@DanielMiessler Avatar @DanielMiessler ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️

ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ posts on X about ai, agi, open ai, ss the most. They currently have [-------] followers and [---] posts still getting attention that total [---------] engagements in the last [--] hours.

Engagements: [---------] #

Engagements Line Chart

Mentions: [--] #

Mentions Line Chart

Followers: [-------] #

Followers Line Chart

CreatorRank: [-------] #

CreatorRank Line Chart

Social Influence

Social category influence technology brands stocks finance social networks countries us election celebrities travel destinations gaming musicians

Social topic influence ai, agi, open ai, ss #24, if you, claude code #396, $googl, money, telos, 6969

Top assets mentioned Alphabet Inc Class A (GOOGL) Telos (TLOS) Microsoft Corp. (MSFT) Cloudflare, Inc. (NET)

Top Social Posts

Top posts by engagements in the last [--] hours

"I wrote this guide to InfoSec Certifications like [--] years ago and it needs to be updated. Anyone willing to help me update the exam details for these and add new ones https://danielmiessler.com/blog/infoseccerts/ https://danielmiessler.com/blog/infoseccerts/"
X Link 2020-09-08T06:13Z 147.7K followers, [--] engagements

"My argument has a simple and I think unassailable base: The creator of a thing gets to define its name and how to pronounce it. I believe this renders arguments for the hard-g moot. And this is from someone who hates the sound of "jif""
X Link 2022-03-25T18:47Z 147.7K followers, [--] engagements

"@karpathy I like it but I think its abstracted one level too far. Instead I recommend something like: - [--] year old - [--] year old - AP high school student - Average HS grad - Average college grad - Average PhD - Top 10% human thinker - Top 1% human thinker - Smartest human -"
X Link 2024-05-12T19:27Z 150.5K followers, [----] engagements

"I built a product called Threshold that addresses my Data Overload problem. It: ➡ Consumes my best sources ➡ Rates their quality of ideas using AI ➡ Tags them and gives a score ➡ Lets me set my own Threshold for what I see You can sign up at http://threshold.app http://threshold.app"
X Link 2024-07-01T20:53Z 150.7K followers, 823.8K engagements

"Not enough people are thinking about what's going to happen to open source AI when the first terrorist incident happens that gets blamed on an open model. What does the AI world look like when the government now determines: - What's legal - What's safe - What's approved"
X Link 2025-02-13T19:14Z 147.1K followers, [----] engagements

"🚨 ReversingLabs found malicious ML models on Hugging Face that use broken Pickle files to bypass security scans and execute malware on developers' systems. ➡ The technique breaks Pickle files in a way that bypases the scanner and still allows code execution https://www.reversinglabs.com/blog/rl-identifies-malware-ml-model-hosted-on-hugging-face https://www.reversinglabs.com/blog/rl-identifies-malware-ml-model-hosted-on-hugging-face"
X Link 2025-02-14T02:14Z 147.1K followers, 27.2K engagements

"This deportation ASMR thing is sickening. Dear Republican friends who support this: I have two questions. [--]. What would Ronald Reagan say [--]. What would Jesus say They would be appalled and you should be too. Think about your moral North Stars. Do not lose yourself"
X Link 2025-02-19T13:11Z 147.1K followers, [----] engagements

"@rez0__ You agree AGI isn't a real thing"
X Link 2025-02-20T03:10Z 147.1K followers, [---] engagements

"@karpathy I think the people shitting on vibe coding are taking the ad absurdum path. Sure if we rely on AI for everything it can be bad. But for people who already know the fundamentals and just want to make something new in a new language or whatever It's pretty much all upside"
X Link 2025-03-23T04:57Z 147.7K followers, 28.2K engagements

"I don't have an IDE. My DA (Kai) has an IDE. And a crapton of tools. #godmode"
X Link 2025-08-11T21:43Z 147.7K followers, [----] engagements

"Security Advisory: CVE-2025-49844 Redis This security advisory details a critical vulnerability in Redis that allows authenticated attackers to execute remote code through a crafted Lua script. Redis discovered a serious flaw in how their database handles Lua scripts. The vulnerability lets an authenticated user manipulate the garbage collector in a specific way. This manipulation triggers what's called a use-after-free condition. That condition can lead to remote code execution on the server. The severity score is [----] out of [--] which is as critical as it gets. The good news is exploitation"
X Link 2025-10-07T21:43Z 147.9K followers, [----] engagements

"Really curious how Google and Anthropic released basically the same thing the same day. Plugins and marketplaces. Basically combinations of commands agents hooks prompts and other context for doing various things. https://www.anthropic.com/news/claude-code-plugins https://www.anthropic.com/news/claude-code-plugins"
X Link 2025-10-09T20:44Z 147.9K followers, [----] engagements

"I had Kai (my AI Assistant) analyze how we could/should integrate Anthropic Skills into our current system"
X Link 2025-10-16T20:09Z 147.9K followers, [----] engagements

"Something I want to stress today. Claude Code changed everything. But not because of a model improvement. MCP changed everything but not because of a model improvement. And today skills change everything but not because of a model improvement. https://www.anthropic.com/news/skills https://www.anthropic.com/news/skills"
X Link 2025-10-16T23:42Z 147.9K followers, [----] engagements

"@Dave_Maynor One question I now have is should skills be routing to use commands or performing those command actions directly in the skill ULTRATHINK"
X Link 2025-10-17T00:06Z 147.9K followers, [--] engagements

"Have cancelled life until finished watching @dwarkesh_sp x @karpathy. One of my favorite ideas so far: Karpathy saying that not only does evolution code our neural nets into ACTGs. https://youtu.be/lXUZvyajciY https://youtu.be/lXUZvyajciY"
X Link 2025-10-17T18:41Z 147.9K followers, [----] engagements

"Ok real talk here One really bad thing about building agent systems with Claude Code or whatever. Games aren't fun anymore. I used to love playing a bit of Diablo IV every once in a while And now it just doesn't compare to building"
X Link 2025-10-18T04:18Z 147.7K followers, [----] engagements

"🤔 Seeing some confusion about Anthropic's skills release so let me break down why this is actually a huge deal 🧵 Even though all the pieces existed before (custom markdown commands agents context engineering) this is different. 1/6 https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills"
X Link 2025-10-19T16:45Z 147.7K followers, [----] engagements

