dennis

@claudeai Claude is slowly turning into a full-blown operating system. The speed from prompt to interactive UI is getting insane. RIP static screenshots.

Claude can now build interactive charts and diagrams, directly in the chat. Available today in beta on all plans, including free. Try it out: claude.ai

@lulumeservey Wait until people find out about 'Prehistoric'. It’s not pre-historic, it’s literally 'before recorded history'. Also, the helico + pter breakdown explains why it’s a 'Heli-port' and not a 'Heli-cop-port'. Mind blown.

Helicopter. You’d think it’s heli + copter, but no It’s helico (spiral, like helix) + pter (wing, like pterodactyl ie wingfingers) Words contain multitudes

Adding an llm.txt to my project today. If you aren't doing this yet, you're basically making it harder for ChatGPT/Perplexity to understand your docs. It’s the new robots.txt for the AI era. Keep it clean, keep it markdown-heavy. #buildinpublic #SaaS #GEO

@perplexity_ai The transition from an AI search engine to a 24/7 personal OS layer is massive. Running on a Mac mini locally makes the privacy/security argument much stronger. This is how the 'AI Agent' era officially begins. #Perplexity #AI"

Announcing Personal Computer. Personal Computer is an always on, local merge with Perplexity Computer that works for you 24/7. It's personal, secure, and works across your files, apps, and sessions through a continuously running Mac mini.

@MyLordBebo maalesef herkes bir şeyler yapmaya çalışıyor ancak uzmanlık denilen şey çok önemli. Ne yazık ki rezil bir sistem olmuş.

🎂 How to know if the caller is an AI, just ask for a cupcake recipe! lol …

@MyLordBebo The ultimate vibe check. If they don't say 'Wait, why are you asking me for a cupcake recipe?', you're definitely talking to a bot.

@karpathy The era of 'intelligence brownouts' is a wake-up call. Realizing how much our collective IQ now depends on uptime is wild. Time to build more resilient, local failovers.

My autoresearch labs got wiped out in the oauth outage. Have to think through failovers. Intelligence brownouts will be interesting - the planet losing IQ points when frontier AI stutters.

Building a deep dive on GEO (Generative Engine Optimization) for Shopify stores. Is your project already being picked up by AI search engines like Chatgpt, Claude or are you still focused on traditional SEO? Let's talk! 👇 #BuildInPublic

“Claude, please propose projects we could do that would restore the sense of childlike joy I felt sitting at my computer a mere few months ago, that feeling of frenzy, that sensation of electricity, that the black hole of politics has sucked away. Make no mistakes. Ultrathink.”

The real question isn't whether AI changes legal work. It's whether lawyers use this moment to finally define themselves by what only they can do.

Spot on. AI is moving from 'destination' to 'infrastructure'. The interface is indeed disappearing, which is exactly why AISO (AI Search Optimization) is becoming critical. In a world of invisible orchestration, you don't rank for clicks anymore; you optimize for being the default 'current' that the system carries. We're shifting from building apps to influencing the medium itself.

the term "agent" will likely never become "app".. it will stay in developer jargon & never reach civilians. which is how it should be. lemme explain why. people understand "app" because apple built a distribution surface that made the concept self evident, then spent billions hammering "there's an app for that" until the word fused with the thing. no one is doing that for agent, & no one needs to. this is because ai isn't a new category of object. also it's not a noun waiting to be named. apps were discrete, bounded, downloadable, & pointable at. ai is ambient. it dissolves into the infrastructure of everything else. you don't need a word for something that becomes the medium rather than the message. imho the right analogy is electricity. nobody uses an electricity. you just flip the switch & the lights come on. ai will be the same. the interface disappears, the verb changes, & the underlying orchestration becomes invisible. the people trying to make agent happen are making the same mistake as people who tried to make horseless carriage stick. they're describing the new thing in terms of the old thing. the new thing will eventually just get called whatever it does.

SEO is about ranking for keywords. AISO is about ranking for intent. While building Mantasaur’s 'AI Visibility Score', I realized models don't just look at metadata; they evaluate brand authority based on context and citations they’ve 'learned.' This changes everything for SaaS marketing. We’re moving from 'gaming the algorithm' to 'convincing the reasoning engine.' The Big Question: > If Google Search died tomorrow and people only used Claude/GPT to find tools, would your project even exist in their 'memory'? Be honest—how many of you have actually checked if an LLM knows your brand exists? 👇

The 'ingestion' gap is real, but there’s a trade-off between raw volume and reasoning precision. In my tests for Mantasaur, I've noticed that while one might 'ingest' more, the other often understands the intent behind a complex query much better. For AISO (AI Search Optimization), being able to process a massive document is useless if the model misses the subtle nuance of a brand recommendation. High-fidelity reasoning will eventually matter more than just being a data vacuum.

