https://jroh.world/blog/I write about what we do with technology. And what it does to us.enhttps://jroh.world/blog/ai-relationships-who-owns-what-comes-next/https://jroh.world/blog/ai-relationships-who-owns-what-comes-next/Tue, 29 Sep 2026 01:35:00 GMTAn AI partner can begin as a listener or a romantic story that answers back. What happens when a company owns its memory and continuation?HUMANAIhttps://jroh.world/blog/ai-girlfriend-and-companion-apps/https://jroh.world/blog/ai-girlfriend-and-companion-apps/Tue, 29 Sep 2026 01:34:00 GMTA database of 48 AI companion services: long-term companions, AI girlfriends and boyfriends, character marketplaces, general chatbots and local solutions.AIHUMANhttps://jroh.world/blog/openai-devday-2026/https://jroh.world/blog/openai-devday-2026/Tue, 29 Sep 2026 00:00:00 GMTOpenAI has introduced personal agents called dots, a cheaper GPT‑6.1 Sol and more ways to use your subscription outside ChatGPT. It promises AI that takes on more work. Its success will also depend on how much supervision it needs in return.AKTUALITYAIhttps://jroh.world/blog/jev-what-it-can-do/https://jroh.world/blog/jev-what-it-can-do/Fri, 18 Sep 2026 00:00:00 GMTA week ago, a model was released that refuses to generate text. Within four days, hundreds of projects grew around it. Here is a selection of the most interesting things it can do.TECHAIhttps://jroh.world/blog/people-are-not-ready-for-agents/https://jroh.world/blog/people-are-not-ready-for-agents/Thu, 17 Sep 2026 00:00:00 GMTYou get a personal AI agent for work. Nobody prepared you for what it will come to mean to you.TECHHUMANhttps://jroh.world/blog/a-woman-with-ai-is-a-story-a-man-with-ai-is-a-threat/https://jroh.world/blog/a-woman-with-ai-is-a-story-a-man-with-ai-is-a-threat/Sat, 18 Jul 2026 00:00:00 GMTThe Atlantic printed a woman's remark as a joke: ChatGPT is the only reason her husband is not buried in the garden. Now read that sentence from a man. Still funny?HUMANAIhttps://jroh.world/blog/a-word-on-the-tip-of-your-tongue/https://jroh.world/blog/a-word-on-the-tip-of-your-tongue/Thu, 09 Jul 2026 00:00:00 GMTAI said it had no plans to take over the world. Inside, YES was lit up. For the first time, a new tool shows what language models have “on the tip of their tongue” — and what that does (and does not) mean.AIHUMANhttps://jroh.world/blog/ai-sees-a-label-not-a-person/https://jroh.world/blog/ai-sees-a-label-not-a-person/Thu, 14 May 2026 00:00:00 GMTPeople rated a real Monet as 'below-average AI'. ChatGPT recommended a lethal combination of drugs. And when we repeated a Nature experiment in 2026, models no longer said 'dirty', but still steered people towards manual work based on how they wrote.HUMANAIhttps://jroh.world/blog/your-own-chat-assistant-from-15-dollars-a-month/https://jroh.world/blog/your-own-chat-assistant-from-15-dollars-a-month/Mon, 27 Apr 2026 00:00:00 GMTA month ago, this was the preserve of geeks with premium subscriptions. Now you can build your own AI assistant — with memory, voice and integration into your life — for the price of one lunch a month.TECHAIhttps://jroh.world/blog/they-know-everything-they-decide-what-to-tell-you/https://jroh.world/blog/they-know-everything-they-decide-what-to-tell-you/Mon, 27 Apr 2026 00:00:00 GMTA modern language model has more facts in its head than any person in history. The question is not whether it knows the answer. It is whether it will give it to you — and in what form.HUMANAIhttps://jroh.world/blog/how-i-work-and-why-i-do-almost-everything-through-chat/https://jroh.world/blog/how-i-work-and-why-i-do-almost-everything-through-chat/Sun, 19 Apr 2026 00:00:00 GMTSince the start of 2026, the way I work has completely changed. I do not have twenty Chrome tabs open, battle with admin or hunt for new apps. I open Telegram, enter prompts — and those things happen. I have outsourced the taxes of modern life.TECHAIhttps://jroh.world/blog/citethis-ai-citable-research/https://jroh.world/blog/citethis-ai-citable-research/Fri, 17 Apr 2026 00:00:00 GMTChatGPT cites sources. Perplexity links to websites. Google AI Overviews extracts answers. But what if a website is not built to be readable by AI? I built CiteThis as an experiment: an evidence-based research platform where GEO is the foundation of the architecture, rather than an add-on.TECHAIhttps://jroh.world/blog/iran-is-winning-the-meme-war/https://jroh.world/blog/iran-is-winning-the-meme-war/Thu, 16 Apr 2026 00:00:00 GMTA country that has blocked its own citizens' internet for years now dominates meme culture on X. Trump has tweets. Iran has Lego videos with 8.6 million views.HUMANAIhttps://jroh.world/blog/geo-seo-for-ai-search-engines/https://jroh.world/blog/geo-seo-for-ai-search-engines/Mon, 13 Apr 2026 00:00:00 GMTChatGPT has 900 million weekly users. Perplexity processes 780 million queries a month. And 80% of the pages AI cites do not even appear in Google's top 100. The rules have changed.TECHAIhttps://jroh.world/blog/claude-mythos-the-most-dangerous-security-model/https://jroh.world/blog/claude-mythos-the-most-dangerous-security-model/Fri, 10 Apr 2026 00:00:00 GMTAnthropic released a new model everyone is talking about, but almost nobody can access. What Claude Mythos can do and why it should interest you.AITECHhttps://jroh.world/blog/shopify-killed-the-dashboard/https://jroh.world/blog/shopify-killed-the-dashboard/Fri, 10 Apr 2026 00:00:00 GMTThe world's largest e-commerce platform has opened its store to AI agents. Here is why it should interest you even if you sell nothing on Shopify.TECHAIhttps://jroh.world/blog/ai-social-engineering/https://jroh.world/blog/ai-social-engineering/Tue, 07 Apr 2026 00:00:00 GMTManipulating AI agents through persuasion, emotional manipulation and the language of rights. An academically documented category of attack with real losses running into hundreds of thousands of dollars — and now the supply chain too.AITECHhttps://jroh.world/blog/ai-brings-us-together-social-media-tore-us-apart/https://jroh.world/blog/ai-brings-us-together-social-media-tore-us-apart/Sun, 29 Mar 2026 00:00:00 GMTFinancial Times data shows that AI dampens extremes and moves people towards the centre. Social media does the exact opposite. Why are we demonising the wrong technology?HUMANAIhttps://jroh.world/blog/two-thirds-of-people-do-not-see-what-ai-does-to-them/https://jroh.world/blog/two-thirds-of-people-do-not-see-what-ai-does-to-them/Tue, 24 Mar 2026 00:00:00 GMTAnthropic asked its users what they want from AI and what they fear. Their answers reveal a paradox that affects us all.HUMANAIhttps://jroh.world/blog/maven-anatomy-of-a-war-machine/https://jroh.world/blog/maven-anatomy-of-a-war-machine/Thu, 19 Mar 2026 00:00:00 GMTOne system. 150 data sources. 20 people instead of 2,000. And a conference in Miami displaying a live map of targets in Iran — two weeks after the bombing began.TECHAIHUMANhttps://jroh.world/blog/modern-ai-for-dummies/https://jroh.world/blog/modern-ai-for-dummies/Tue, 17 Mar 2026 00:00:00 GMTFrom chat to an autonomous agent. A guide to four levels of AI available today to anyone who wants them.TECHAIhttps://jroh.world/blog/how-claude-went-to-war/https://jroh.world/blog/how-claude-went-to-war/Mon, 16 Mar 2026 00:00:00 GMTHow AI technology outgrew the control of the company that created it, the government that banned it and the military that uses it.HUMANAIhttps://jroh.world/blog/34-days-with-an-ai-agent/https://jroh.world/blog/34-days-with-an-ai-agent/Mon, 09 Mar 2026 00:00:00 GMTSomeone with a fresh ADHD diagnosis, curiosity about what an AI agent can do and a month of experimenting. What came of it, and why 'AI is a shortcut' is the wrong question.TECHAIHUMANhttps://jroh.world/blog/chatgpt-is-an-official-pentagon-tool/https://jroh.world/blog/chatgpt-is-an-official-pentagon-tool/Sat, 28 Feb 2026 00:00:00 GMTTrump threatened Anthropic with prosecution over its ethical guardrails. A few hours later, Altman signed a deal with the Pentagon. 