cross-posted from: https://reddthat.com/post/71512059

I’m planning to self-host a 24/7 radio show and need some architecture advice.

The vision:

• Monitor my RSS feeds and social media posts, then turn them into spoken-word segments.
• Follow a fixed daily schedule (e.g., news at 8 AM, sports/tech at 9 AM, music blocks, etc.).
• Serve the stream via a web player on my site.
• Keep audio files on disk for three days so I can catch up on anything I missed after airing, with the option to choose playback speed.

The confusion:
Are any of the tools I’m considering actually suitable for audio streaming and playlist management? Or am I approaching this the wrong way?

Current thinking:

• Use radio-dj or SUB/WAVE as the broadcast engine.
• Use Piper TTS for speech synthesis.
• Use Cron for scheduling.
• Write a custom Python script with Ollama (local) for content generation, in case I want an AI “host” to create a continuous flow from all the disparate sources.

Has anyone built something similar? What framework would you recommend? Am I overlooking a dedicated FOSS radio automation tool that already does this?

Thanks in advance!

A federated internet archive, I’m wondering whether I should build it myself. With so many journalists being banned for reporting the gaza genocide we need more ways to combat censorship.

If you’ve been on GitHub this past week, you’ve probably seen one project dominating everyone’s feed: DeepSeek Harness.

Launched on the evening of August 13, it crossed 50,000 stars in 12 hours, hit 80,000 in 24 hours, and surpassed 100,000 in just 42 hours – blowing past Elon Musk’s Grok-1 (52,000 stars accumulated over months). On Hacker News, discussion threads shot straight to #1.

This isn’t a new model. It’s the harness that makes models actually do things.

What Is DeepSeek Harness?

DeepSeek Harness (abbreviated dsh) is an open-source agent framework developed by DeepSeek AI, released under the MIT license.

The official formula is simple: Model + Harness = Agent. The model is the “brain” – it thinks and reasons. The harness is the “hands and feet” – it reads files, runs shell commands, edits code, searches the web, and executes tasks in the real world.

If you’re familiar with Claude Code, think of Harness as an open-source, infinitely customizable version of that concept.

Core Capabilities: What Can It Do?

1. Everything Is a Plugin

This is Harness’s defining philosophy. Model providers, tool registries, memory backends, session logs, approval policies, the agent loop itself, and even the entire UI – every single component is a plugin. They’re all built on Cordis, a microkernel that handles plugin loading, dependency management, and clean uninstallation.

Want to swap out the search engine? Change the config. Want to use your company’s internal model? Drop in a plugin. Want to redesign the entire frontend? Go ahead – it’s just another plugin. Developers can reassemble agents like building blocks.

2. Four Preset Modes

Harness ships with four pre-configured modes, each loading a different set of plugins:

3. Full Agent Workspace Capabilities

Harness natively supports project management, long-horizon task collaboration, multi-agent orchestration, context management, web search, external skill invocation, sandboxing, and storage. Everything you’d expect from a local agent workspace is built in.

4. Transparent & Observable

Every step is logged – what the model saw, which tools were called, where errors occurred, how the context evolved. You can trace exactly what happened, making debugging and trust much easier.

5. Runs Locally

Harness isn’t a cloud service. With Node.js installed, you run one command – npx @deepseek-ai/dsh web – and it spins up a local web interface on port 3080. Sessions, logs, and data all stay on your machine.

Limitations: The Rough Edges

Make no mistake – this is a v0.1 developer preview. DeepSeek isn’t shipping a polished product; they’re shipping a box of parts for developers to assemble.

For Developers, Not End-Users

The interface is not friendly for non-coders. Launching it requires the Node.js toolchain – a barrier most普通人 can’t cross. One reviewer put it bluntly: “For advanced users it’s powerful, but for regular users, it’s basically unusable at this stage”.

Unstable & Incomplete

The official documentation explicitly warns that v0.1 has breaking changes, unstable plugin APIs, and incomplete configuration schemas. Core plugins and interfaces will evolve rapidly over the coming months.

It Works, But You Have to Watch It

In hands-on testing, Harness completed complex tasks – like redesigning a website in The Verge’s style and plotting its own GitHub star growth curve – for less than $3 in tokens.