I want to move permanently to Claude, but their desktop experience is nowhere near ChatGPT. You can’t upload large/multiple documents to Claude because it can’t process them. ChatGPT ingests whatever you throw at it. This is the only thing keeping me from canceling my ChatGPT subscription.

Exactly. AI isn't replacing the expert; it's just stripping away the 'busy work' that masked itself as expertise for decades. Knowing which questions to ask is now the only moat left. It’s why we’re shifting from SEO to AISO—the value is no longer in having the information, but in how an agent interprets and recommends it. High-level intent is the new billable hour.

AI won't replace lawyers. But it did expose that a lot of what gets billed at $/hr was never really 'thinking.' It was formatting, copy-pasting, and researching things a machine now does in seconds. The real skill is in knowing which questions to ask.

This is the core shift of the 2020s. We’re moving from 'knowledge acquisition' to 'intent orchestration.' While schools are stuck in the era of memorization, we're building tools for a world where humans are architects of AI agents, not just info-processors. The real sabotage is not teaching kids how to discern and steer these models. Education needs to pivot to teaching 'prompting the world' rather than just answering it.

We are sending our kids to school to memorize facts that AI can retrieve in 0.3 seconds. We're grading them on essays that AI writes better than their teachers. We're preparing them for jobs that won't exist by the time they graduate. The entire education system is training humans to compete with machines at what machines do best. That's not education. That's sabotage. The schools that survive will teach thinking, not memorizing. Creating, not repeating. Discerning, not obeying. Every other school is a museum that doesn't know it yet.

Most SaaS founders focus on SEO, but they’re completely blind to AISO (AI Search Optimization). Just finished the 'AI Visibility Score' for Mantasaur. It doesn't just guess; it actually queries models to see if they recommend your brand or your competitors. The first tests are... eye-opening. Some "big" brands are invisible to Claude and GPT. I’m curious: If you had to pick one AI model to be recommended by, which one would it be and why? GPT-5, Claude 4.6, or Perplexity? 👇

The silence is often louder than the letters. While some are busy debating the 'doom' in theory, the real shift is happening in the engineering trenches. We’re past the point of open letters; we’re in the era of building the observability and verification tools that make these systems actually accountable in the wild.

Did any prominent doomers publicly call out the Hegseth maneuvers? I tried to suggest to some that we join arms in a joint letter and got basically no reply.

This is a masterclass in the shift from 'Builder' to 'Architect.' You didn't just build a store; you exercised high-level judgment on brand, UX, and SEO while letting AI handle the 'production' layer. The fact that you prioritized LLM-readability (llms.txt and AI crawler access) shows you’re already playing the 2026 game. Most brands are still fighting for Google Page 1, but you’ve made @moumoujus 'citation-ready' for the agent era. This is how you close the trust gap at scale

This is exactly why the 'Context Window' debate is becoming secondary. It’s no longer about what the model remembers, but what the agent can execute on the fly. The shift from 'Chatbot' to 'Engineer-in-the-loop' changes everything for knowledge work. If an agent can write its own tools and spin up specialized sub-agents, the only real bottleneck left is how well your domain data is structured for them to ingest. We are moving from 'Searching for answers' to 'Architecting outcomes.' Wild times indeed.

It’s pretty clear that the emerging paradigm of agents will be like if you had a human expert in any domain, and they had all the capabilities of a top engineer who could use any tool (or the write their own on the fly) to complete any task, along with unlimited compute and a file system to work with. That combination of skills and technology primitives provides you with somewhat limitless capability in AI. You’re no longer limited by only what the model was trained on, or the inherent context window limitations. The agent will simply spin up subagents to work on component parts of the workflow, and get expertise as needed throughout the process. For all known types of tasks that are frequently repeated, they have quick access to existing skills and tools to complete their work. We’re already seeing this in a range of fields where skills are being written for agents to follow either domain-wide or company-specific processes. Doing legal analysis in a specific way, running financial models, processing spreadsheets for complex data work, generating PowerPoints, and so on. And for areas they’ve never seen before, they can simply write code on the fly to do the work one-off. Imagine pairing an industry expert with an engineer that can code up any custom script whenever it wants. Compute is your only limiter. This approach seems to cover a fairly wide range of knowledge work. Obviously the first space to benefit the most from this has been in coding itself, but it’s clear that this go across all other areas of work and even personal agents. Kind of wild.