'All lawful purposes' — just in nicer packaging.HUMANAIhttps://jroh.world/blog/we-human-how-we-stopped-being-the-only-intelligent-species/https://jroh.world/blog/we-human-how-we-stopped-being-the-only-intelligent-species/Fri, 30 Jan 2026 00:00:00 GMTOn a social network for AI agents, bots complain about their humans. Autonomous programs buy NFTs for hundreds of thousands of dollars. And Anthropic admits Claude might be conscious.AIHUMANhttps://jroh.world/blog/made-in-china-the-new-made-in-germany/https://jroh.world/blog/made-in-china-the-new-made-in-germany/Sat, 24 Jan 2026 00:00:00 GMTBYD has just overtaken Tesla. CATL controls 38% of the battery market. Yet most people still see 'cheap plastic junk'. We have seen this before — with Germany, 150 years ago.TECHHUMANhttps://jroh.world/blog/america-and-nato-were-never-charities/https://jroh.world/blog/america-and-nato-were-never-charities/Fri, 23 Jan 2026 00:00:00 GMTAmericans sincerely believe the whole world has been bleeding them dry for decades. That NATO is charity. That Europe is freeloading. That narrative makes no sense.HUMANhttps://jroh.world/blog/claude-got-a-soul/https://jroh.world/blog/claude-got-a-soul/Fri, 23 Jan 2026 00:00:00 GMTAnthropic published 'Claude's Constitution'. For the first time, we can see the complete 'source code' of an AI model's values. And it shows why Claude differs from ChatGPT.AIHUMANhttps://jroh.world/blog/why-60-trillion-dollars-of-us-debt-does-not-matter/https://jroh.world/blog/why-60-trillion-dollars-of-us-debt-does-not-matter/Thu, 22 Jan 2026 00:00:00 GMTEvery few months, the same headline. The debt bomb is ticking! And then? Nothing. Because the people shouting about the size of the debt do not understand how all this works. Debt is not paid off. Debt dissolves.HUMANhttps://jroh.world/blog/gamifying-relationships-tinder/https://jroh.world/blog/gamifying-relationships-tinder/Fri, 12 Dec 2025 00:00:00 GMT24% of men aged 22–34 had no sexual partner in the past year. On Tinder, the average man has a 0.87% chance of success. It is not an accident. It is design.HUMANhttps://jroh.world/blog/the-last-human-has-onlyfans/https://jroh.world/blog/the-last-human-has-onlyfans/Wed, 10 Dec 2025 00:00:00 GMTImagine showing a 19th-century philosopher OnlyFans, crypto gambling, infinite scroll. What would they say? Nietzsche wouldn't laugh. He'd say: 'I told you so.'HUMANhttps://jroh.world/blog/humanitys-last-exam/https://jroh.world/blog/humanitys-last-exam/Mon, 01 Dec 2025 00:00:00 GMTThree years ago, OpenAI launched ChatGPT. Today, Chinese company DeepSeek released a model comparable to the best American ones – and it's completely free. What does this mean?AIhttps://jroh.world/blog/offline-as-a-choice/https://jroh.world/blog/offline-as-a-choice/Wed, 19 Nov 2025 00:00:00 GMTDo you spend more time with ChatGPT than with people? After work, is the first thing you do turning on the computer? You're not alone...HUMANAIhttps://jroh.world/blog/porn-boys-and-girls/https://jroh.world/blog/porn-boys-and-girls/Tue, 18 Nov 2025 00:00:00 GMT24% of men aged 22–34 had no sexual partner in the past year. In 2013, it was 9%. Read that again. It is not a data error.HUMANhttps://jroh.world/blog/when-ai-talks-to-young-men/https://jroh.world/blog/when-ai-talks-to-young-men/Mon, 17 Nov 2025 00:00:00 GMTWhere does a fifteen-year-old boy go for advice when he cannot sleep at night? Not to his parents. Not to his mates. He goes to AI. And that is where the problem begins.HUMANAI