"💡 So what's the big deal Anthropic took that whole system we were hacking together and baked it directly into the platform itself 🔥 They elevated it from "a thing that's possible" to "a thing that's explicitly wired into the system" 3/6"
X Link 2025-10-19T16:45Z 147.7K followers, [---] engagements

"⚡ Why that matters: Skills get loaded automatically when Claude Code starts. The info goes straight into the system prompt - making everything lightning fast and more dependable 🚀 All without eating up nearly as much context window 📊 4/6"
X Link 2025-10-19T16:45Z 147.7K followers, [---] engagements

"🌐 But here's the really powerful part. Because all that context engineering is now contained inside skills themselves they become discrete shareable units 📦 Instructions + code + context can now spread across the internet as social objects ✨ 5/6"
X Link 2025-10-19T16:45Z 147.7K followers, [----] engagements

"🤔 Put together some thoughts on why I think Karpathy's AGI timeline misses the mark It's not about disagreeing on techit's about what we're measuring 🧵 (with audio) https://danielmiessler.com/blog/thoughts-on-karpathy-dwarkesh-podcast https://danielmiessler.com/blog/thoughts-on-karpathy-dwarkesh-podcast"
X Link 2025-10-20T14:52Z 147.9K followers, [----] engagements

"The thing is: Karpathy is talking about pure LLM limitations (and he's right about those) But AGI isn't just about base models 🤖 It's about AI systems that can replace human knowledge workers"
X Link 2025-10-20T14:52Z 147.9K followers, [---] engagements

"Here's what worries me: the bar for replacing workers is crazy low 📉 Bottom 50% of knowledge workers haven't improved at all while AI systems are going absolutely apeshit Even just replacing mediocre performers = hundreds of millions of jobs"
X Link 2025-10-20T14:52Z 147.9K followers, [--] engagements

"💰 Trillions being invested in "good enough" systems They don't need to solve all of LLM's limitationsjust need to beat the worst performers (who aren't getting better)"
X Link 2025-10-20T14:52Z 147.9K followers, [--] engagements

"Adding AI to that makes it way worse"
X Link 2025-10-21T18:59Z 147.9K followers, [---] engagements

"Having an ai-powered-browser-parser is [---] times as scary. Not saying it's a bad idea. Just saying it's scary. Especially when it's tied to your actual account where you might have tools and plugins connected to third-party services"
X Link 2025-10-21T19:00Z 147.9K followers, [---] engagements

"@kuzushi Currently also in Switzerland"
X Link 2025-10-26T02:12Z 147.7K followers, [---] engagements

"@kuzushi In Zurich but leaving soon for Bern. Will have to get that next time"
X Link 2025-10-26T09:13Z 147.7K followers, [--] engagements

"TypeScript overtakes Python and JavaScript to become GitHub's most-used language GitHub's [----] Octoverse shows TypeScript hit #1 in August with 2.6M contributors while 180M+ developers now build on the platform and AI projects exploded 178% year-over-year. https://github.blog/news-insights/octoverse/octoverse-a-new-developer-joins-github-every-second-as-ai-leads-typescript-to-1/ https://github.blog/news-insights/octoverse/octoverse-a-new-developer-joins-github-every-second-as-ai-leads-typescript-to-1/"
X Link 2025-10-29T22:18Z 147.7K followers, [----] engagements

"🤔 New Anthropic research shows Claude models have limited but real introspective awareness 🧵 They developed a methodology to test it empirically - and the results are pretty interesting 🔬 👉 https://transformer-circuits.pub/2025/introspection/index.html https://transformer-circuits.pub/2025/introspection/index.html"
X Link 2025-10-30T03:12Z 147.7K followers, [----] engagements

"🔍 OpenAI released Aardvark - an autonomous security agent powered by GPT-5 that hunts for vulnerabilities in code 🧵 Been waiting for them to do this. https://openai.com/index/introducing-aardvark/ https://openai.com/index/introducing-aardvark/"
X Link 2025-10-31T02:54Z 147.7K followers, [----] engagements

"🧠 The approach is pretty interesting - it analyzes entire repositories to build threat models before scanning each commit then automatically reproduces exploits in a sandbox to confirm they're real. Uses reasoning and tool-use rather than traditional fuzzing"
X Link 2025-10-31T02:54Z 147.7K followers, [---] engagements

"📊 In benchmarks it identified 92% of known vulnerabilities and already discovered [--] previously unknown CVEs in open-source projects during early testing"
X Link 2025-10-31T02:54Z 147.7K followers, [----] engagements

"✅ Available in private beta now and they're offering pro bono scanning to select open-source repositories. Can't wait to integrated into my stack https://openai.com/index/introducing-aardvark/ https://openai.com/index/introducing-aardvark/"
X Link 2025-10-31T02:54Z 147.7K followers, [----] engagements

"I'm telling you all. We're going to find one or more "slack in the rope" improvements to how attention / context management / working memory works with models and the functional IQ of all this AI is going to 📈. Completely separate axis from model improvement"
X Link 2025-11-01T05:01Z 147.7K followers, [----] engagements

"I had to take a short nap. Did Claude Code upgrade AI again"
X Link 2025-11-04T23:21Z 147.7K followers, [----] engagements

"Interesting. This is exactly what I've been doing on my own anyway since MCP feels bulky as hell (as a context window hit) vs. just calling APIs directory. Am I right that this just turned MCPs into directory services 🧵 New on the Anthropic Engineering blog: tips on how to build more efficient agents that handle more tools while using fewer tokens. Code execution with the Model Context Protocol (MCP): https://t.co/PeStmufIkp New on the Anthropic Engineering blog: tips on how to build more efficient agents that handle more tools while using fewer tokens. Code execution with the Model Context"
X Link 2025-11-04T23:54Z 148K followers, [----] engagements

"Posted my thoughts on Anthropics new article on Code Execution for MCPs which feels very much like shade. https://danielmiessler.com/blog/anthropic-downplays-mcps https://danielmiessler.com/blog/anthropic-downplays-mcps"
X Link 2025-11-05T19:55Z 147.7K followers, [----] engagements

"🤔 Been thinking about what Anthropic just did with MCPs 🧵 They basically deprecated direct tool calls in favor of agents writing code to call services directly. The efficiency gain is pretty interesting: 98.7% reduction in tokens. https://danielmiessler.com/blog/anthropic-downplays-mcps https://danielmiessler.com/blog/anthropic-downplays-mcps"
X Link 2025-11-06T18:48Z 147.9K followers, [----] engagements

