Dave Winer - Scripting.com

Scripting News - 2025-08-09T22:26:43Z

- 2025-08-09T15:06:10Z
I did a rewrite of the FeedCorps page in FeedLand. You get to it through the Reading Lists sub-menu of the Tools menu. There are three lists in the new version. A lot were false starts, they didn't make the cut. I'm always adding feeds to my blogroll and news.scripting.com. Unfortunately I can't say the same for podcasts, which are not hooked up to my podcast app. I really want a hot connection there. We'll get there. The reading lists feature is going to play a big role going forward in the open social web we're building. BTW, I really like the name FeedCorps. I haven't talked about it very much. It's a cause, like peace or freedom. Open those suckas up. Feeds all the way babe.

- 2025-08-09T15:27:22Z
An archive of the previous version, built around GitHub.

- 2025-08-09T22:09:49Z
There's a difference between reading a site in a web browser and it being part of the web. As it turns out what became Web 2.0, all built as silos, could more accurately be called Anti-web 2.0. Underneath all the silos, the heart of the web is still beating. Ready for us to build on it again.

- 2025-08-09T13:30:12Z
A few days ago I worked with ChatGPT to generate an RSS feed of news that interests me. Here's a writeup with a place to comment and perhaps to collaborate on doing this for real. ChatGPT has real limits. This has to be done off on the side. It certainly could be done with their API. I'm head-down on other projects and can't do it myself but as I explain in the writeup, it would plug in beautifully to stuff I'm doing and it would all be open, so a new kind of feed reader is possible. And we could find news from other bloggers that the journalists aren't reporting on, the same way we relied on blogs in the early days to learn about what was going on on the web. It's time to do that again.

- 2025-08-09T22:26:43Z
I had an RSS-specific blog starting in May 2004. I had forgotten about it. Lots of stuff here, I just read through a few months.

I want the web to be like Harvard - 2025-08-08T13:52:09Z

Back in the beginning I wanted the web to be like Harvard.

It's one of the reasons I went there as a research fellow in 2003.

I wasn't satisfied with the web, the density of minds wasn't up to my expectations, so I thought if I brought the web to the minds, that might work better. I wanted to work on communication systems for people trying to work together to solve big problems.

I chose Harvard over MIT, because the web was already strong at MIT, and they weren't friendly to my cause, they were busy trying to reinvent the web when it imho didn't need that. I hoped instead, Harvard, more a school of the liberal arts, would be more open to innovation in communication between people as opposed to between machines. I felt, in 2003, that was a solved problem on the web.

It didn't happen there either, some great stuff did come out of that period, and some of it was lasting, but eventually the university wasn't on board with the web as a community for intellect. Instead what came out of Harvard in that period was Facebook, which is a great accomplishment. I am a FB user to this day and appreciate how well it works at keeping friends from long ago in contact.

I still want an intellectual community on the web. That's why I'm putting so much effort into rebuilding it as a writing environment. All this because the writing products of Silicon Valley are always will be silos. As heroic as Ghost has been in supporting ActivityPub that is not the web. The barriers are too high and the features too limited. The web is the web, it's well defined and well understood. And that's where we have to go, and I want the minds to be there or else it wasn't worth the effort.

UI in AI is the next frontier - 2025-08-08T12:53:02Z

Yesterday I wrote a quick review of ChatGPT 5, after trying a simple programming exercise with it. It was an awful collaboration because it lied to me about its capabilities, leading me to develop a feature in my software to support the capability that it later admitted that it didn't have. I've never had a programming partner do that to me. I've had people who never showed up for work. Or people who were in so far over their heads that it would have been better if they had never shown up. It lied repeatedly, and when cornered it had the audacity to gaslight me. "I understand why you feel that way." Oh nothing pisses me off more than that.

I don't understand why other people don't review it based on its functionality, after all it isn't just a money sink, or a technological breakthrough, it's software that we pay actual money for that is meant to help us in our work. I don't know if it's by design that it's such an asshole. I have a feeling that people are going to focus on this eventually.

ChatGPT is not human, it's like the world of Rick and Morty. And we're meant to be one of the many versions of Rick and/or Morty, but which is the real one? Before we lose sight of that, let's keep our heads on, we're still the ones discovering stuff and being creative and entertaining other humans, not the machines. It'll all work a lot better when they're programmed to behave like computers, not these vain and corrupt pseudo humans. Still a breakthrough in search. Really sticks it to Google. Yaaay. That's good enough for me.