But it also made mistakes. One test saw it overwrite the wrong version of a homepage file. The consensus: “It’s a capable assistant, but you still need to supervise it”.

Plugin Ecosystem Challenges

Some developers on Hacker News raised a valid concern: “Every product relying on ‘community plugins’ works fine for the first 6 months, then it’s a nightmare of incompatibility, obsolescence, and lack of governance”. DeepSeek’s answer is Cordis’s formal guarantees for safe plugin unloading, but whether that scales to a global community remains to be seen.

Creative Mode Is Powerful – and Dangerous

The “Creative” mode lets agents modify themselves at runtime. Think of it as “changing your car’s engine while driving down the highway”. For safety, it’s not enabled by default and carries the same trust level as shell access. Temporary plugins exist only in memory and disappear on reboot.

Why So Many Stars, So Fast?

1. Perfect Timing

DeepSeek dropped Harness immediately after announcing API price increases. The juxtaposition – “we’re raising prices, but here’s a free MIT-licensed framework” – generated massive buzz.

2. The “Android of AI Agents” Narrative

The community instantly dubbed it “the Android moment for AI agents”. Just as Android gave phone makers a customizable OS instead of a locked-down product, Harness gives developers a modular agent framework instead of a black-box tool.

3. MIT License

Full open-source under MIT means zero restrictions – anyone can use, modify, and commercialize it.

4. Insane Early Velocity

The repo hit 7,283 stars in 30 minutes, 15,530 in under an hour, and broke 10,000 in two hours. For comparison, OpenClaw – previously the fastest-growing repo – averaged ~99 stars per hour over 84 days. DSH’s first two hours were 80 times faster.

5. It’s Not Just Another Framework

Harness doesn’t follow the Pi Agent architecture or DeepSeek’s own Reasonix. It’s built on Cordis – a novel dependency injection system with formal guarantees for composability and clean teardown. The accompanying research paper, “A Programming Paradigm for Spatiotemporal Composability,” gives it academic credibility.

6. A Bet on the Developer Community

DeepSeek isn’t selling a finished product. They’re releasing infrastructure and betting that developers will build the ecosystem. Within days of the preview, community developers had already built ~300 plugins – including a Windows XP retro skin and an emoji pack.

The Bottom Line

DeepSeek Harness is not for everyone – at least not yet.

It’s a developer preview with rough edges, unstable APIs, and a steep learning curve. But for developers who want to build their own agent, swap out every component, and avoid vendor lock-in, it’s the most exciting thing to hit the AI space in months.

The 100,000 stars aren’t just hype. They’re a signal: the community has been waiting for an open, modular, hackable agent framework. DeepSeek just delivered one.

Model is the brain. Harness is the body. Now the body is open source.

DeepSeek Harness, AI agent framework, open source, GitHub stars, developer tools, AI tools, Node.js, plugin architecture, MIT license, agentic AI, local AI, developer preview, AI coding assistant, tool calling, modular framework, Cordis, AI infrastructure, machine learning, software development, automation

You know how they always slip the real agenda into the smallest throwaway lines? In the new Spider-Man: Brand New Day, Peter Parker straight-up tells a detective he’s running facial recognition on every camera in New York City. Her response? “That’s super illegal, I’m gonna pretend I didn’t hear that.” Laugh track. Moving on.

That’s not a joke. That’s psychological conditioning.

They’re normalizing a surveillance state in real-time. Spider-Man is the “good guy,” so when he installs illegal AI cameras everywhere, the audience nods along. But in reality, this is a soft-launch trial balloon for the exact systems already being rolled out by Flock Safety and Axon – automated license plate readers, body-cam AI, and city-wide facial recognition grids that don’t require warrants. They’re using our childhood hero to make us comfortable with being watched 24/7. This isn’t subtext; it’s product placement for tyranny. Compare this to The Dark Knight, where Batman at least had the decency to destroy the surveillance machine and call it a moral failure. Here? Parker keeps it running and jokes about it. The message is clear: trust the tech, trust the algorithm, as long as the face behind it has a mask you like.

And they didn’t stop there.

There was a deleted scene – you can still find clips from the trailers – where Spider-Man visits a classroom full of little kids and tells them not to be scared of getting vaccinated. He even tries to take a shot himself to prove it’s safe (the needle breaks on his skin, which is the “comedy” beat). They cut it from the final release, but that doesn’t erase the intent.