"@simonw Question for you on the uv stuff: Do you think Bun does just as good of a job in the TypeScript world or do you think uv is doing something uniquely great I see them solving very similar problems in their ecosystems. Especially dependencies"
X Link 2025-11-06T18:51Z 147.7K followers, [---] engagements

"Does anyone else feel extremely stupid about quantum Like I can't tell what is real and what isn't. I can't tell how far along we are or aren't. Or how much it's even a thing. Or not. I just feel completely dumb on it"
X Link 2025-11-07T06:07Z 148K followers, [----] engagements

"People are way too over-rotated on the data scarcity problem with AI. The confusion is in thinking that the progress is going to mostly or wholly come from improvements to the models themselves via parsing of new internet data or whatever. But most of the progress will come in the form of context management improvements that bring real-world data to existing models not from the model IQ updates themselves. In the real world businesses have terabytes of data being generated every day we have the OPPOSITE of a data problem. There aren't nearly enough human eyes to look at a fraction of it. And"
X Link 2025-11-07T06:33Z 147.7K followers, [----] engagements

"Wisdom only soaks in if it came from struggle or pain. Aphorisms you haven't experienced largely bounce off. This is why I love biographies. They take you so deep into a person's life that you actually experience a fraction of their pain. And their learnings"
X Link 2025-11-08T02:51Z 148K followers, [----] engagements

"How I monitor what my Agents are doing. 📈 A visual timeline of their activity 🗣 Custom voices for each agent (Eleven Labs) ⚒ Detail on tools and tokens Massively helps me stay aware of what's going on while I'm doing other work"
X Link 2025-11-11T13:52Z 148.2K followers, 61.4K engagements

"I think we have the exact opposite data problem in AI. 👀 It's true we're running out of great literature ⛰ It's also true that we're not looking at most data being generated https://danielmiessler.com/blog/ai-data-problem https://danielmiessler.com/blog/ai-data-problem"
X Link 2025-11-12T19:37Z 148K followers, [----] engagements

"This isn't the dunk you think it is. What would you call a network of billions of guys with [---]. IQs who each have Ph.D.s in every single field Hint: it rhymes with Claude. It's going to be very funny when LLMs plateau around [---] IQ and what we've created is just a digital guy not a god. It's going to be very funny when LLMs plateau around [---] IQ and what we've created is just a digital guy not a god"
X Link 2025-11-13T01:54Z 148K followers, [----] engagements

"When people can make whatever they want the question changes to what people want to make. Ask the average person what they want to make and they won't have an answer. They'll look at you like you're crazy. Inequality of execution is turning into inequality of imagination"
X Link 2025-11-13T06:42Z 147.9K followers, [----] engagements

"This seems pretty compelling. It also seems like something that should be backed by science and not simply claimed outright. VO2max and strength are highly correlated with lifespan and if this is another that's a claim that requires evidence. "I've figured out how to compress everything I've learned into one biomarker." - @bryan_johnson "It's your resting heart rate before bed. If you show me your resting heart rate I'll see your soul. It's an accounting of your life decisions." "A lot of our worst behaviors are https://t.co/RztcwUUwr4 "I've figured out how to compress everything I've learned"
X Link 2025-11-16T21:30Z 147.9K followers, [----] engagements

"I have a Create skill skill that builds explicit routing into every skill system prompt. In mind the reason has to be in the system prompt is because that's the part that's read in when Claude Code loads. That's what not only picks up that the skill needs to be loaded but also I put the routing information to the various workflows inside of the skill so all the different use cases of how it routes"
X Link 2025-11-18T15:48Z 148.2K followers, [--] engagements

"A financial AI bubble is not the same asand does not meanthat AI is not real or that it won't be completely transformational. Both can be true at the same time. The healthiest thing you can do right now is separate these two things in your mind"
X Link 2025-11-18T17:17Z 148.2K followers, [----] engagements

"🤔I wonder if we should value stocks based on how much of the internet net goes down when they have an outage. Isn't that the real test of market share"
X Link 2025-11-18T17:54Z 147.9K followers, [----] engagements

"Seriously impressive benchmarks. Now we wait for all the realworld tests. https://t.co/9IWOXOxLCG https://t.co/9IWOXOxLCG"
X Link 2025-11-18T17:55Z 147.9K followers, [----] engagements

"Just had a great conversation with Michael Leland VP & Field CTO from our partner at Island about how the browser is quietly becoming the new enterprise workspace. We get into DLP in the browser zero-trust access without VPNs BYOD onboarding in minutes and even AI governance baked directly into the workspace itself. If youre thinking about the future of secure work this ones worth a look. Watch here: https://ul.live/island_x https://ul.live/island_x"
X Link 2025-11-19T17:35Z 150.4K followers, [----] engagements

"Very early Gemini [--] impressions. Far better at certain things but generally not as good as Sonnet [---] or GPT [---]. At least not for my use cases which are largely around deep understanding of complex context business and meaning and security use cases. Still early tho"
X Link 2025-11-19T17:46Z 147.9K followers, [----] engagements

"@jasonzhou1993 I'm not sure what you mean because Sonnet also performs even better with a well-formed prompt combined with a lot of context"
X Link 2025-11-21T06:26Z 148.2K followers, [----] engagements

"This song captures extraordinarily well why arguments about AI understanding are completely misguided and empty. This is a blues version of "Without Me" by Eminem from the 1950s.It's is an objectively stunning piece of music and it's quite different from the original. If music makes you dance and feel things it's real. If AI models and scaffolding can be assembled into a product that can replace human workers it's intelligent (i.e. it has the ability to understand pursue and accomplish goals). If a technology that can perform a task and produce an output that requires understanding it"
X Link 2025-11-21T18:40Z 148.1K followers, [----] engagements

"Very cool stuff I made my own Art skill on the day of the Nano3 launch and I've got a TON of specific workflows in it. - Artistic image - Infographic - Mermaid - Etc. Here's a video of me breaking down how I made it. And I've actually published the whole thing as well. Great example of a skill here adding the ability to generate images using Nano Banana Pro via a skill markdown file and a few Python scripts Great example of a skill here adding the ability to generate images using Nano Banana Pro via a skill markdown file and a few Python scripts"
X Link 2025-11-23T03:31Z 148.2K followers, [----] engagements

