) unlikely to be resolved any time soon.

I've been following litestream for a while, and it seems like the project has been hijacked by a vibe coder. I wouldn't trust it for critical tasks anymore.

Thanks for the kind words! I hadn't heard of tatoeba before, so thanks for sharing that as well.

Is that a "hard problem" though? Really?

Personally, as a DM of casual games with friends, 90% of the fun for me is the act of communal storytelling. That fun is that both me and my players come to the table with their own ideas for their character and the world, and we all flesh out the story at the table.

If I found out a player had come to the table with an LLM generated character, I would feel a pretty big betrayal of trust. It doesn't matter to me how "good" or "polished" their ideas are, what matters is that they are their own.

Similarly, I would be betraying my players by using an LLM to generate content for our shared game. I'm not just an officiant of rules, I'm participating in shared storytelling.

I'm sure there are people who play DnD for reasons other than storytelling, and I'm totally fine with that. But for storytelling in particular, I think LLM content is a terrible idea.

Really cool project. A quick note, I had to dig in your FAQ to find your definition for LOQ:

> "What does 'LOQ' mean in your results? > > Limit of Quantification (LOQ) is the lowest concentration we can reliably measure. Results below LOQ are marked "<LOQ" - this doesn't mean zero, just below our measurement threshold."

IMO this definition should be on every results page, since most of the pages have more LOQs than anything else.

Viewport size and deciding where to paginate makes a naive approach to this surprisingly difficult. That being said, if you can control the css / html, you can often solve these problems with a short media query and some hints at where to break pages (e.g. https://developer.mozilla.org/en-US/docs/Web/CSS/break-after).

It seems like before you think about monetization, you should think about how you can generate the amount of traffic you had in February repeatably. February had 6k visitors, but March had 350, and August had only 240. It's going to be pretty hard to generate any revenue from a website that links to free books with only 8 visitors per day.

I don't want to be mean, but this post has exactly the problem the person you're replying to was complaining about. The person you're replying to, I think, would like an explanation that reads more like "It's like Twitter, but not tied to a mega-corp, just for you and your pals". I don't know if that description actually fits Nostr though because, like the person you're replying to, I have a pretty hard time understanding what Nostr actually _is_.

A few ways:

1. An American company benefited from their labor

2. American consumers benefited from the goods / services they contributed to providing

3. American citizens benefited from the services provided by the taxes they pay

4. Other American businesses benefited from their patronage

Please don't just copy paste LLM output into comments.

> Another way to look at this is you’re outsourcing your understanding to something that ultimately doesn’t think.

You read this quote wrong. Senior devs outsource _work_ to junior engineers, not _understanding_. The way they became senior in the first place is by not outsourcing work so they could develop their understanding.

The irony is that a valid reason for in-house developers to not want to use an external product is concern about the long term support availabilty for that external project. You could make a case that this product shutting down is proof the in-house developers were right not to trust it.

I don't think that's totally fair in this case, since it seems they open sourced their software. But also, in general, I think NIH syndrom gets a bad rap. Sometimes a "worse" solution you control really is more reasonable compared to a technically superior solution made by an external company.

Agree! I bought my kindle in 2016 and have read hundreds of books on it. The cost per book at this point is in the pennies. It's not like you have to buy a new device every book, or even every few years.

Also I was very confused by the assertion that ebooks are a niche market. By the authors math more than 20% of people in the UK use e-readers to read on a regular basis. That seems like a very healthy market to me!

I know you're joking a little, but I personally would love to see them! I'm very interested in how people manage simple deploys.

> type `print(function_name.__doc__)`

Or even easier: `help(function_name)`!

The problem is that I'm not requesting Mozilla do anything. Firefox isn't a "service" it's a web browser. When I input a seach query, _I_ am acting on my behalf, not Mozilla.

I don't want any language where they get to insert themselves into that chain of behavior. Curl doesn't need a TOS, why does Firefox?

I think Goodhart's law only applies if you have fixed, published criteria for funding. That's why I mentioned transparency explicitly. I wonder if you could avoid some of the worst of Goodhart's law by saying something like "the formula changes every year, and we will publish it only after 5 years, but the goal is to reward value provided, and de-risk the ecosystem". The idea being you're explicitly trying to incentivize broadly valuable work rather than specific metrics.

It's a bit like the SEO dance. Publishing the exact formula makes it much easier to game SEO, so instead search engines say stuff like "we're trying to gauge the overall quality of the site, we use metrics like [...] but not exclusively, focus on making good, responsive, accessible content above all else". Obviously it doesn't work perfectly and the more money there is, the more incentive to game the system, but it seems better than the alternative of publishing the exact ranking algorithm.

I think this is a really interesting model for providing funding to open source software. There's something about the "Index Fund" approach that is really appealing. I also think it's interesting that the author was both balancing "value" and "risk". I do wonder, if this became a more dominant strategy for providing funding for open source how you would deal with a couple potentially adverse incentives:

1. Publishing the exact formula for funding is great for transparency, but then leads to an incentive to game the system to capture funding. Doing things like breaking your project up into many small packages, or reducing the number of maintainers are not good for the ecosystem but may lead to more funding. Also, there starts to be an incentive to juice download numbers.

2. In general, rewarding "risk" with additional funding seems like it creates an adverse incentive. This seems like a really tricky problem, because lack of funding is a major source of risk. It seems like there could be a pendulum effect here if you're not careful. Is there a way to structure the funding so it encourages long term "de-risking"?

That sounds like useful engineering, but not useful science.

3Blue1Brown has a fantastic video on GPT that I'd highly recommend called "But what is a GPT? Visual intro to transformers" [1]. You'll need to have some background in mathematical reasoning, but not too much. He does a great job of drawing pictures to show the mathematical transformations that happen under the hood.

[1] https://www.youtube.com/watch?v=wjZofJX0v4M

Is Superset a decent tool if you're just a single person doing data analysis? Say I have a handful of sqlite databases, and just want to be able to develop some queries / charts. I was looking into Tableau / Power BI / Superset, and all of them seemed pretty heavyweight for a single user, and none of them seemed super easy to get setup locally.

Any recommendations for a good piece of software for the single user case? Or a more convenient way to run the heavyweight tools?

We won't know for sure until there's actual law or court decisions to reference, but until then, it's worth looking at something like the Monkey Selfie copyright dispute[1] for guidance. The article references a 2014 opinion from the United States Copyright Office

> only works created by a human can be copyrighted under United States law, which excludes photographs and artwork created by animals or by machines without human intervention ... Because copyright law is limited to 'original intellectual conceptions of the author', the [copyright] office will refuse to register a claim if it determines that a human being did not create the work. The Office will not register works produced by nature, animals, or plants

The extension of this to AI would be saying basically that the copyright office simply wouldn't extend copyright of an AI created work to any party.

In the case of one of these songs, though, if you wrote your own lyrics, you would still have the copyright to those lyrics, if not the full piece of music generated from them.

[1] https://en.wikipedia.org/wiki/Monkey_selfie_copyright_disput...