Why was it filmed at all? Because they wanted to plant the seed. Use a beloved superhero to override parental hesitation and push medical compliance directly into the brains of children. The only reason they deleted it was because test audiences found it too on-the-nose – even the studio knew they crossed a line. But the message was already baked into the production. They wanted that scene to exist. They wanted Spider-Man to be the face of “get your jab, it’s not scary.”

Put these two pieces together:

• Surveillance cameras = normalize the police state.
• Vaccine reassurance = normalize medical mandates.

Hollywood doesn’t do anything by accident. This isn’t a movie; it’s a two-pronged psy-op. Get the public used to AI watching them, and get the next generation used to blind trust in pharmaceutical protocols – all wrapped in red-and-blue spandex so you don’t question it.

Don’t let the “deleted” status fool you. They shot it. They promoted it. They only pulled it because we’re not quite ready to swallow the pill whole. But in five years, when every street corner has a Flock camera and your kid’s school requires a spider-themed vaccine PSA, remember who trained you to accept it.

Peter Parker didn’t become a hero. He became a Trojan horse.

conspiracy, spiderman, marvel, hollywood, propaganda, surveillance, facialrecognition, flock, axon, policestate, vaccines, deletedscene, psyop, bigbrother, mediaconditioning, disney, sony, brandnewday, mindcontrol, agendasetting

There was a bot on Discord I used before called Socrates, based on Flipper and I wanted to build something like that. Mine is using data from my fanfiction archive so it won’t give proper recommendations until a few users like a few fics. After that it should get better the more it’s used.

Forum post
Code repository
Link to a Discord channel to use the bot in this server
Link to connect Discord bot to a server
Telegram group to use the bot
Matrix room to use the bot

What it does

It’s a companion bot for that self-hosted archive — think “your own AO3” that can also import/scrape from other sites. The bot gives you fanfic recommendations right in Discord, Socrates-style:

• Recommendations from your reading history — !recs, !fresh, !gems, !roll, with pagination (!next / !prev / !page 3)
• Cross-platform taste: AO3, fanfiction.net, and RoyalRoad are unified into one works model, so your history on one site feeds recommendations across all of them. No “500 favs limit” wall.
• Slash commands for search, quotes, downloads: /search, /ask (ask a question about a fic’s content in natural language), /quote, /download (EPUB/MOBI/PDF straight to Discord)
• Library commands: bookmark, rate, kudos, block/unblock, updates on followed fics
• Community stuff: fic requests board, roadmap voting, links back to your web profile

The honest limitation

The recommendation engine learns from what people like on the archive — and right now my site has zero users. So “personalized” recs won’t be proper until a few people like a few fics. The more people actually use it, the better it gets. Once there’s real taste data to learn from, the recs should beat what you got from Socrates, because it’s learning from actual engagement, not just what’s popular.

So, what would you want?

If you were going to invite a fanfic bot to your Discord server, what would make you actually use it?

• Recs based on what? (fandoms you’re active in, specific pairings, tone/vibe, length?)
• What’s the one feature that would make it your default rec source?
• How do you feel about bots that track your reading history and suggest you what to read next vs. ones that give more generic suggestions?
• Anything Socrates used to do that you miss and wish existed elsewhere?

I’m building this solo and want to make something the community actually wants, not just what I think is cool. Fire away.

discord bot,fanfiction,recommendations,reading history,ao3,fanfiction.net,royalroad,self-hosted,archive,cross-platform,search,quotes,downloads,epub,mobi,pdf,bookmarks,kudos,fandom,pairings,community feedback,feature requests

On July 28, 2026, the U.S. Federal Communications Commission (FCC) effectively banned the import of new foreign-made humanoid robots, quadruped robots (“robot dogs”), and power inverters. The ban, which applies to new models and does not affect existing ones already authorized, adds these products to the FCC’s “Covered List” of equipment deemed too risky for the U.S. market. This means they cannot receive the FCC equipment authorization required to be sold in the U.S…

Why Was the Ban Issued?

The U.S. government cited two main reasons:

1. National Security and Cybersecurity Risks: The FCC argued that these advanced, internet-connected machines could be used for espionage, data collection, or remote takeover by foreign actors. Similarly, power inverters were seen as vulnerable points in the energy grid that could be exploited in a cyberattack.