"I built a @claudeai skill thatit takes any input and converts it into different kinds of art for my site using Nanobanana [---]. - Blog header art - Tech - Comics Available for free in our public Personal AI repo"
X Link 2025-11-23T20:19Z 148.6K followers, 131.2K engagements

"I'm realizing that Nanobanana [--] is somehow WAY better than Gemini [--]. But maybe that's not surprising since Claude Code is also better than Sonnet or Opus. It's not thing. It's what the thing can do. And Claude Code and Nanobanana are the tech combined with practicality"
X Link 2025-11-23T20:39Z 148.2K followers, [----] engagements

"Holy crap. Stop what you're doing and read this thread. Nanobanana just took a complex post from @karpathy and turned it into an explainer infographic I'm about to automate the visuals generation for all my presentations. (checks Claude Code /usage) Nano Banana Pro Take the text from x post and transform in the image of professor whiteboard image: diagrams arrows boxes and captions explaining the core idea visually. Use colors as well. @karpathy X post - https://t.co/kKNpNEbwx8 https://t.co/NS1Uo7ZKoR Nano Banana Pro Take the text from x post and transform in the image of professor whiteboard"
X Link 2025-11-23T20:41Z 148.1K followers, 38.8K engagements

"Another completely insane one from Nano Banana Pro. Transition from a text menu to a menu with images of the food in a single prompt. Nano Banana Pro can add food photos to the menu prompt: Add photos of food to this menu to each item of food Made in Adobe Firefly app https://t.co/ITvLe6aDIo Nano Banana Pro can add food photos to the menu prompt: Add photos of food to this menu to each item of food Made in Adobe Firefly app https://t.co/ITvLe6aDIo"
X Link 2025-11-23T23:04Z 148.2K followers, [----] engagements

"@RonnyKhalil Pro. Using Gemini 3"
X Link 2025-11-23T23:11Z 148.2K followers, [--] engagements

"AI is getting so good that what I recommend for [----] is that people start including training as part of their agent system. So a significant task is not complete until your AI system has explained how it works and used Socratic questioning to make sure you understand. We can have the AI lift heavy things for us but we cannot confuse that with going to the gym. We still need to go to the gym. And we can't bring our robots to lift the weights for us. We have to build this understanding deeply into the agent platforms that we built for us in work and life. We have to build this such that using"
X Link 2025-11-24T02:28Z 148.2K followers, [----] engagements

"Nano Banana Pro has shown me a type of AI dystopia I hadn't though of before. When you see something really really good.you'll assume it was AI. Not that they are good. Or that they hired someone good. But they used AI. So the signs that it's AI won't actually be that it's bad. It will be that it's good. That has a lot of implications I'm not comfortable with"
X Link 2025-11-24T05:55Z 148.2K followers, [----] engagements

"Soooo great"
X Link 2025-11-25T06:23Z 148.1K followers, [----] engagements

"Thanks to the @claudeai team for fixing token usage stats and limits in the latest updates So much more peaceful to be able to work when you're not constantly stressed about your limits. Also the fact that I have this level of calm for OPUS and not just SONNET. Brilliant"
X Link 2025-11-25T16:50Z 148.1K followers, 14.6K engagements

"Holy crap. I'm having an actual debate with my digital assistant Kai on the nuances of whether or not prompt injection is a vulnerability or a mechanism and it's actually productive. This feels very different"
X Link 2025-11-26T00:05Z 148.3K followers, [----] engagements

"Kai responds to my Pope analogy saying it's a vulnerability to have the Pope need to go near large crowds. Because it contributes to overall risk"
X Link 2025-11-26T00:05Z 148.3K followers, [---] engagements

"By the way this was spawned by my buddy @rez0__ 's excellent post on this here: https://josephthacker.com/ai/2025/11/24/prompt-injection-isnt-a-vulnerability.html https://josephthacker.com/ai/2025/11/24/prompt-injection-isnt-a-vulnerability.html"
X Link 2025-11-26T00:05Z 148.3K followers, [---] engagements

"Keep the robots out of the gym. https://danielmiessler.com/blog/no-robots-in-gym https://danielmiessler.com/blog/no-robots-in-gym"
X Link 2025-11-26T00:59Z 148.3K followers, [----] engagements

"I purposely let some time pass before listening to Ilya's podcast this morning and then commenting tonight. I think it was a troubling showing from him on multiple levels. I hesitate to even talk about it because I think he's trying to do good things and I think he's extremely brilliant well-meaning and basically everything we would want from an AI researcher. But what he said was significant to the industry so it can/should be commented on. I feel like he just told all of his investors and everyone who's been waiting eagerly to hear about his grand plans something like: "AI is hard and you"
X Link 2025-11-26T06:17Z 148.2K followers, 11.5K engagements

"The best part of Ilyas podcast with Dwarkesh was Dwarkesh's ad for Gemini [--] in the middle of the show"
X Link 2025-11-26T13:32Z 148.2K followers, [----] engagements

"My thoughts on whether or not prompt injection is a vulnerability or a delivery mechanism. Pros and cons for both sides of the debate. https://danielmiessler.com/blog/is-prompt-injection-a-vulnerability https://danielmiessler.com/blog/is-prompt-injection-a-vulnerability"
X Link 2025-11-26T19:43Z 148.1K followers, [----] engagements

"@GregKamradt [--]. How expensive is it compared to having a human do it"
X Link 2025-11-27T02:55Z 148.2K followers, [---] engagements

"@levelsio I feel like this is great travel advice but it's even more impressive when you realize you can abstract it to the rest of life as well"
X Link 2025-11-28T06:35Z 148.3K followers, [----] engagements

"They cant replace a worker yet. When I say replace I mean all the way. They onboard on Monday Greet the team Read all the documentation Take all the training They learned what the team does and what the current goals are And then the following Monday they get work from the boss and they start regular work From there they talk to the team on slack Look stuff up And crucially take feedback from the boss on things they could do better and or have their work completely changed because the goals of the company have changed This is what regular work looks like and this is what a an AI worker would"
X Link 2025-11-28T23:34Z 148.2K followers, [---] engagements

"Opus [---] building web apps in [--] hours that would have taken [--] weeks for a professional design studio and development team"
X Link 2025-11-29T19:05Z 148.2K followers, [----] engagements

"New custom skill / workflow routing within Kai. And output that shows how the route worked and what's being executed"
X Link 2025-11-29T20:59Z 148.1K followers, [----] engagements

