Joshe

I think it's possible Trump and the Project 2025 people are planning an actual hostile takeover and won't even try to cheat in 2028. If you know what the Heritage Foundation is and how it operates, and how deeply embedded in our political system it is, when you hear it's leader say "The revolution will be bloodless if the left allows it to be.", and when you realize that dude is serious as fuck and has serious plans in the works, backed by billions of dollars, that statement reads to me like a direct confrontation. Not some slow soft coup. Just a theory, but it seems possible they just outright say "we're taking over, no more democracy". I know it sounds crazy and hard to imagine how they'd pull it off, but these people have the deepest pockets and influence and a vast network of operatives (tens of thousands), and they're building underground bunkers with mini cities inside them. Maybe not likely but I'm keeping it on the table.

I considered that, haha.

I'm sure there's going to be a massive response to this. It'll be interesting how it plays out. Basically, this is a huge pressure campaign to change voting laws before the midterms.

You gotta download Claude Code CLI and then just send it that. It will build it for you.

lol, the solution to your problem.

I would also include an instruction to have it run OCR on all the scraped images so you could search the image posts as well.

Just have Claude Code do it for you. Here's a prompt that should get you an app to fetch the blog daily and let you search it, all hosted locally: # Build prompt: Actualized.org blog reader Build me a GUI app that lets me read and search every blog post from https://actualized.org/insights — a local, offline archive with full-text search that stays current on its own. Start with the backend: getting all the posts off the site, stored durably, OCR'd where the content is locked inside images, and kept up to date via a daily automated check. The GUI comes after. You own the architecture. Everything below is reconnaissance and constraints, not instructions — make your own engineering calls, and tell me if you disagree with any framing here. I'm not a developer, so flag tradeoffs in plain terms. ## What I already checked (verified 2026-07-16 — don't repeat these dead ends) Every convenient structured-data route is closed: - **WordPress REST API: unavailable, and WordPress doesn't render these pages anyway.** A WordPress install *does* exist at `/wordpress/` — it serves the blog's images — but the public site is not WordPress-rendered. Evidence: `/wp-json/wp/v2/posts` 404s at the root; `/wordpress/wp-json/wp/v2/posts` 404s; the permalink-independent fallback `/?rest_route=/wp/v2/posts` returns the homepage HTML rather than JSON or a JSON error, meaning nothing WordPress-shaped handles requests at the root; and a sample post page carries no WordPress fingerprints (no generator meta, no `wp-includes` scripts, no `api.w.org` REST link, no `wp-block-*` or `postid-N` classes). Don't burn time looking for a WordPress-shaped door — there isn't one. - **RSS: useless.** `https://actualized.org/rss` is a valid feed but only carries 10 items, description excerpts only — no `pubDate`, no `content:encoded`, no post bodies. - **No sitemap anywhere.** `/sitemap.xml`, `/sitemap_index.xml`, `/wp-sitemap.xml`, and `/sitemap-index.xml` all 404. `robots.txt` contains no `Sitemap:` directive. So HTML scraping is genuinely the only path. Post URLs must be enumerated by walking the pagination at `/insights`, which runs to roughly 274 pages. ## Site structure - Index with numbered pagination at `/insights`, ~274 pages. - Individual posts at `/insights/[slug]`. - **Canonical host is `www.actualized.org`.** Post links on the index are absolute and include the `www.`, though the site also answers without it. Pick one form and normalize, or you'll end up with duplicate records for one post. - Each post has a title, a publication date, body content, and social share buttons (Facebook, Twitter, Pinterest, Email) that are chrome, not content. **The site's age matters.** `robots.txt` disallows `/includes/`, `/Templates/` (capital T), `/flash/`, `/downloads/`, `/wordpress/`, and several hand-listed `.php` files under `/pages/misc/`. Combined with the total absence of WordPress markup on the public pages, this reads as an old hand-built PHP site with a WordPress install sitting in a subdirectory doing something behind the scenes — plausibly the media library and/or an authoring back-office — while custom code renders the front-end. I never established what WordPress's actual role is, and it doesn't matter for this build: you're parsing hand-rolled HTML either way. The implication is what matters — assume post markup varies across eras, and verify that early rather than trusting page 1 to represent page 274. ## Content characteristics - Posts are short and structurally simple: usually a single headline followed by rich text. - YouTube embeds appear throughout, sometimes several per post. These need to survive into the archive and be playable/visible in the app. - **Roughly 30–40% of posts are just an image, and the image is almost always text.** Not a photo or a diagram — an image containing the actual written content of the post. For these, the HTML body is nearly empty and everything I care about is pixels. That's a large minority, not the majority — most posts are ordinary rich text — but it's far too many to treat as an edge case or handle by hand. See "OCR" below; this is core scope. - Images are hosted under `/wordpress/wp-content/uploads/`. ## Constraints - **Respect `robots.txt`, with one decided exception.** `/insights` and `/insights/[slug]` are not disallowed — that's the content, and it's fair game. `/wordpress/` is disallowed, so don't crawl or enumerate it. However, post images live under `/wordpress/wp-content/uploads/`, and OCR requires downloading them. I've decided that's fine: those are assets embedded in pages I'm permitted to read, and fetching them is the same request my browser makes when I read the post normally. Fetch images **only** by following `src` URLs found in posts I've already legitimately retrieved. Never traverse, enumerate, or spider `/wordpress/` itself. - **Be a good citizen.