There does seem to be evidence that X (formerly Twitter) is a dying platform, but what surprised me here is that longtime platforms like Snapchat, Reddit and even Pinterest get more MAUs than X - and this is more October 2025:
Are they not the same users? They're all part of the same org, so it seems likely that accounts for the others are mirrored or auto-created or whatever.
Reddit is on the first page or at the top for well over half of my searches. Sometimes I find myself in complete physical-memory typing -site:reddit.com.
X/twitter is #1. reddit is #4, NYT is #11, Fox is #16, AP is #18, CNN is #21.
That's not a dying platform as much as you clearly wish that were true. The question is why are you so hellbent on convincing people something that is clearly not dying; is dying?
If X is dying, CNN, AP, Fox and NYT are stone cold corpses with reddit having its last gasp.
Threads surpassed X in DAU only for mobile, with a slow decline shown in X (see plot), with "dying" being a misleading word. For web, X has 18x more DAU than threads [1].
Total daily active users (all access methods) is overwhelmingly for X. I can't find the trend for web. Please post the link you found.
- If I ask Claude to go and build a product idea out for me from scratch, it can get quite far, but then I will hit quota limits on the pro plan ($20pm).
- I have not drunk the Kool-aid and tried to indulge in ClaudeMaxxing (Max plan at $200pm). I need to sleep and touch grass from time to time.
- I don't bother with a Claude.md in my projects. I just raw-dog context.
- If I have a big codebase, and I'm very clear about what code changes I want to make Claude do, I can easily get a lot of changes made without getting near my quota. It's like Mr Miyagi making precision edits to that Bonsai Tree in Karate Kid.
My last bit of advice - use the tool, but don't let the tool use you.
Methane is the most harmful gas right now, but the solution for some reason is more money to the gov. Maybe we should do something about that meat industry
The Twitter social graph was an amazing data asset. I worked at a consumer insights firm and the data on followers/followings was quite powerful.
Using a custom taxonomy of things (celebrities, influencers, magazines, brands, tv shows, films, games, all kinds of things), we could identify groups of people who liked certain things, and when you looked at what those things were, it gave you a way of understanding who those people were.
With that data, you could work out:
- What celebrities/influencers to use in marketing campaigns
- Where to advertise, and on which tv/radio channels
- What potential brands to collaborate with to expand your customer base
- What tone of voice to use in your advertising
- In some cases, we educated clients about who their actual customers were, better than they understood themselves.
One scenario, we built a social media feed based on the things that a group of customers following a well-known Deodorant brand in the UK would see.
When we presented that to the client, they said “Why are there so many women in bikinis in this feed?”
The brand had repositioned themselves to a male-grooming focussed target market, but had failed to realise that their existing customer base were the ones that had been looking at their TV adverts of women on beaches chasing a man who happened to spray their Deodorant on them. Their advertising from the past had been very effective.
That was the power of Twitter’s data, and it is an absolute shame that Twitter went the way that it did. Mark Zuckerberg once said that Twitter was like “watching a clown car driven into a gold mine”.
I’m pretty sure he must be delighted with how things have panned out since.
This reads very dystopian. You are not optimizing to understand people, you are optimizing to weaponize that understanding against them.
When you know what someone will buy based on exploiting their unconscious preferences, and you are paid to increase sales, you will do it. Especially if your competitors are doing it too.
And this happens at scale, invisibly. People never see the manipulation.
In any case, it is not useful for most people. It is useful for the people doing the deceiving.
The tech is interesting and useful, no need for the scary moral framing.
The original application of the entire field of data science or ML is/was actually based on this paradigm of finding "unconscious preferences" (your words) and hidden patterns. How one chooses to deploy the tech should be judged on its own.
On the current trajectory of tool/data abuse where Palantir et al. are leading the way, this is very low on the sinister scale.
I am not disputing that the tech is interesting. My point is about how it is being applied. The examples above are not about understanding people, they are about exploiting their latent preferences (before: "unconscious preference") for persuasion at scale.
Attempting to normalize that by saying "Palantir is worse" does not make it any less manipulative and sinister.