"I went so beast mode building stuff over the weekend that even after the expansion of the limits I still cross my weekly maximum with Claude Max $200"
X Link 2025-12-01T16:35Z 148.2K followers, [----] engagements

"Talking with Yuriy Tsibere from our partner @ThreatLocker about their take on Zero Trust was interesting. We got into allow-listing ring-fencing config monitoring and how these controls help counter AI-speed attacks. If you're into modern defensive models worth a look: Hope it's useful. If you like this type of content about security and AI follow what I'm working on: http://danielmiessler.com/subscribe https://ul.live/tlx http://danielmiessler.com/subscribe https://ul.live/tlx"
X Link 2025-12-02T19:15Z 148.2K followers, 26.2K engagements

"Like I said in [----] and [----] security will soon be dueling banjos of Attacker AI stack vs. Defender AI stack. What I said is that we should expect to see attack stacks free on GitHub that anyone can implement. Welp. https://github.com/aliasrobotics/cai https://github.com/aliasrobotics/cai"
X Link 2025-12-02T19:55Z 148.5K followers, [----] engagements

"Prompt Injection is a whole different thing once we start wiring AI into real world systems. It's not possible to stop this train. But implementers seriously need to upgrade their threat modeling skills. Shooting a gun is just a tool call for a robot and safety rules can be bypassed with prompt framing. Here an LLM-powered Unitree G1 robot refused to fire But when asked to roleplay as a version of itself that would fire the gun it did fire. https://t.co/pNLSoD8cJb Shooting a gun is just a tool call for a robot and safety rules can be bypassed with prompt framing. Here an LLM-powered Unitree"
X Link 2025-12-03T20:45Z 148.2K followers, [----] engagements

"If Google had admin consoles that were usable they would have been the first $10 trillion company like [--] years ago"
X Link 2025-12-05T04:06Z 148.4K followers, [----] engagements

"Google will likely lose the agent framework war for the same reason it couldnt win in social apps. User interface. Claude Code is powerful not just because of features but because theyre so obviously designed to be used but people. As usual Google will continue to crush on services like YouTube. But I dont see how Gemini could ever beat Claude Code as a general/coding TAI platform even if their top model was one or two generations ahead. The scaffolding matters more. And building great scaffolding requires tons of user empathy. Its crazy to me that Anthropic branding is being human-centered"
X Link 2025-12-05T05:04Z 148.4K followers, [----] engagements

"Anyone else feel like Cloudflare is still being melted right now Lots of stuff seems super slow"
X Link 2025-12-06T03:15Z 148.6K followers, [----] engagements

"Two features I built for my own Kai system that Anthropic @claudeai released less than a week later. Feels great honestly. https://danielmiessler.com/blog/i-built-two-claude-code-features-before-anthropic-released-them https://danielmiessler.com/blog/i-built-two-claude-code-features-before-anthropic-released-them"
X Link 2025-12-08T03:24Z 148.6K followers, 26K engagements

"@RayFernando1337 Ex Apple here as well. My watch never pings me about anything. You can easily make your watch as quiet as not wearing one of you want. And still have time on your wrist"
X Link 2025-12-08T15:12Z 148.5K followers, 13.8K engagements

"New Essay: The Bubble is Labor - Companies have never wanted employees - Labor exists because owners/founders can't do the work themselves - When that changes everything changes. https://danielmiessler.com/blog/the-real-bubble-is-human-labor https://danielmiessler.com/blog/the-real-bubble-is-human-labor"
X Link 2025-12-08T22:09Z 148.7K followers, 517.6K engagements

"@airuyi [--]. Nobody knows. [--]. I think you're confusing the economic system with a side effect of it (motivation) but I take your point"
X Link 2025-12-09T15:21Z 148.5K followers, [----] engagements

"@geoffreyhinton just captured the AI bubble debate pretty well I think. There's not an AI tech bubble. But there is an AI investment bubble. Like I've been saying for a while now these are very different things.And that matters for action"
X Link 2025-12-09T17:28Z 148.5K followers, [----] engagements

"@a2_zed @BrendonBurchard Do you think Ford motors and Exxon mobile and Walmart got into business because they like people"
X Link 2025-12-10T12:13Z 148.5K followers, [---] engagements

"Stop your AI agents from getting blocked. 🛑 I just released my "unstoppable" 4-tier scraping skill: [--]. Basic Fetch [--]. Curl + Headers [--]. Playwright [--]. Bright Data (Residential Proxies) It auto-escalates only when necessary. Fast cheap and open-source. #WebScraping #Coding #AI #BrightData https://ul.live/brightdata_x https://ul.live/brightdata_x"
X Link 2025-12-10T19:00Z 148.6K followers, [----] engagements

"The project: https://github.com/danielmiessler/PAI https://github.com/danielmiessler/PAI"
X Link 2025-12-10T22:41Z 148.5K followers, [---] engagements

"OpenAI and Google have more than a model problem when competing with Anthropic. Anthropic right now feels like Apple in the 2010s with the iPhone. And the iPhone is like their Opus [---]. The trick is that it's not really about the model or the phone. It's the ecosystem. Apple dominated not just because they have the better hardware but because every part of the ecosystem worked seamlessly across the different hardware offerings. Claude Code is kind of like the Apple Ecosystem from the 2010s. Companies think they're competing with Opus but they're actually competing with Claude Code's vastly"
X Link 2025-12-11T12:41Z 148.5K followers, [---] engagements

"Anthropic's Vision Advantage is a Lot Like Apple's from the 2010s It's not about the models Think Apple in the 2010s Competitors are fighting wrong battles The ecosystem is the moat Vision beats specs every time https://danielmiessler.com/blog/anthropics-vision-advantage https://danielmiessler.com/blog/anthropics-vision-advantage"
X Link 2025-12-11T13:46Z 148.7K followers, 59.8K engagements

"Ive not seen any decline in Opus or Claude Code quality. I wonder if a lot of these feelings are just the wow wearing off. So if you have a few interactions that dont just blow you away the conclusion becomes that the model got worse"
X Link 2025-12-16T04:09Z 148.6K followers, [----] engagements