I'm more worried about AIs killing us by overwhelming our nervous systems than taking our place at work. Has anyone actually used any of the software an AI has written? I know people who have never written software before are in awe, as I am about drawing characters I make up, but how does the software it builds rate? If it lies to itself like it lies to me, the software doesn't work at all, even slightly. I'm pretty sure "garbage in garbage out" applies in the new world of software.

• 2019: "I'd love to understand how one can be a programmer and not always be searching for the truth. Unlike just about any other profession I can think of, ours depends on telling the truth. You can't lie to a compiler is one of my programming mottos. Garbage in garbage out."

The tech industry doesn't get its relationship with paying human customers. Had a problem with my cell phone service. I started getting emails saying it was shutting off my service because my credit card was being denied. Went to the account on the web and it looked fine. The messages kept coming. I reached out for help. The people couldn't understand what I was saying. I eventually gave up because I didn't understand either. I might have two accounts, I don't know, in some ways it looks like two, in other ways just one. I presume they have a way to find out, assuming there are actually any reasoning empathetic human beings who care about doing quality work. If we were talking to each other human to human, not as peripheral devices for a computer network, we might have figured it out in five minutes, instead of having the customer report it on their blog after weeks of bullshit that did not uncover the truth. We still don't know how many accounts I have.

The company is Google. They're good at dealing with imaginary customers not real ones. If they ever understood how to communicate with people, they have forgotten. And of course there's a real possibility that all my conversations were with their ChatGPT. If that's true, no wonder we were running around in circles never getting the answer.

- 2025-08-07T19:51:07Z
ChatGPT 5 is even more of a bastard than 4.

- 2025-08-07T20:10:05Z
That said, it got pretty far toward solving the problem I asked it to work on. Here's a demo of a page it put together for me of the top 25 articles in US news.

- 2025-08-07T18:05:28Z
A summary of recent work via my WordLand blog.

- 2025-08-07T17:34:03Z
New version of feedlandSocket. It's now an NPM package you can include in Node projects. The demo is more useful, and there's a video of what it looks like as it scrolls through the JavaScript console. WebSockets + feeds. A fairly important component of an open social web system.

- 2025-08-07T11:22:39Z
I wanted to put together a demo of a very simple but interesting Node.js app, so I hooked up with replit, and was surprised to find out that it's now an AI bot. But when I asked it to make a sandbox for this app, which is in a repo on GitHub, rather than take the direct route, and run the demo.js app (which is what I asked it to do), it concocted a pointless user interface, that hid all the interesting bits. This was a demo for programmers for crying out loud. I want to show them the machinery in motion. I'm going to try again today.

- 2025-08-07T11:24:57Z
I recorded a good podcast yesterday about how I develop new stuff like podcasting, and how it only works if there isn't a huge dominant platform vendor to FUD the project. That's why RSS worked, sort of -- we did get FUDded by the RDF folk, who had a few famous people on their list of co-authors, and it put a big scar on RSS when it was just starting to grow. We cleaned up their mess after a while, and then it boomed, thanks to the New York Times. This is what usually happens, in markets that are controlled by a big company. Google has that effect, so does Apple as does the W3C. They force everything to stop and go through a process that simply doesn't work, it doesn't yield innovation in the market, because it wasn't bootstrapped. That's why I don't like the idea of "Podcasting 2.0" whatever it is. Adam Curry says on Masto that it's all a big misunderstanding, but that's what the RDF people said too. Who cares what the name is. Everyone cares, when it comes to the name of a standard or protocol. And when you have an Apple or Google or IETF trying to confuse people about what you're doing, well you can't make any progress on interop. They freeze everything because users and developers will wait for them. Of course that's what "Podcasting 2.0" is all about. To make all the ideas flow through one place. And whoever they are (ChatGPT says it's Adam Curry btw) they didn't even bring the power of an Apple or Google to the table. They have nothing other than their aspiration to own the name of an open format and protocol that made the world a little happier for a while. For that of course someone has to try to own it. That's basically what I said, in a very long-winded way, in the podcast. I also talked about ways we could move forward anyway even though there is a 800-pound would-be gorilla trying to own our playground. Maybe I'll just release the podcast as-is and ask you all to indulge me for my need to say the same thing far too many times with so much vigor. :-)

My uncle Ken - 2025-08-07T13:02:51Z

This is a picture of my dear departed Uncle Vava taken in the early 1970s in my parents' house in Flushing. It just showed up in my On This Day list on Facebook, thought it belongs on the blog too. .