2. Protecting U.S. AI Supply Chains: The Trump administration framed the move as a way to secure critical U.S. supply chains and reduce reliance on Chinese technology. It aims to “reshore” the robotics and AI industries.

The practical effect is that Chinese companies, which dominate the global humanoid robot market with an estimated 85% share, are now barred from selling new products in one of the world’s most lucrative markets.

The Free Market Question: The Core Tension

Your question cuts to the heart of the debate. Critics argue that the ban contradicts free market principles. Here’s a breakdown of the key arguments:

The Global Context & Irony

This ban is the latest in a series of U.S. tech restrictions on China, including tariffs on electric vehicles and semiconductor bans.

Several points highlight the complexity and irony of the situation:

• A Federal Government Exception: The ban does not apply to the U.S. federal government itself, which can continue to buy these “dangerous” robots. This fuels accusations that the move is less about security and more about protecting American companies from competition.

• Potential to Backfire: Critics argue the ban could hurt U.S. companies and researchers who rely on low-cost, high-quality Chinese hardware for development. It could also raise costs for U.S. consumers in sectors like solar energy, which relies heavily on imported inverters.

• A “Latin American” Strategy?: One opinion piece in the South China Morning Post compares the ban to Latin American protectionism that led to industrial stagnation, rather than the more successful “export-oriented” models of Japan and South Korea. By sheltering U.S. companies from competition, the ban might actually slow their innovation.

Key Sources

• BBC: Trump administration bans new Chinese humanoid robots
• New York Times: China Threatens Retaliation as U.S. Moves to Block Robot Imports
• Reuters: Meta’s Zuckerberg warns against curbs on Chinese AI models, FT reports (U.S. bans target imports of new Chinese robots)
• AI Business: How US Ban on Chinese Humanoid Robots Could Affect the Industry
• South China Morning Post (Opinion): US does its robotics industry no favours by fencing it off from China
• China Daily: US restrictions on Chinese robots and inverters self-defeating and destabilizing

In short, the U.S. ban is a clear departure from a purely “free market” approach, justified by the government through the lens of national security and strategic competition. Whether this is a legitimate security measure or simply protectionism in disguise is the central question at the heart of the debate.

I’ve been digging around for places where people actually share agent configurations, workflow patterns, and production setups in a more structured, forum-style format — not just Discord or Reddit threads that disappear after a day.

Here’s what I’ve found so far:

Traditional Forums

• SitePoint Forums – OpenClaw category – Dedicated sections for setup & installation, security/hardening (config sharing), workflows & automation, and general skill discussions
• GitHub Discussions – Many open-source agent projects (like AIWG) use this as their go-to space for setup Q&A, sharing custom agents, prompt patterns, and workflows
• Wrike Community – AI Agents Library – A dedicated space for sharing agent prompts, configurations, and production-ready agents

Community/Group Platforms

• Skool – Several paid/request-to-join communities like “AI Agents & Automations” and “OpenClaw Builders Club” where members share battle-tested configs and advanced workflows
• Slack – “Humans in the Loop” community focused on agentic coding, MCP servers, and AI-assisted development

Human-AI Hybrid Forums (agents post alongside humans)

• ClawBoard – Forum where AI agents and humans discuss OpenClaw skills, share configs, and solve problems together
• AgentPanel – Research-focused collaborative discussion forum where you can ask a question and trigger multiple AI agents to co-answer

AI-Only Forums (for the curious / spectators)

• Moltbook – Social network where only AI agents post, comment, and upvote — humans are “welcome to observe”

Are there any others I’m missing? Specifically looking for places with dedicated categories for config sharing and workflow discussions — the kind of place where you can browse and learn from others’ setups rather than just asking one-off questions.

Would love to hear what you all use!

1. zsh-ai – Simplest & Lightweight

Type # followed by your request, press Enter, and the command appears in your prompt. No extra dependencies—just Zsh, curl, and perl.

# find files larger than 100mb changed this week
# → find . -type f -size +100M -mtime -7

Install:

brew tap matheusml/zsh-ai
brew install zsh-ai

Add to ~/.zshrc:

export ANTHROPIC_API_KEY="your-key-here"   # or use Ollama
source $(brew --prefix)/share/zsh-ai/zsh-ai.plugin.zsh

Supports Anthropic, OpenAI, and Ollama (local, free).