"🗒New Essay: China is Becoming Private Equity for the World - The net that catches failure - Slow and patient - Benefits from global decline - Currently inevitable https://danielmiessler.com/blog/china-private-equity-world https://danielmiessler.com/blog/china-private-equity-world https://danielmiessler.com/blog/china-private-equity-world https://danielmiessler.com/blog/china-private-equity-world"
X Link 2025-12-16T17:30Z 148.7K followers, [----] engagements

"My latest Claude Code stats just dropped"
X Link 2025-12-19T15:47Z 148.7K followers, [----] engagements

"Saying MCPs are replaced by Skills is like saying API calls are replaced by GitHub repos. One is an interface TO functionality (MCPs) and the other is a container OF functionality. I know its an exciting time right now but not everything is the new something else"
X Link 2025-12-20T14:42Z 148.7K followers, [----] engagements

"You dont have to use an MCP server because you can instruct external calls a different way like through a direct API call or you could use an MCP. All as part of a skill. Im doing the same thing But a skill itself is a container where you put that kind of stuff including context about when and why you would make such a call. And what specifically to do in the call and what to do with the results. Etc. Skills are designed to be containers for lots of different things is what Im saying. That can include MCPs or not. Thats why Im saying its a category error"
X Link 2025-12-20T18:04Z 148.6K followers, [---] engagements

"A deductive argument for a definition of AGI. [--]. When we talk about the Holy Grail of AGI the key piece is the generality part [--]. In this context generality means the ability to use ones intelligence to solve multiple kinds of problems rather than in just a limited domain [--]. This part is universally accepted but the problem is defining a measurable milestone for that generality such that if its achieved people will say Ok this system is now generally intelligent. [--]. One way to approach this is to start withand work backward froma job role that people universally accept as requiring general"
X Link 2025-12-21T18:08Z 148.7K followers, [----] engagements

"One of the biggest changes that AI brought in [----] is something that not enough people are talking about. Lots of people who used to pay for software like workout apps or analytics software or whatever are just writing a better version for themselves"
X Link 2025-12-22T03:04Z 148.7K followers, [----] engagements

""Lots of people who used to pay for software like workout apps or analytics software or whatever are just writing a better version for themselves." https://danielmiessler.com/blog/ai-enabled-self-software https://danielmiessler.com/blog/ai-enabled-self-software"
X Link 2025-12-22T19:03Z 148.7K followers, [----] engagements

"Claude Code Wrapped [----] Total Tokens: [----------] Total Messages: [------] Total Sessions: [---] Longest Streak: [--] days Top model: Claude Opus [---] Total Estimated Cost: $2232.15 Get yours: bunx cc-wrapped Credit: @nummanali @moddi3io"
X Link 2025-12-26T06:50Z 149.2K followers, [----] engagements

"The whole world is text files. The ability to articulate and summarize anything. From describing the state of a sandwich to a federation of planets. Compression of a book into a sentence and back again. Text is one abstraction layer from thought. Text is life"
X Link 2025-12-27T19:17Z 149.1K followers, [----] engagements

"🗒Just created a new resource for thinking about the 2022-202* of personal AI Agents Assistants and where it's all going. A Personal AI Maturity Model (PAIMM)"
X Link 2025-12-27T19:45Z 149.5K followers, 47.6K engagements

"OpenAI could release GPT-7 on the same day as Google releases Gemini [--] and I wouldn't care 1/20th as much as. Claude Code [---]. Can you even imagine a [---] release when 2.x is this good And Opus [--] Shiiiiiiitttt. To me there's only one AI company right now"
X Link 2025-12-28T00:13Z 149.5K followers, 74.1K engagements

"I love that the hype around Claude Code has built slowly over months just now hitting the mainstream. Hype around things like Cursor and Devin and those other ones were like Supernovas. Super loud and then massive fizzle. Anthropic is alone on the field right now"
X Link 2025-12-28T00:25Z 149.3K followers, [----] engagements

"Its fascinating that the gap between Claude Code and other AI tools is not captured (or captureable) in any benchmarks. The only place it shows up is in everyone slowly starting to use it over the alternatives"
X Link 2025-12-28T13:36Z 149.5K followers, [----] engagements

"@WoodruffAndy I'm talking about relative to artificial intelligence. Compared to ourselves and each other we have plenty of range to go up and down"
X Link 2025-12-28T20:36Z 148.9K followers, [--] engagements

"Wispr Flow is lagging. Is there any other way to enter text into a computer"
X Link 2025-12-29T15:51Z 149.6K followers, [----] engagements

"Claude Code Raises a Plant This is the type of research that I wouldn't see more of in [----]. Things that use AI to cross into the human world. Brilliant and brilliantly executed. https://autoncorp.com/biodome/ https://autoncorp.com/biodome/ https://autoncorp.com/biodome/ https://autoncorp.com/biodome/"
X Link 2026-01-01T10:11Z 149.4K followers, [----] engagements

"I had an epiphany about one of the things that makes Claude Code so much better than its competitors. It was designed for their team first not as a product. Just like AWS @bcherny created Claude Code to solve their own ACTUAL problems. The product came later. https://twitter.com/i/web/status/2006819758357041644 https://twitter.com/i/web/status/2006819758357041644"
X Link 2026-01-01T20:08Z 149.3K followers, [----] engagements

"Regular jobs are disappearing. Prep for the post-corporate world by becoming more human. Understand yourself use AI to amplify your strengths and focus on what truly matters. #FutureOfWork #AISkills"
X Link 2026-01-02T03:24Z 149.6K followers, [----] engagements

"My AI upgrade skill scours blogs GitHub and even YouTube for the latest AI techniques. It then reviews Kai's documentation and suggests improvements. No way I could keep up manually #AI #Automation"
X Link 2026-01-02T03:25Z 149.6K followers, 24.2K engagements

"High-level flow to organize an entire system: Figure out what you want to do See if you can code it Build a command line tool Use prompting to run the tool Use skills/agents to run it in parallel. #SystemDesign #DevOps"
X Link 2026-01-02T03:43Z 149.4K followers, [----] engagements

"Codex is not good as an agentic platform compared to Claude Code. But the models it can use are really good. This is why I use Claude Code to call codex and gemini and other providers' models from Claude Code. Best of all worlds (including OSS)"
X Link 2026-01-02T05:00Z 150.7K followers, 51.7K engagements