** Personal-use archive of one site I read. Throttle politely; there's no reason to hit the server hard. A slow crawl is fine — run it in the background. - **Fetch each post once.** The initial crawl should be a one-time cost. See "Keeping it current" below. - **Fail loudly, not silently.** A post that doesn't parse should be recorded as a failure with its URL, not written as an empty or half-empty record. I'd rather see a list of 40 problems than discover hollow posts in six months. - Windows. Assume nothing else about the stack — recommend what fits. ## OCR For roughly a third of posts, the image *is* the article. If the archive only stores the HTML body for those, it stores nothing, and they're invisible to search — a third of my archive would be a hole, which defeats the point of the whole app. So: **Size this properly before choosing an approach.** If the archive turns out to be ~2,700 posts, 30–40% means somewhere around 800–1,100 images to OCR — and every one gets re-processed if I later want a better pass. That's enough volume that per-image cost and total runtime are real design inputs, not rounding errors. Tell me what your approach would cost me in both, and if it's a paid service, what the bill looks like for one full pass. - Text extracted from post images must land in the database and be **searchable on equal footing** with regular post text. If I search a phrase that only ever appeared inside an image, that post has to come back. - **Track which text came from OCR.** I need to know whether a given post's text is authored HTML or machine-read pixels — it affects how much I trust it, how it should be presented, and whether it's worth re-processing later. Don't blend the two into one indistinguishable blob. - **Assume the first OCR pass won't be good enough.** OCR engines vary a lot on this kind of content, and I may want to re-run everything with a better one later. Re-processing must not require re-downloading anything from the site. - **Bad OCR must be visible, not silent.** An image that yields zero text, or obvious garbage, should be flagged as a problem I can review — not written to the database as if it were a legitimately empty post. - **These are quote cards, which is good news.** Confirmed image filenames follow the pattern `leo-quote-<slug>-01.png`, e.g. `leo-quote-life-is-not-a-physical-process-01.png`, `leo-quote-my-work-is-10-percent-wrong-01.png`, served from `/wordpress/wp-content/uploads/`. So these are machine-rendered text on a background — clean, high-contrast, probably consistent typography — not photographs or scans. That's the easy case for OCR and should push accuracy expectations up. Verify the consistency holds across the archive's full history before relying on it. - **The filename is a free sanity check.** The slug in each filename is a human-readable rendering of the quote itself. It's not a substitute for OCR — it's truncated and lossy — but it's an independent signal you can compare OCR output against to catch a pass that's silently producing garbage. Worth storing regardless. - Still sample real examples across the archive before committing to an approach. The `-01` suffix hints at multi-image posts or variants; I don't know what that means. ## Keeping it current The archive needs to check for new posts **daily**, unattended, and pull anything it finds into the archive without me doing anything. Constraints that make this harder than it looks: - **The machine is a personal Windows desktop.** It sleeps, it gets shut down, it won't be on at any predictable hour. A schedule that silently skips its window and waits for tomorrow is not acceptable — missed runs need to catch up on the next opportunity. Days may pass between runs, and the archive still needs to be complete afterward. - **The daily check must be cheap.** Re-walking all ~274 pagination pages every day to find the zero-to-two posts that are new is wasteful and rude to the server. The daily path should cost far less than the initial crawl. - **New posts need the full treatment**, including image download and OCR. A daily sync that grabs an image-only post but never OCRs it has archived an empty post and won't tell me. - **Silent failure is the main risk.** This runs unattended, so the realistic bad outcome isn't a crash — it's the site changing its markup, the scraper quietly finding nothing, and me not noticing for months while believing the archive is current. I need to be able to tell at a glance whether the sync is actually working. Surfacing "last successful sync" and "last error" somewhere I'll see matters more than clever recovery logic. - Don't let a failed or interrupted run corrupt the existing archive. A bad sync should leave yesterday's good data intact. ## Things I don't know - Whether I want to be **notified** when new posts arrive, or just have them appear silently in the app next time I open it. Raise this with me — don't assume either way. - Whether the daily check should also catch **edits to existing posts**, or only genuinely new ones. Same — ask me, and tell me what it costs to do both. - The real post count. ~274 pagination pages; posts-per-page unconfirmed. - How many distinct markup variants exist across the site's history. - Whether publication dates are reliably present and parseable on every post. - Whether pagination is stable enough to enumerate cleanly (new posts shift it). - The 30–40% image-post figure is my rough estimate from reading the site, not a count. Confirm it while you're sampling. - Whether the quote-card format is consistent all the way back. I only saw recent ones; the oldest posts may look nothing like them. - What WordPress is actually doing at `/wordpress/`, beyond serving images. I didn't establish this and don't think it affects the build — but if you find a legitimate, robots-respecting structured route into the content that I missed, I'd rather use it than parse HTML. - Whether OCR'd text should be **shown to me as readable text**, or only used behind the scenes to make image posts findable by search. These have very different accuracy bars. Show me sample OCR output before I decide. ## What I want from you first Before writing the full scraper, show me what you actually found: what the markup looks like at both ends of the archive's date range, how many variants you're dealing with, and what your storage plan is. Then build it once we agree.