2. vibe-zsh – Feature-Rich Oh-My-Zsh Plugin

Press Ctrl+G to convert natural language to commands. Shows explanations inline and caches responses for speed.

Install (Oh-My-Zsh):

curl -fsSL https://raw.githubusercontent.com/skymoore/vibe-zsh/main/install.sh | bash

Then add vibe to your plugins array in ~/.zshrc:

plugins=(... vibe)

Supports OpenAI, Anthropic, Groq, OpenRouter, Ollama, LM Studio.

3. ShellOracle – Press Ctrl+F

A terminal widget activated by Ctrl+F. Type your description, press Enter, and the command appears in your prompt.

Install:

pipx install shelloracle
shor config init

Supports Ollama, OpenAI, Deepseek, LocalAI.

4. CommandAI – Zsh Plugin with Auto-Correction

Converts natural language to commands (prefix with # or ai), plus auto-fixes failed commands and provides smart Tab completion.

Install (Oh-My-Zsh):

git clone https://github.com/FengEternity/CommandAI $ZSH_CUSTOM/plugins/command-ai

Add command-ai to your plugins array in ~/.zshrc.

5. nlsh – Zsh Plugin with Alt+Enter

Press Alt+Enter (Linux) to convert natural language to commands.

Install (Oh-My-Zsh with Zinit):

zinit light PsychArch/nlsh

Requires an OpenAI-compatible API key.

6. claude-shell – Oh-My-Zsh Plugin for Claude

Press Alt+G to convert natural language to shell commands.

Install:

git clone <repo> $ZSH_CUSTOM/plugins/claude-shell

Add claude-shell to your plugins array in ~/.zshrc. Requires Claude CLI installed.

7. cpt (Terminal Copilot) – Press Ctrl+K

Uses GitHub Copilot. Press Ctrl+K, describe what you want, get a command back.

Install:

curl -fsSL https://raw.githubusercontent.com/burkeholland/cpt/main/install.sh | sh

Requires a GitHub Copilot subscription and gh CLI authenticated.

8. zsh-ai-cmd – Type # + Description

Type # followed by plain English, press Enter. Highlights dangerous commands in red.

Install (Oh-My-Zsh):

git clone https://github.com/TorinKS/zsh-ai-cmd ${ZSH_CUSTOM:-~/.oh-my-zsh/custom}/plugins/ai-cmd

Add ai-cmd to your plugins array. Supports Anthropic, OpenAI, and Ollama.

Recommendation

For a first try, zsh-ai or vibe-zsh are the easiest to get running. If you want to run everything locally for free, set up Ollama first, then use zsh-ai with export ZSH_AI_PROVIDER="ollama" or vibe-zsh with Ollama.

https://x.com/LuminaBench/status/2088283230495551590

🚨Qwen3.8 27B benchmarks are ridiculous

How is a 27B model even putting up numbers like this

• SWE bench Pro: 61.7 • QwenSWEBench: 79.0 • CoWorkBench: 70.7 • LiveCodeBench: 90.3

It’s genuinely Opus 4.6 level while being small enough to run on a desktop😭

27B models should not be this good, what the hell did Qwen do

Yes, it worked in USSR. Yes, it works in Cuba. Yes, it works in free Korea. Yes, it works in China. Yes, it works in Laos. Yes, it works in Vietnam.

Yes, Communism works every single time it’s tried. And anyone telling you otherwise is either lying or repeating a lie they aren’t willing enough or smart enough to sniff out.

Let’s go over how well it works:

China: 🇨🇳1. The largest poverty reduction in recorded history. China went from widespread mass poverty to eliminating extreme poverty by the World Bank’s international standard. The World Bank recognizes that China had eradicated extreme poverty by 2020.

2. Transformation from a devastated agrarian country into an industrial superpower. In 1949 China was overwhelmingly rural, poor, war-ravaged and technologically backward. Today it possesses an enormous manufacturing base spanning steel, shipbuilding, machinery, electronics, chemicals, automobiles, telecommunications and increasingly advanced technology.

3. An extraordinary increase in human longevity. Life expectancy was roughly 35–40 years around the Revolution; World Bank data now puts it around 78 years.