- 2025-08-06T20:43:26Z
A Brian Lehrer segment on specialized high schools in NYC. I went to one of them, signed up for the test mostly to get out of school for a day, and got in. Back in those days (the early 70s) no one studied for the test as far as I know. It has become very competitive and there's an issue of the racial makeup of the student body. As always Lehrer does a great podcast.

- 2025-08-06T20:42:22Z
Milestone: Chuck Shotton got his FeedLand up today. That's why I spruced up lists.opml.org, that's how we're connecting our servers. He's building this stuff to flow into his LLM via the FeedLand websockets interface. He'll be updated on my subscriptions, and have his own, and the news will flow into his AI system. All of it can emanate from anywhere RSS is supported with a focus on WordPress. It's the secret sauce. 😀

- 2025-08-05T13:27:17Z
news.scripting.com is still going strong. Open to the public, no login.

- 2025-08-05T21:53:48Z
A podcast I recorded this morning, prime time, while getting things done, and having ChatGPT getting in the way. It needs to become more invisible, there's no suspension of disbelief when you're working with it. I think we can do much better at finding a robot that can really augment human intelligence. This is awful stuff. We have to work on these dynamics.

- 2025-08-05T20:50:27Z
Whenever you have to get something done with a company, get ready for lots of phone tag, waiting on hold, talking to bots, getting screened, trying to convince a computer that you have legitimate business, and no, what you're looking for isn't on their website (believe me I looked). The stupid thing about it is that ChatGPT is becoming more like those things every day. Companies have built awful systems for getting anything done that might eat into their profits. Google is the absolute worst. Even for services that cost real money, they absolutely will not help. You better hope everything goes perfectly if you buy their service.

- 2025-08-04T12:54:05Z
WordPress will make a much better open social web server than any other software out there. We can all develop any component around a solid, documented, simple and widely supported open source API.

- 2025-08-04T20:49:33Z
My contribution to 715-999-7483.com. Now it's your turn!

- 2025-08-04T16:19:32Z
Since the govt is no longer funding NPR maybe they could stop bending over for the Repubs. Lay it all out there, stop spinning all the crazy fascist authoritarianism as both-sides and normal partisan politics. They know we're in a lot deeper. Since we're now paying the bills, how about plain facts.

- 2025-08-04T14:00:42Z
Gaslighting is everywhere. ChatGPT just said to me "I understand why you feel that way." It has nothing to do with my feelings. I don't have feelings about computers. It lied to me over and over just now. I said you're lying to me. "I understand why you feel that way." As if it were the all-knowing feeling-inferring god-like creature it is not. The real question for me is this -- Does Open AI program it to be this way. Think about the opportunities it has to introduce true feelings of insecurity and worthlessness. That's the purpose of gaslighting. It's evil.

Antisemitism is everywhere - 2025-08-04T14:33:15Z

Academics, who speak scientific language, won't say "there's no antisemitism at UCLA" because they know that's not true. Antisemitism is everywhere. The question is how limiting and dangerous is it. And how often is it encountered.

I grew up in the 1960s in a Queens neighborhood where antisemitism was a real thing. We were blamed for World War II. If it weren't for Jews my uncle Pete would still be alive. That was the story. No doubt the kids heard that at home.

Antisemitism was bred into us at home. Because of all the abuse my parents and grandparents, uncles and cousins, suffered in Europe, before coming the US during the war, their own idea of what it meant to be Jewish was not entirely positive. You can't help but feel responsible in some ways for the abuse you suffer. And you can't help but pass that on to your kids.

There was and is a fair amount of pain associated with being Jewish in the United States. Fact.

But: At UCLA? Harvard? Columbia? Please.

At the same time scientists can admit that there is antisemitism at these places we must also acknowledge that to single these institutions out among all, is complete utter total fucking bullshit.

That's a scientific appraisal, btw. Provable in peer-reviewed publications (though not submitted).

Antisemitism is complex. We should be hearing, openly, what exactly the case is against these universities. The government is theoretically representing the Constitution and the people of the United States. The old "no taxation without representation" thing. Let's hear the case in clear terms, and why the universities are supposedly so bad.

- 2025-08-03T16:38:41Z
A podcast about a podcast users' API.

- 2025-08-03T15:22:09Z
It's always a good idea to get a second opinion with AI stuff. ChatGPT may give you a convoluted answer where Claude.ai gives you a concise one.