His work would be much more appealing and people would want to see where it could go if he wasn't so weird. Point is, it's not primarily about the work. The work is not what people are repulsed by, at least not most. You have the whole anti-transhumanist crowd, but that's not most people.

He wrote a grievance post about why people don't like him and got it completely wrong. The main reason people mock him and want to see him fail is because the things I mentioned. Not because the world wants him to die or because he's shattering paradigms. You seem to have some good sense. Do you really buy his reasoning in that post? I'm actually not bothered by him and have enjoyed watching some of his stuff. His experiments are interesting IMO, and I find him benign for the most part, so I don't personally have a problem with him. That said, he's a weird mf.

Interesting. I can see how it could develop like that, thanks. Although I'm not sure I would attach "Godlike" to the capacity, as that seems like a story being added onto it.

People don't dislike him because he's aiming high with longevity. They dislike him because they sense something is seriously off with him. And they're right. He's a grandiose charlatan and a weird as fuck elite.

The best I can do is a still image and the only way to get animated scenes is with conscious effort. On a psychedelic, the scenes are automatic and require no conscious effort. Are you saying you experience fluid-motion scenes that can go on at length without conscious effort, while sober? If so, how do you think this capacity developed?

Right. A lot of intellectuals are also mad they no longer have their monopoly. They spent their whole life reading books, wikipedia, doing research, collecting data and connecting dots in their preferred rigorous way and now someone with the drive could construct the models they worked so hard for with 1/1000th the effort. They hate that shit. Ego losing its specialness and moat.

Do you mean you can conjure up vivid images of anything you want? Or are they more like animated scenes? If so, how long do the animation last? Do they run on forever until you stop it them?

"What is sadism?"

You serious? lol. Maybe I'm mixing you up with someone else but it seems like you were one of the people who was defending Leo's much worse rhetoric than "non-dual advice given at the wrong time".