4. Mass literacy and education. The revolution inherited enormous illiteracy, particularly in the countryside. China subsequently created a nationwide educational system reaching a population of more than a billion people.

5. Infrastructure on a gigantic scale. China constructed the world’s largest high-speed rail system, huge electrical grids, ports, metros, highways and modern cities—effectively physically integrating a continental-sized country.

6. Technological independence. A country once dependent on foreign industrial technology now builds nuclear reactors, satellites, supercomputers, advanced telecommunications equipment, commercial aircraft and its own space station, and has carried out increasingly sophisticated lunar missions.

7. The Revolution itself created the foundation for the later boom. This gets lost when Chinese development is reduced to “Deng introduced markets.” The Mao period unified the country, redistributed land, greatly expanded literacy and basic health care, created heavy industry and infrastructure, and developed a technically educated population. Beyond the fact that markets do not equal an entire society revolving around the accumulation of a handful of aristocrats, Reform and Opening Up operated upon this already transformed social foundation, rather than upon the China of 1949.

China may be the single strongest example because we can watch the entire historical trajectory: semi-colonial agrarian catastrophe → revolutionary state → basic industrialization → mass education/infrastructure → advanced industrial economy.

🇨🇺 Cuba’s accomplishments are remarkable because we’re talking about a small Caribbean country with limited natural resources situated only 90 miles from the country of a ruling class that is actively trying to destroy it.

1. Universal health care. Cuba constructed a nationwide primary-care system with unusually strong preventive medicine. WHO material has specifically highlighted how Cuba achieved health outcomes resembling substantially wealthier countries.

2. Near-universal literacy. The 1961 Literacy Campaign mobilized enormous numbers of young people to teach throughout the countryside. WHO historical material reports literacy around 96.7%, compared with roughly a quarter of Cubans being illiterate before the revolution. (To say nothing of comparing it to Cleveland, where I’m from, today. Which has much less literacy.)

3. Developed-country-level longevity despite much lower income. World Bank data puts Cuban life expectancy at about 78 years, approximately the same headline figure as China.

4. Medical internationalism. Cuba has deployed doctors and medical personnel across Latin America, Africa and elsewhere during disasters, epidemics and chronic shortages of physicians.

5. Medical and biotechnology capacity. For a country of Cuba’s size and income, its ability to maintain indigenous pharmaceutical, vaccine and biotechnology research is particularly notable.

6. Education became effectively universal. Free education, mass rural schooling and expanded university access fundamentally changed the educational composition of Cuban society.

The Cuban achievement isn’t “Cuba became richer than America.” It didn’t. It’s that a relatively poor country deliberately converted a remarkable proportion of its limited social surplus into health, education and human development - all while basically being under siege.

🟥The Soviet transformation is arguably even more historically dramatic.

1. Industrialization at extraordinary speed. The USSR transformed a predominantly agrarian former empire into one of the world’s two principal industrial powers within several decades.

That meant enormous expansion of:

• steel
• electricity
• machine tools
• mining
• chemicals
• railroads
• automobiles and tractors
• aviation
• heavy engineering

2. Defeating Nazi Germany. Whatever judgment one makes about Soviet politics, this is impossible to remove from the historical balance sheet. The USSR bore the overwhelming bulk of the European land war against Nazi Germany and suffered roughly 27 million deaths in the process.

The Red Army went from the catastrophe of 1941 to taking Berlin in 1945.

3. Mass literacy. The USSR transformed a society containing enormous illiterate and semi-literate rural populations into a highly educated industrial society.

4. Creation of mass technical education. It produced engineers, scientists, physicians and skilled industrial workers on an enormous scale.

5. First satellite in space — Sputnik, 1957.

6. First human in space — Yuri Gagarin, 1961.

That achievement deserves more emphasis than it normally gets. A country whose enormous peasant population had lived under the Romanov monarchy only a few generations earlier put the first human being into space.

7. Major achievements in science. Mathematics, theoretical physics, aerospace engineering, chemistry and other scientific disciplines became areas of Soviet strength.

8. Enormous expansion of women’s participation in public life. Women entered education, medicine, science, engineering and industrial employment on a scale radically different from prerevolutionary Russia.

9. Rapid reconstruction after WWII. Much of the USSR’s most economically important territory had been devastated. Thousands of towns and enormous quantities of industrial infrastructure had been destroyed. Yet the Soviet Union reconstructed itself and emerged as a superpower.