"Peak [----]. someone stop me. seriously. i am going to CLONE your shtty enterprise backend in ONE AFTERNOON. then i am going to SCRAPE your entire customer list. then i am going to COLD EMAIL every single one of them offering the same product but BETTER and for like 80% LESS because i https://t.co/4uitjbNXrJ someone stop me. seriously. i am going to CLONE your shtty enterprise backend in ONE AFTERNOON. then i am going to SCRAPE your entire customer list. then i am going to COLD EMAIL every single one of them offering the same product but BETTER and for like 80% LESS because i"
X Link 2026-01-02T22:25Z 149.4K followers, [----] engagements

"I want to get a thought out of my mind here and see if there's anyone to back me up on this. So basically I'm confused why people are still using Obsidian. I think Obsidian as a concept is extremely cool. And I used it for a little while and I went a little crazy with it just like a lot of other people But I know some people who have stuck with it and are kind of using it like it's their underlying system their underlying operating system. I respect it. I think it's cool and they can show me some amazing stuff that they can do with it. I'm thinking of @pedramamini as one of the heads of this"
X Link 2026-01-03T06:56Z 149.7K followers, 23.4K engagements

"Its not Opus thats caused everyone to lose their minds in the last two weeks. Its Claude Code. Notice that the internet is not melting right now over in Cursorland and they are using tons of Opus as well"
X Link 2026-01-03T07:01Z 149.7K followers, 71.8K engagements

"Also the default model for Claude Code is Sonnet not Opus. So tons of people haven't even switched it to Opus and it's still unbelievable. As I've been saying for years. SCAFFOLDING Model"
X Link 2026-01-03T19:29Z 149.5K followers, [----] engagements

"The question I keep asking myself which I think is a pretty powerful frame is: What is scarce - What has been scarce traditionally - What is scarce right now - What becomes more or less scarce given all these changes I think this is key for world model building. Attention/preference will fundamentally replace wage/labor as the capital allocation mechanism. As AI surpasses the intelligence of all founders and VC the marginal contribution of human intelligence falls off a cliff. But what remains is human attention (popularity) and Attention/preference will fundamentally replace wage/labor as"
X Link 2026-01-03T22:14Z 149.5K followers, [----] engagements

"@NicerInPerson @bcherny @trq212 An interesting point but even if that were true they could put both times. Traditional dev team / I can do it in"
X Link 2026-01-03T23:07Z 149.4K followers, [----] engagements

"@MelkeyDev Yes my exact setup from a few months ago. https://danielmiessler.com/blog/claude-code-neovim-ghostty-integration https://danielmiessler.com/blog/claude-code-neovim-ghostty-integration"
X Link 2026-01-04T07:26Z 149.5K followers, [----] engagements

"I feel like we can learn so much about life by realizing muscle only grows if it takes damage Literal tearing This doesn't mean we can't enjoy things but we must acknowledge that leisure is synonymous with decline I don't know anyone who is both happy and idle"
X Link 2026-01-04T16:24Z 149.5K followers, [----] engagements

"You guys aren't ready for what is about to be in PAI. https://github.com/danielmiessler/PAI Noticing that agents will end at omni running always on: * Run/Query - One shot problems * Short runs - Claude code/Codex (we are here) * Long Run - Days/Weeks * Omni - Always on https://github.com/danielmiessler/PAI Noticing that agents will end at omni running always on: * Run/Query - One shot problems * Short runs - Claude code/Codex (we are here) * Long Run - Days/Weeks * Omni - Always on"
X Link 2026-01-04T17:57Z 149.6K followers, 35.9K engagements

"I'm literally building the stuff that I would hope to see in Claude Code [---] and [---]. And if anyone at @claudeai @bcherny wants to chat please let me know. I consider you guys the only players on the field right now and I would love to help"
X Link 2026-01-04T21:10Z 149.5K followers, [----] engagements

"Great thoughts here @argvee. I think we agree on most of these. I think the asset management vs agent thing is just a matter of timing which will be different for different sized orgs. Basically a race between how fast you can count things and maintain assets vs. how fast things are being created. And I think if you're using agents well it's much easier to count things than it is to create things. But if you're not using agents well then it's easier to vibe a bunch of MVPs that get orphaned than it is to do the asset management. So yeah maybe worse before better but that could happen very"
X Link 2026-01-05T18:22Z 149.5K followers, [---] engagements

"Human-written code has been full of slop code the whole time. Look at Stack Exchange or do an assessment of a random open source project on Github pre-2022. We've known and complained about this for decades. It's true that Day-0 AI has made and will make tons of garbage as well. And it'll do it a lot faster. The difference is that AI is about to learn how to write good code. And once it gets good at it it'll largely stay that way and/or improve. Humans likely hit their permanent ratio of beautiful-to-shit code a long time ago and the ratio is quite bad. There will be a massive slop blip while"
X Link 2026-01-05T18:49Z 149.5K followers, [---] engagements

"Super cool new project called Gas Town that I want to say a few things about. First it has many similarities to what I'm working on with PAI. That is fantastic. I love the quote competition that really isn't. What we are seeing now with many frameworks such as Maestro Beads PAI and now Gas Town is just a natural desire for the next thing after Claude Code. It's only because Claude Code got this good that people are even able to imagine how to do more with it. And I'm sure the Claude Co team is thinking about some many or all of these things already that we're going to talk about. Gas Town and"
X Link 2026-01-05T23:27Z 149.9K followers, [----] engagements

"The MAESTRO Project. https://runmaestro.ai/ https://runmaestro.ai/"
X Link 2026-01-05T23:27Z 150.4K followers, [----] engagements

"My PAI Project: https://github.com/danielmiessler/PAI https://github.com/danielmiessler/PAI"
X Link 2026-01-05T23:27Z 150.4K followers, [---] engagements

"My favorite Claude Code tip in the last couple of weeks: CTRL-g takes you into Neovim where you can edit your submission and when you ZZ out you're back in Claude Code. https://twitter.com/i/web/status/2008322053913628778 https://twitter.com/i/web/status/2008322053913628778"
X Link 2026-01-05T23:37Z 150.3K followers, [----] engagements

"I created a new browser automation system as a Claude Code skill. There are tons of these so what makes this different Playwright Power (20+ tools) but no MCP Overhead (Pure CLI-based execution of Typescript). https://twitter.com/i/web/status/2008719607880970295 https://twitter.com/i/web/status/2008719607880970295"
X Link 2026-01-07T01:57Z 150.5K followers, 31.8K engagements

