An “everything” feed reader. Its a plugable framework that allows you to push anything into an RSS feed reader type interface. Email, Slack notifications, RSS, etc.

I want one place to manage ALL notification settings. So if I want to be notified of Slack messages that contain the word “cat”, I can do that.

I am also looking to add summarization and tagging using a local SLM. Trying to find a method that can run on older hardware.

let's connect, I can hook you up with a custom SLM -> dm on X

I had a similar situation using Jekyll (before Docker) so I wrote my own SSG, and its compatible with Jekyll frontmatter.

Its been a great experience and completely solid to use. Need a URL shortener feature? Code it into the generator. Want some static search? Code it into the generator. Want to generate a newsletter for certain posts? Wire your generator up to SES and do it.

I guess my first thought here is that I would just implement this on my blog platform as a series specific RSS feed. Generating an RSS feed per series and even per tag would be trivial to implement and make for a decent user experience. IMO.

I want RSS with micropayments. I want to consume information in my own interface, and am willing to pay. I am not willing to pay for a full subscription to a publication when I only find a few articles a year that I want to read.

I want Spotify for text, but with a business model that makes sense for all involved.

Does somebody remember the service Flattr from a decade ago? You’d set a fixed amount to pay every month. Like a subscription.

Then, this amount got distributed between the sites you visited. If you only visited one site, they would get everything. 2 different websites = 50% each. And so on. This way there were no surprises in your monthly spending. I still see this as a pretty ideal model.

There's that new thing CloudFlare has that lets you set a price for A.I. crawlers, maybe that could be used to set a price for anybody. If the price was reasonable at all I'd have my crawler pay it for maybe 300 articles a week.

Micropayments have been tried plenty of times and never succeeded. People say they'd be willing to pay, but they're not.

If it was trivial for me to spend 5 cents for a one off article that someone recommended me, I probably wouldn't mind it at all. Payment processor fees make that essentially impossible, requiring more thought and investment in systems like Flattr to group together the small payments.

We need to decouple online payment infrastructure from the duopoly/oligopoly of private corporations that control how and when users can exchange money online.

What if it’s not so cheap, though? I think the problem is that the ad-tech industry offers publishers more than the median reader will pay, and does so reliably, while all of the alternatives mean they have to convince a whole ton of people who’ve been trained their entire lives that the internet is free to start paying non-trivial sums for previously free content.

I would like to live in a world where we do pay for what we use but I’m not sure how we get there by now.

Maybe you wouldn't mind, but nevertheless, it has been tried and didn't work. Here's how it went for Blendle, which had the New York Times, The Washington Post, The Wall Street Journal, the Financial Times, The Economist, Time, and more on board. It wasn't payment processors that killed it.

https://www.niemanlab.org/2023/08/the-poster-child-for-micro...

Is it really about people not being willing to pay though?

I hadn't even heard of Blendle, had I heard that it was a thing and had the mentioned publishers I would've used it.

I've visited those sites so many times and I've never even seen a mention of Blendle.

They haven't been tried plenty of times, what are you talking about? It's been discussed since the dawn of the Internet and the ideas been around since forever, but actual attempts have been very limited in scope. The rails are there though, they're just not user friendly (http://micropayments.fyi).

Unfortunately you aren't willing to pay enough.

Micro payments for journalism don't work

Data Story: Ethics, Productivity, Age and AI:How do these correlate with Developers’ Stances Towards AI at work, 2024

A smaller scale yet interesting look at how ethical issues impact stance on AI in developers. This post looks at data from the Stack Overflow Developer Survey 2024.

So for people building real things today are you saying instead of stringing prompts together with logic we should just sit on our hands for a year and wait for the models to catch up to the agent paradigm?

I’d be interested to see what this does to overhead costs of businesses that actually need a large vehicle.

I have a small horse farm and drive a 2010 F-150. I haul horses, 900lb. hay bales, feed, lumber for repairing fences and building shelters, etc.

If my truck breaks down I am going to have to scrape through used inventory to find something that fits my needs. I don’t need leather, 16 cameras, seat warmers, and a high end sound system. I need a truck that can get done everything I need around the farm, and I need it to be cheap enough that I’m not worried about it getting scratched.

All cars seem to be luxury vehicles now, I don’t know what folks are doing the just need something more utilitarian.

So I discovered the opposite. My reading was always tied to a goal of some sort. I started reading on a kindle and turning off progress indicators and suddenly I had my attention back. I wasn’t worried about improving or reading a certain amount per day or finishing a book by a certain date. I just read.

I have no use for this but my first thought is wondering how safe the features like InfiniteSpin are from hallucinations. That’s assuming it’s using an LLM of course.

The InfiniteSpin is optional and have a strict and extensive prompt that keeps the original structure, goal, and tone, but rephrases the message WITHOUT changing any important info like numbers, names, data, etc.

We only use latest LLM models. We used Claude 3.7 Sonnet (and are now updating to 4).

I’ve been using radicale for contacts and calendar for quite a few years now. Plays nice with iOS Mail and Contacts as well as Gnome Evolution and has overall been a great experience.

I'm a happy radicale user too but I haven't been able to get it to work w macOS mail and contacts. I use EMClient with it which works great.

Thanks, that's very helpful. I'll have to see if importing a Letsencrypt certificate works?