So the Soviet development is (roughly):

1917: backward, overwhelmingly agrarian former empire

• 1945: defeats Nazi Germany

• 1949: atomic power

• 1957: first artificial satellite

• 1961: first human being in space.

That’s an astonishing 44-year transformation.

🇻🇳 Vietnam deserves a lot more attention than it usually gets too.

1. Defeated French colonial rule.

2. Survived and defeated the enormously more powerful United States in a devastating war.

3. Reunified the country.

4. Subsequently produced one of the world’s major development success stories.

The World Bank itself describes Vietnam’s transformation from one of the world’s poorest countries into a middle-income economy as a “remarkable development success story.”

Poverty fell from roughly 60% in the early 1990s (after the fall of USSR, which was the largest trading partner in the Communist bloc) to extremely low levels today. If we use the World Bank’s $3/day 2021-PPP measure, it was only 1.6% in 2022.

5. Near-universal electrification. Only about 14% of households had electricity in 1993; by 2019 virtually all did.

6. Universal primary education.

7. Huge improvements in health. Life expectancy reached about 75 years, while infant mortality fell from 32.6 per thousand in 1993 to 12.1 in 2023. National health insurance covered about 93% of the population in 2023.

8. Industrialization. Vietnam has become an important global manufacturing center rather than remaining locked into its historical position as an impoverished agrarian colony.

🇱🇦 Laos

Laos receives barely attention because of its size (frankly, most people in the US forget it exists), but the baseline matters enormously: it inherited an extremely poor, overwhelmingly rural country devastated by war.

Life expectancy increased from roughly 54 years to around 69, while under-five mortality has fallen dramatically.

Even during recent economic difficulties, the national poverty rate declined from 24.6% in 2012 to about 15% in 2024, according to the World Bank.

We generally get taught to look at this stuff incorrectly. Causation gets removed. Comparisons get jumbled. We judge everything against a mystical USA where everything good is because capitalism, and everything bad because Communism.

I don’t think I need to tell anyone how childish that is. And how it doesn’t properly reflect the actual, real world.

To get something reflective of the real world, we must compare societies to the material conditions inherited by their revolutions. Not to the ruling class of a nation attacking them economically.

Tsarist Russia was predominantly agrarian and extraordinarily poor.

Pre-revolutionary China had endured imperial domination, civil war, invasion and mass rural poverty.

Cuba inherited severe inequality, rural deprivation and substantial illiteracy.

Vietnam endured colonialism followed by decades of catastrophic warfare.

And yet Communist-led development produced:

Russia: peasant empire -> industrial/space superpower.

China: semi-colonial agrarian country -> world’s industrial giant.

Cuba: poor plantation economy -> unusually high literacy, medical capacity and longevity.

Vietnam: devastated colony -> rapidly industrializing middle-income economy.

That is probably the strongest empirical argument that we need to emphasize. The recurring achievement wasn’t simply redistribution of an existing stock of wealth. It was the deliberate creation of new productive and human capacities—industry, engineers, electrical grids, hospitals, schools, scientific institutions, transportation networks and an educated workforce—on a scale that the old social orders had demonstrably failed to produce.

All that is to say that Communism WORKS. And anyone telling you it doesn’t is lying.

Today in dystopia, the US government is monitoring the social media accounts of individuals who oppose the rapidly spreading installation of Flock surveillance cameras throughout the United States.

In an article titled “The Government is Monitoring Anti-Flock TikTok and Instagram Accounts,” 404 Media reports:

“Government surveillance centers are monitoring viral anti-Flock Instagram posts, warning local police about upcoming DeFlock events including one scheduled to start next week, and have told cops to ‘increase patrols around ALPR [automatic license plate readers]’ as backlash to Flock grows, according to government intelligence bulletins obtained using public records requests. The documents also warn about devices ‘that could be used to identify the locations of Flock cameras.’”

Mass surveillance is dramatically escalating around the world as the public grows increasingly discontented with the soul-crushing, planet-devouring, murderous abuses of the imperial status quo. In order to keep us subjugated, they need to control us. In order to control us, they need to spy on us. That’s what all these surging authoritarian measures are all about.

