Nostr Web Client

3 minutes - create SEO metatags from rss feeds and chatGPT (or llama, etc). https://www.youtube.com/watch?v=RLN72EwRm6Q

Until everyone moves to Nostr, X/Twitter should identify censored content with actual blacked out content in tweets labeled with who requested the censorship.

Replying to

PABLOF7z

A few days ago I had lunch with nostr:npub1dergggklka99wwrs92yz8wdjs952h2ux2ha2ed598ngwu9w7a6fsh9xzpc and he told me about this awesome idea:

> A relay that rejects long notes.

This morning I woke up thinking about it.

And now it's live.

wss://140.f7z.io

140-character limit. With a pretty strict moderation, so spam should be basically non-existent.

It's also sucking short notes from other relays, so it's global should be pretty useful.

Check it out!

PS: No coincides; I saw this nostr:npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6 note just now 😂

@note10nvzcqzst4c5yee279u2j5qpzzkffhwxn5qf5t7jxqfx53h6agdqte34ay

c3

Cole Albon 2y ago

I made wss://140.f7z.io the default relay for cafe-society.news nostr global page. https://github.com/colealbon/cafe-society.news/commit/5900de7490997f4ba5e519a3b8a43a4981a6f5fb

Replying to

jb55

Almost done the filter interface for nostrdb. Soon you will be able to query it like any relay. Having a fully queryable local relay inside clients will be wild.

c3

Cole Albon 2y ago

I'll likely be using nostrdb for storing a ~1 megabyte machine learning model so people can share their preferences real time. It would be a long anticipated long delayed conversion from blockstack/stacks. I've already spiked on yjs (I think I got yjs from the nostrdb code) Any objections?

Replying to

Lyn Alden

Im always shocked that retailers still exist. I rarely do in-person clothes shopping anymore.

I had been meaning to go to Macy’s to buy some coats but kept delaying it. So I go today, get some coats, and it turns out they are 60% off for Black Friday even though it’s Wednesday, and even though there is no line/craziness today. So I’m like, “alright, thanks”.

Huge store, so many products, not many shoppers, and they must have a decent amount of unsold product for certain things. Based on what I paid for these coats I can’t see how they made any money from me and this is like, my once-per quarter trip to Macy’s.

I’m glad they exist so I can go sometimes, but I just look around and think “how are you all still here!?”

c3

Cole Albon 2y ago

real estate prices, duh!

Replying to

iefan 🕊️

Binance CEO pled guilty to money laundering. BINANCE will pay $4.3 billion in fines per the WSJ

Remember: Not your keys, not your coins.

Also, don't let your hard-earned money be used by these transnational criminals.

c3

Cole Albon 2y ago

I assume the criminals you refer to are the government agencies collecting the fines because "money laundering". Do you know if the fine money comes out of customer funds?

c3

Cole Albon 2y ago

I was on the marketing team at GoDaddy during the SaveToby controversy and the banned Superbowl ad controversy. Both stunts landed a phenomenal amount of free press and subsequent traffic. I was on the Sigaba sales team when they rented 6 heavily decaled humvees to escort journalists around Las Vegas during Comdex. As a salesperson and aspiring software engineer, I converted the melissa virus into an email guesser to bump my sales leads, making me one of the first to transition from fax spam to email spam. Prior to the Mt Gox collapse, I commandeered the "payments" weekly developer meetup in San Francisco to steer developers (and MicroStrategy consultants) to the bitcoin meetup. I got personal training from Regina Brady who's favorite direct marketing ad campaign was an apology letter for an event that never actually happened.

In short, I've watched teriffic guerilla marketers apply their trade and I've done a bit myself.

I see guerilla marketing written all over the Sam Altman business.

P.S. I'm looking for javascript work and might be guerilla marketing to you now.

Replying to

Edward Snowden

It just occurred to me, due to a nostr bug, that it might actually be pretty cool to have social media hide *who authored* the posts on your TL to force a kind of double-blind perspective. You'd know the post or reply is (or probably is) from one of the people you follow, but you wouldn't be so quick to agree (and perhaps might even be more willing to disagree) if you aren't just smashing the heart button as a reflex because you like the guy. It might make discussions more communal and disagreements less personal if you knew that you're still talking to someone you generally like and respect, but you don't know whether it's "that one guy on the internet" or, like, your actual Mom.

(Nostr, due to ramshackle Tor support, often displays notes from people without correctly fetching their profiles for me, so I find myself liking or replying or following people without having any idea as to their nym or avatar)

c3

Cole Albon 2y ago

like this right? https://cafe-society.news/nostrposts

Replying to

Edward Snowden

OK, so you guys showed me McDonald's jacked prices at least 130% on the core budget item: the mcchicken. Mentally recalculating McDs as "literally more than twice as expensive is a doozy, sure but what I really wanna know: did they do any shrinkflation on top of that to really wring out the margin? Smaller patties, cups, fries, cheese or something? Or was it all just pure price action?

c3

Cole Albon 2y ago

You can engage with whomever you want. I will just point out that you are arguing for more french fries or "...cheese or something". You are approaching a certain age where you should be happily paying a fortune for smaller higher quality portion sizes.

c3

Cole Albon 2y ago

Head of some of the biggest AI projects on earth, Andrew Ng, says more than once in this interview that he expects AI processing to move towards the edge (on laptops). I agree, ChatGPT is doing it wrong (my words not his). If you try to anticipate every single question from every single user on earth with one huge 8 figure training session then you have to bombard the press and congress with advertising just to break even. Great for transferring investor money to advertisers like Alphabet, and MSN, but otherwise huge waste of electricity. https://www.youtube.com/watch?v=KDBq0GqKpqA&list=TLPQMDUxMTIwMjONZgcGdpBDgw&index=1

Replying to

Edward Snowden

I think inertia is leading us toward a quantified state, in which the costs of tracking and categorizing objects — primarily people, phones, cars, and money — as they move from place to place becomes so low that most institutions are doing it by reflex, without either considering or caring for consequences. Eventually, somebody is going to exploit the gap in information (between what is available to ordinary people, and what is available to organizations) in a way that causes enormous, historic harm, as happened in WWII when a particular state realized that IBM had compiled a database that was perfect for populating a list of names of those who would soon be marched off to camps. And that was just census data. Now they know what you read, where you go, who you're with, and how you'll vote.