And to be more on topic, Twitter's value as dataset is overstated. Hardly the panacea people make it out to be.
To not frame the amorality and negative effects centrally and primarily is to be dishonest. There is absolutely not a single person whose wage doesn't rely on not seeing it, that doesn't see that that entire branch of tech has strictly negative value to society.
But of course, line must go up, and it's not you personally being negatively affected, so it doesn't matter.
It doesn't need to lead to AGI or a good coding agent. Some of the only people who are actually profitable in the LLM industry are the people making actual chatbots. There are several bootstrapped startups that run open-weight models with a $10 or $20 monthly sub and make millions in profit off of inference from people just talking to the things, usually for character roleplay / "AI boyfriend/girlfriend" stuff etc. Some of them even took those profits and invested it into training their own bespoke models from scratch, usually on the smaller side although finetunes/retrains of Llama 70b, GLM, and Deepseek 670b have also been done. Grok could probably be profitable if it targeted this space, as the most "intelligent" conversational/uncensored model.
This is already presupposing that profit even matters, though. Musk already burned some $50 billion dollars to control messaging on political discourse with his acquisition of Twitter. It was not about money, but power. After you already have infinite money, the only thing left to spend it on is acquiring more power, which is achieved through influencing politics. LLMs represent a potentially even better propaganda tool than social media platforms. They give you unprecedented access to people's thoughts that they would probably not share online otherwise, and they allow you to more subtly influence people with deeply-personalised narratives.
As an aside that quote from MZ does bother me. There's more to making a web-scale human rights respecting (because it has to, it's the internet, social media needs guidelines) than just making money (which Zuck doesn't seem to care much about anyway if he's sinking apparently billions into metaverse while having no account support)
Of course he would only see it through the lens of cash. I have no idea how profitable Twitter was under Dorsey but it felt the spirit of the company at first was relatively neutral, it was a tool, it was what Jack came up with
Zuck replaced people's email addresses[1], the feed has been wildly unchronological for years. Fix some of those problems wrt. lack of user respect and maybe you can make statements like "all else being equal, clown car goal mine". Or was it "dumb fucks"[2]?
It _was_ a great asset, however, just like models need proper data, as soon as musk removed the clamps on valuable social signals, well, he basically took a dump where he intended to eat.
I wanted to share the demo to show what's possible with Claude Code in a short window of time (I only started building this on Friday evening, and I've spent probably no more than 6-7 hours in total this weekend on it).
Point taken on not doing a write-up on this. I think I will write a blog post about my approach and learnings and then share later. I'll let you know once it is up.
I thought it might be worth sharing that I'm a fullstack developer with about 20 years of industry experience, but I didn't study CS at university, I studied Management & Systems instead (Business Studies with Maths, learning about Linear Programming, Time-Series Forecasting, Critical Path Analysis, Monte-Carlo Simulations, and Systems Thinking).
I have a GitHub profile here so you can see all the open source software I've written over the years: https://github.com/paulbjensen.
I'm also the author of Manning Publications' "Cross Platform Desktop Applications", a book about Electron and NW.js. https://manning.com/jensen.
I still write code in my day job, but I'm having a lot of fun creating PoCs with Claude Code in my spare time.
And if that description about a category of HN users who only became software developers for the high pay was referring to me, I thought it would be worth mentioning that my friends at university in London back in 2006 went into Investment Banking as that had the high pay, but I took a different route and became a self-taught programmer.
I never did it for the high pay (it didn't exist in London back then). I did it because I grew up around computers (my dad was a software and hardware engineer), and I realised that I love creating things with them.
Some background. I'm a 20-year software full-stack developer (started with Ruby on Rails, then moved to Node.js. These days doing a lot of Next.js/React, and some Svelte in my spare time) - as well as Astro.
So, last year after finishing a contract at Volvo Cars, I took some downtime and dabbled with learning Svelte, and managed to recreate a silly little prank I made when I was working at a web agency called New Bamboo many years ago. That little prank then got iterated on, and is now the music editor app at https://lets-make-sweet-music.com.