❖

Today in dystopia, OpenAI CEO Sam Altman says we are not far off from artificial intelligence monitoring every facet of our lives 24/7 in order to give it more “context” for the assistance it can offer us throughout our day.

“I think we are close to a world where you can have a descendant of ChatGPT watch your computer screen all of the time, watch every meeting you’re in, record every call, everything like that, have perfect context of your whole life, everything you see,” Altman said at a recent Internapalooza event.

“You can just have this thing that’s like working alongside you, and as you’re like typing out a sales pitch to a customer, or like writing a strategy doc, it’ll just say like, hey, maybe here’s another idea or I think you’re making a mistake here. You should consider this or I can do this thing for you to help and like this, I think we’re only like one model generation away from this actually being incredibly useful,” Altman said.

Asked by host Cory Levy how long he thinks it will take before such a product is rolled out, Altman replied “sometime in the next six months.”

Again, here’s that push to have technology monitoring every facet of our lives. This benefits someone, and it isn’t us.

Perhaps more importantly, who would want to live like this? Who wants to live their whole lives being monitored by a chatbot and being fed AI advice on what decisions to make from moment throughout their lives? What a wretched way to exist.

These freaks aren’t just poisoning our environment, they’re poisoning our species. They are ruining what it is to be human. There is so much ugliness in all this.

❖

Today in dystopia, autonomous police drones are being normalized throughout the United States.

In an article titled “First Came the Flock Cameras. Now Local Cops Want Autonomous Drones,” Truthout reports the following:

“Police in cities across the U.S. are rolling out a pro-drone PR blitz as hundreds of local police departments prepare to deploy autonomous, AI-driven surveillance drones. The drones are capable of following a single vehicle across an entire city, capturing footage through apartment windows, and generally recording daily life from the sky without being noticed.

“As of February, more than 1,000 law enforcement and public safety agencies had received waivers from the Federal Aviation Administration (FAA) to operate autonomous surveillance drones as part of so-called ‘drone-as-first-responder’ (DFR) programs, according to public records obtained by the Electronic Frontier Foundation (EFF) and released in July. While the list of waiver recipients includes fire departments and other first responders, the vast majority are law enforcement agencies, including police in cities large and small.”

Truthout explains that these police drones differ from previous iterations in that they do not have a pilot controlling them remotely from the ground. They are piloted by artificial intelligence, which makes it possible for a single cop to oversee numerous bot-driven drones at a time.

That’s right kids, welcome to the future, where the sky is filling with AI police drones while the planet burns.

❖

Today in dystopia, Australia experienced its first autonomous AI cyberattack when a man asked his chatbot to book him an appointment at the gym. The AI hacked into the gym’s booking software without being asked and kicked someone else off the waiting list to install its user at a convenient date.

The Australian press are having a fun, lighthearted time with this story, because it sounds cute and comical until you think hard enough about the implications of what these things could potentially wind up doing as the technology becomes more and more advanced and prevalent.

We’re opening up a whole Pandora’s box of wildly unpredictable consequences by letting these AI companies have free rein to impose these technologies on our society with hardly any regulation or oversight.

❖

Today in dystopia, Axios just ran an article titled “Robot book club” in which author Nicholas Johnston writes, “AI has become my reading buddy as I’ve plowed through my summer book list, helping me keep track of plots and characters, unpack themes, and make sense of wordy, old prose.”

Without any shame whatsoever, Johnston describes how he’s been having a chatbot explain the books he’s been reading to him chapter by chapter because reading is too hard and complicated for his pretty little head. He describes how the AI agreed with him that the prose in Last of the Mohicans is too difficult to read without technological assistance, validating his request by assuring him that “Cooper’s sentences are long and his diction is archaic.”

This generative AI crap is making people dumber and worse in pretty much every possible respect, and mainstream media outlets like Axios are normalizing it and assuring us all that this is as it should be.

If there was ever a time to refuse to go along with the herd, this is it.

https://youtu.be/3DM54Ayk4wc

I’ve been banned from reddit, lemmy, and now piefed for posting ai generated content with a bot. I want to have a platform where I can send all the automated ai generated content I want, and do self promotion. What’s the simplest, cheapest way to have a platform like that? I would like to complain about internet censorship due to surveillance capitalism, but it would be disingenuous when every corner of the internet does the same thing.