- 2025-08-03T14:04:22Z
I've been working on the top level of WordLand, and finally got to a place where navigation feels good, like this is the right track. So I took a snapshot so I can come back and look at this later.

- 2025-08-03T13:22:08Z
As you get older your memory gets less reliable. It makes programming more of a challenge because as your software gets more features, there's more to remember and at the same time you're getting older. I wonder if there isn't some way to use ChatGPT to augment the aging mind dealing with more software complexity. It's very much in line with the idea I've had for a long time of putting all my writing in an AI database so I could then ask it to edit it down to book length. Or get a table of contents of what I think and then be able to read chapter-length sections on, say platforms, or how important prior art is, or what interop makes possible, and why everyone should give back when they take from open ecosystems. Each one of those topics has lots of associated stories over the years, but I probably couldn't find most of them, but an AI database certainly could. How to set that up? I've not figured that out so far.

- 2025-08-03T13:26:05Z
One of the more depressing things in having so much code that I can easily search, is the number of times I've rewritten the same code without remembering I had written it before. Yesterday I did it knowingly, I wanted a function that could tell me if one of a set of categories applies to a WordLand draft. So I could say "don't list drafts with 'linkblog' as a category." It could have other categories, and there could be more than one category to exclude. I wanted to take the time to write perfect code for this one problem. Not dense, not particularly efficient. No matter how inefficient the code is, on today's hardware such a function couldn't use any time at all. It's fair to say I've solved this problem before, but there was always more to it. Here's the function. Now you know something about what I think is readable, understandable, respectful of a human more than I am of the computer. I know the computer would come up with much more dense and tricky code, but it doesn't have a mind, and I do. At times it can be relaxing, I magine, to not have a mind. And btw, in a couple of years this layer of code will be obsolete. We already are able to tell the machine how to do this in human language, it understands what we mean. One tiny little but hugely significant breakthrough made possible by ChatGPT and its cousins.

- 2025-08-03T13:40:05Z
A basic question I had about the ChatGPT agents that I can answer now that I have the feature, is whether or not the code you create can run on a server, where you can give it a URL and make it an endpoint other networked software can call. Or if it could run periodically, say once every five minutes for a function that was creating an RSS feed anyone could subscribe to. The answer is no -- it can't do either of these things. I'm sure they could do it at a technical level, but they don't want to host applications. But now I may understand better why they want to make a web browser, I bet you will be able to call these agents from apps that run in the browser. And in their case, they might not even have to support JavaScript? Heh. A wholly different programming model? Maybe I'm overestimating how much they're biting off? I wonder if anyone at OpenAI reads this blog and might want to get me in a tighter loop, so I can be among the first to try new features like this, rather than, in this case, among the very last.

Let's all be anonymous! - 2025-08-03T22:07:58Z

What if we all wore ICE-style masks as we went around doing our business.

If a baseball player's uniform didn't have a number or name.

Let's all be anonymous. It's very practical, we don't want to get doxxed either!

- 2025-08-02T21:34:45Z
I think I'm going to put in my will that ChatGPT should run the Dave Winer persona on all social networks, and my blog, as long as the money lasts. It would tell stories that I would likely tell, take political stands that I would take, draw meta-pictures of my sad and depressed programmer friend and a cute and adorable kitty getting into all kinds of trouble. The seasons would come and go, and there would be Dave, still diggin. And of course he would continue to develop software, using some of the greatest tried and true tools, reminding everyone of how great Frontier is -- but -- if only it ran on Linux. The long-lived fearless and fully paid-up version of Uncle D.

- 2025-08-02T16:41:29Z
A change would do us good too.

- 2025-08-02T16:40:23Z
I got hoisted with my own Picard.

- 2025-08-02T14:47:40Z
I have a lot of code written before ChatGPT. Sometimes, as I read my old code, I wonder how the h*ck I ever figured that out without it.

- 2025-08-01T15:54:38Z
Podcast. It's time for things to change. Two examples, Wired and Harvard. Change was always coming, but now you can't turn away from it very long before it re-appears.

- 2025-08-01T15:25:17Z
Welcome to August. More than half the summer is behind us. We're still here. The shelves in the supermarkets are full of good stuff. We've had about the normal amount of hot weather, a good amount of rain but not too much. The Mets are doing OK. They opened a Pho restaurant in Bearsville and it's good. Health, not too bad, all things considered. Still diggin!

- 2025-08-01T15:37:09Z
Source OPML for July 2025 blog posts.

#rss