After that, I created a simple ball-table game using Svelte and Threlte, which is a wrapper around the excellent ThreeJS library. Once I realised that Threlte had support for Rapier physics, I realised that I could make a game, so I made this: https://3d-garden.vercel.app
Usually when you hit a blocking issue, you'd google around and probably end up on Stack Overflow. But last year I increasingly found myself using ChatGPT (I have a $20pm subscription) to ask questions and copy/paste snippets of code for suggestions on how to resolve a blocker.
That process turned out to yield some good results, so I was able to iterate on my ideas and get quite far. My only real limitation is that I have a tendency to lose interest in projects and end up jumping from one project to the next. I also started looking at making a LLM-bot interface to try and create a cyclical loop of bots making other bots do things and implementing the feedback loop.
I also managed to implement an isometric game engine based on an example I'd made using Löve2D (a game engine in Lua), but this time written in TypeScript and using HTML5 canvas. I managed to work on that for quite a bit, but as usual ran out of passion.
In theory, if I can resurrect my passion on it, then I could get it to the stage where it would be possible to play a SimCity-2000/3000 like game entirely in the web browser.
Based on the work across those side projects, I managed to extract out a library for emitting events http://github.com/anephenix/event-emitter - the purpose being to easily decouple Svelte components and make the logic within those components more modular.
I took up the opportunity to work with a former boss at an e-commerce company based in Sweden, so I recently moved here. It was during that time that I learned that the team had been dabbling with Claude Code, which I hadn't used at all.
I then decided one weekend to have a go at trying to build a pixel editor for Babsland, but rather than do it by hand, I decided to try out Claude Code with the Zed Editor (I needed to switch from VS Code because my apartment didn't have electricity for the 1st week, so I relied on charging devices at work and a great hotel - Scandic Centralen in Gothenburg). I realised that VS Code being an Electron app (after all I wrote a book on Electron and Nw.js) was using a lot of battery, so I switched to Zed.
In a couple of hours at the hotel, I managed to build the beginnings of a pixel editor - which is now https://www.babspixel.com. It was a revelation of an experience.
You can see an example of dragging and dropping a MIDI file of Guns 'n Roses Welcome to the Jungle and playing it with the guitar instrument.
I also managed to code up a silly little game called "Mr Spanky's Meatballs" where you lob meatballs are characters and try to survive for as long as you can, again built in a few hours with Claude Code.
The golf game was an idea I joked with making with my former colleagues from New Bamboo (we keep in touch online to this day, a testament to how good that group of developers were). I then started coding it with Claude Code on Friday evening, and posted about it less than a day later.
I'm using the Claude Code Pro subscription ($20pm), not the Max subscription, so I regularly run into the session quota limits, but that's fine because that allows me to sleep and have a normal work/hobbies/life balance.
As someone who I guess you could describe as a creative developer, I'm so excited by what is possible with these AI tools, and I think that in terms of the software industry, the genie is out of the bottle - I think the biggest challenge is going to be ensuring that the quality of the output improves and meets the threshold we expect - for example, that game has various bugs which others in the comments have noted.
For games and fun side-projects, I'm happy to delegate the coding to AI tools and perform a manual feedback loop of asking it to make one feature/fix at a time, reviewing the output, and either iterating on that, or moving onto the next feature/fix to do.
As for work, I'm not yet confident enough to use AI tools the way I have been using them with side projects. In fact the first thing we're doing with our new project is putting E2E tests in using Cucumber and Playwright so that we can verify that the application (the combo of backend, frontend, databases and other services) works and that we can have confidence in deploying it when those E2E tests pass.
I think that we are in a major transition phases in our industry, and that there is still work to be done to yield the results we want from AI (in terms of quality, accuracy, not hallucinating libraries or API calls that do not exist).
I also want to say that having spent years working with React, I really love the developer experience of using Svelte, and I choose to use it for my side projects now.
https://www.statista.com/statistics/272014/global-social-net...
It would be really interesting to learn if brands and advertisers are seeing the same thing?
reply