The unfortunate reality of contemporary political thought is that consequences do not exist until they are felt. Rare is the chicken that fears the farmer. That means we've got two options: we either wait for the pain, or we stop waiting for political consensus.

nostr:nevent1qqsq94302eeh5r82vytl2r3n0hdhtmeg3ehxxcavk767j5xlqufmd2spzemhxue69uhhyetvv9ujumn0wd68ytnzv9hxgq3qpvgcusxk7006hvtlyx555erhq8c5pk9svw57snlxujpkgnkup89sxpqqqqqqz27cnxw

c3

Cole Albon 2y ago

Journalists and whistleblowers are too eager to self censor when it comes to exposing names and actions taken by the people who are members of organizations. Organizations are immune to accountability and punishment. People are not. It is changing though, when Russel Brand was cancelled, some journalists were quick to publish Dame Caroline Dinenage's name, her husband's political connections, and letter she wrote. As a news consumer, I can't do much, but I can at least seek out and click more on journalists who write about people instead of vague institutions.

Replying to

BITwise 🐳 >>= RIGHTish

You say its boring around here but I see you follow like 60 people. Thats like complaining its wet when youre sitting in a pool.

I follow 1000+ people and my feed is constantly full. Yes there's a lot of bitcoin and nostr content but there's also a ton of completely unrelated content as well. Art, music, homesteading, agriculture, diy, cars, cooking, just to name a few.

c3

Cole Albon 2y ago

I wrote a tool to train your own filter against the global nostr feed. More signal less noise from the main vein. https://cafe-society.news/nostrposts. The default is my own personal preference for breaking news, less bitcoin. You can delete the default and start training your own filter - all client side training and prediction.

Replying to

bumi

spam and impersonation is increasing on nostr recently… :/

c3

Cole Albon 2y ago

"spam and impersonation is increasing on nostr recently… " - not in my hizzy. I trained against 4000 nostr posts to suppress bitcoin price articles, zapathons, spanish poetry, vulgar language and whetever else. https://cafe-society.news/nostrposts. Not good yet, but much better than IRL. Here is the winknlp compatible model if you're curious. https://github.com/colealbon/squelch/blob/50772762d7ed6b7c4f920903b243083972e069c2/src/defaultClassifiers.ts#L38

Replying to

verbiricha

Which clients support private mute list? I have some words and tags there and the ones I use are not respecting it so far. I think is important to be able to choose what you mute in public and what you mute in private. #asknostr

c3

Cole Albon 2y ago

https://cafe-society.news/nostrposts supports private mute list. read only. still pretty buggy work in progress

Replying to Cole Albon

yes! cafe-society.news/nostrfeed considers a post read if it is an 80% string similarity score.

c3

Cole Albon 2y ago

nostrposts, not nostrfeed - sorry

Replying to

jb55

Do you think “read” states would be useful? Jump to the last unread note in your feed?

c3

Cole Albon 2y ago

yes! cafe-society.news/nostrfeed considers a post read if it is an 80% string similarity score.

c3

Cole Albon 2y ago

I am trying to filter global feed with machine learning, specifically wink NLP bayes filter. My problem is that big long posts score better than short posts. Anyone know how to solve that problem? Suggested books or articles? https://cafe-society.news/nostrposts.

Replying to

utxo the webmaster 🧑‍💻

There is borderline anime child porn in global. Much is not marked sensitive. This makes us look really bad to new users imo.

I think global feeds should be more of a hidden/advanced feature if we are going to have them at all.

Which we probably shouldn't have either, there are better ways to do discovery...

Thoughts? 🤔

c3

Cole Albon 2y ago

- start with global nostr feed.

- add a button to block addresses

- add ML buttons to train ML

- train ML with a few thousand posts

- https://cafe-society.news/nostrposts

c3

Cole Albon 2y ago

I've been blocking nostr keys and training a wink nlp model to filter out bitcoin shilling and other noise in the global feed. The model is here: https://cafe-society.news/classifiers/raw blocked keys here: https://cafe-society.news/nostrkeys/raw scored, filtered feed is here: https://cafe-society.news/nostrposts. You can train your own.

Replying to

Guy Swann

We should stop and take an enormous appreciation for the total lack of control and censorship of the RSS protocol.

Its easy to forget when things work really well, but holy fuck through all of this, through the social censorship, through the advertiser attacks, the video hosting platforms, the search engines, and every other damn thing under the sun... if you have the RSS feed of the podcast you trust or value, you can always find that fucking show.

Imagine how much worse all of this would be without that avenue still open to us to share ideas?

The reason I bring this up is simply to remind us that this CAN work. It has worked. #Nostr could be the backbone for the social world.

c3

Cole Albon 2y ago

I run cafe-society.news - a site that aggregates user configured rss feeds with the nostr global channel. Filtered by client side machine learning training and prediction. I'm looking for ideas how to connect the reader to the author and other readers financially.