Yes, good share. That's the core distinction. It seems there's a conflation of the two. "What would make one biased against science?" is a good thread to pull on.

lol

@Inliytened1 To your point about replacement, I do think that'll be part of the process, but the replacements will be closer and closer to reality, not further away from it. My guess is, eventually all religion will be shed and replaced with spirituality.

I have a theory I'm putting together. I'll share more when it's fleshed out. The basic idea is that technology is changing the speed at which humans shed falsehood. For new generations, coerced installation of bad ideas will fail at a much higher rate than ever, because the kids will have an instant confirmation tool in their pockets. It's interesting if we look at history and analyze how false beliefs and bad ideas were shed over time, what made them shed, how resistant they were, what made them persist, etc. It's not enough for bad ideas to be thoroughly up-ended with reason. We all know reason alone is super slow at propagating and we basically have to wait on entire cohorts of people to die off (science advances one funeral at a time). But there are other variables. The main one being "knowledge transfer". This is what AI is. Knowledge transfer is now in hyperdrive. You can look at what the printing press, a much less effective knowledge transfer tool, did to bad ideas. We've only had high-tech knowledge transfer for a very short time. When every single religious claim can be investigated on-demand by new generations, that's going to have massive implications. When I was first shedding Christianity as a teenager, I was the only one I knew doing it and I didn't dare tell people around me that I figured out it was all bullshit. I couldn't discuss it with anyone because it was taboo and blasphemous and people feared I was going to burn in hell, lol. Most believers have questions, and historically they had no good way to seek answers. But this isn't the case for new generations. There's a lot more to it than this but I think that's the broad stroke. The case studies of bad ideas shedding over time is really interesting. 250 might be too optimistic, but I think it's possible and maybe even likely. If not 250, there's no way it'll make it to 500. Just think about how old the US is and much false shit has been shed since its inception - and that was just with the printing press, radio, and TV. Now, add a tool that cranks the transfer mechanism up to 11 (AI), and realize nearly every kid will have access to it. That's not going to bode well for religion.

If you contemplate some complex and coherent framework long enough, it eventually clicks. Careful not to mistake the feeling of the "click" for something other than what it is.

The argument is that your confidence in your estimation was built from a very shallow understanding of what AI is. You're extrapolating from "chatbot". If you have no real-world understanding of agentic coding, you can't begin to reason about AI's value intelligently. About a month or two ago you claimed AI couldn't code this forum, and I'm guessing you still believe this based on things you've said here. And you're just wrong and refusing to update, no matter how many people tell you you're wrong. As an experiment, spin up Fable 5 right now and say "recode actualized.org's forum. No questions. Just get it done." Watch what happens. Then ask "how powerful would this tool be in capable hands?" "Fable 5, build an Excel clone". This is now possible by one human. So you have two choices. Collect proof that I'm right and update, or tell me (a daily claude code user who follows multiple senior SWEs) and everyone else using these tools that none of us know wtf we're talking about.

"I can be wrong" is said from above the error. It's the master looking at his fallibility in the abstract, granting the error is possible. The self stays elevated the whole time. Nothing is lost. Whereas owning a live error where others are right and you are wrong means climbing down off your perch and standing level with the people who corrected you. It means being an ordinary person who got it wrong and got told so before there's any story available that turns the error into wisdom or the correction into something you already actually knew, or the critics into people who missed your larger point. People don't want "I can be wrong". They want you to demonstrate in real-time you're capable of owning your error. That tends to matter to people with integrity. When truth and self-image collide, can you put truth first? Instead of saying "I can be wrong", say "It appears I am wrong and haven't even explored a large part of this space where the actual value is, let me reconsider all this stuff you guys showed me". This is all people want you to demonstrate, because the inability to demonstrate what everyone considers a basic virtue points to something much larger than being right or wrong. But, I've been trying to tell everyone. It's not going to happen bro. It's just how Leo is and there is no path to correction because it's deeply structural and load-bearing.

https://sketchplanations.com/ is badass. Gonna buy the book. Reminds me of Visualize Value by Jack Butcher.

The actual right messages are spreading. https://www.facebook.com/reel/3298375760325691