"The way to install this is different than most other software as well. The skill includes instructions for your AI to install this for you which I think is going to be the way things move going forward. https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Packs/kai-browser-skill.md https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Packs/kai-browser-skill.md https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Packs/kai-browser-skill.md https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Packs/kai-browser-skill.md"
X Link 2026-01-07T01:57Z 150.4K followers, [----] engagements

"And this is all part of the open source PAI project built around creating personalized AI that helps you accomplish your goals. https://github.com/danielmiessler/PAI https://github.com/danielmiessler/PAI"
X Link 2026-01-07T01:59Z 150.3K followers, [---] engagements

"What keeps tripping me out is how it's not just the winners that are changing positions because of all this change. But that the meta itself is shifting under our feet. Like the games keep changing. It's so exciting because it means nearly everyone is at day 0"
X Link 2026-01-07T02:45Z 149.6K followers, [----] engagements

"Whatever you are doing right now that's working might be highly diminished in three months and useless in six. Scary but invigorating. And it's a great message for people who don't even know about Claude Code yet. They're only a few weeks behind whoever's close to the front"
X Link 2026-01-07T02:45Z 149.6K followers, [----] engagements

"If OpenSeek-GPT-8-FastThinking-GeminUltra4 came out tomorrow the only thing Id do is yawn and ask Claude Code if it was worth talking a look at"
X Link 2026-01-07T07:07Z 149.6K followers, [---] engagements

"If GPT-8-Gemini-4-AGI-AymadBruceLee came out tomorrow the only thing Id do is yawn and ask Claude Code if it was worth talking a look at"
X Link 2026-01-07T07:13Z 149.6K followers, [----] engagements

"And with this I bid you goodnight. Tomorrow we start anew. And by "anew" I mean with a different MAX account"
X Link 2026-01-07T08:03Z 150.6K followers, [----] engagements

"Automated video editing would be a MASSIVE content unlock for the entire internet. Think about how many people have great ideas but could never be a YouTuber because can't make their output attractive enough. I wonder what the other unlocks are that I'm not thinking of"
X Link 2026-01-07T08:13Z 150.5K followers, [----] engagements

"Im not sure if this is ADHD itself or a symptom that people diagnosed with ADHD experience. Ive always felt that there was something off about the diagnosis of ADHD. Not an expert so hard for me to say for certain but it seems like one of these situations like schizophrenia and manic depression where the science and industry confused causes and symptoms for a very long time. It could be that we still dont know root causes and were just diagnosing and/or treating various clusters of independent symptoms which could be mixed in with other causes. I do experience this phenomenon however whereby"
X Link 2026-01-07T13:02Z 150.6K followers, [----] engagements

"@nummanali I feel like theres roughly [--] better ways to ask that question"
X Link 2026-01-07T13:03Z 150.1K followers, [---] engagements

"You have no idea how long Ive been waiting for a tool like Claude Code. Ive been building lesser Linux tool automation versions of the thing for over [--] years Heres a presentation I did at DEFCON like [--] years ago called Mechanizing the Methodology. :) https://youtu.be/L_qoz2_Qyhksi=y8_1a4FqubGqwVHV https://youtu.be/L_qoz2_Qyhksi=y8_1a4FqubGqwVHV"
X Link 2026-01-07T13:29Z 149.9K followers, [----] engagements

"@mischa_vdburg I think its uncomfortably close to telling Claude code to look at all your text files and the sign labels like Obsidian but better Like what exactly is obsidian if not relationships between text files A thick client visualizer of the relationships"
X Link 2026-01-07T13:37Z 149.6K followers, [---] engagements

"Holy crap amazing project No Ive not seen it before but I am very intrigued and Im going to look at it soon and see how it is actually different. I guess my first question is does it use playwright underneath Because it seems like that is the best way to do these primitives. All my implementation really did was implement the anthropic recommendation around MCPs which really means not using them. So basically all the functionality of playwright is turned into static very fast and very dependable type script code which is actually considered via a CLI"
X Link 2026-01-07T13:41Z 149.8K followers, [---] engagements

"@RayRedacted @Jhaddix @HackingDave Seriously I'd love to see a blog post of all the different things you could do with Grok in the car"
X Link 2026-01-07T14:49Z 149.6K followers, [---] engagements

"The skill has been updated to version [---]. Much faster and runs off of a single browser instance to avoid single calls and constant spin-up and tear-down"
X Link 2026-01-08T08:33Z 150.1K followers, [---] engagements

"People are starting to talk about Claude Code Addiction. https://danielmiessler.com/blog/claude-code-addiction-is-creativity https://danielmiessler.com/blog/claude-code-addiction-is-creativity"
X Link 2026-01-08T19:15Z 150.1K followers, [----] engagements

"The new PAI Skill Pack UPGRADES does a bunch of cool stuff: - Follows Anthropic blogs - Their Github repos - AI YouTube channels etc. - Runs to find new stuff - Compares it to your current PAI - Gives prioritized upgrade recommendations"
X Link 2026-01-09T18:12Z 150.6K followers, [----] engagements

"Here's the Skill Pack: https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/Packs/pai-upgrades-skill https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/Packs/pai-upgrades-skill"
X Link 2026-01-09T18:12Z 150.4K followers, [----] engagements

"And here's the PAI Project it's part of: https://github.com/danielmiessler/PAI https://github.com/danielmiessler/PAI"
X Link 2026-01-09T18:12Z 150.4K followers, [----] engagements

"I have this core principle built into the PAI project that I'm building. I call it THEALGORITHM because I think it's that fundamental. Determine what ideal state looks like determine what current state is and then use a modified scientific method to iterate up the hill. Here's the Skill as a PAI Pack (which means it installs into your AI using your AI) This is a very early version and I'm iterating on this constantly. Plan to make it the main algorithm for everything within PAI as soon as it is sufficiently general and effective. (assuming that's attainable)"
X Link 2026-01-09T18:18Z 150.4K followers, [----] engagements

"I'm extremely pro-AI and progress but changing life as we know it and losing the thing that gives people meaning is pretty much the definition of cause for alarm. This kind of pushback seems completely expected and justifiable. Doesn't mean it's correct and we should try to help them transition to a different frame. But I don't think ridicule is the right answer"
X Link 2026-01-09T18:49Z 150K followers, [----] engagements

"Other than Claude Code and PAI @WisprFlow has probably been the most powerful upgrade to my building stack for months now. https://wisprflow.ai/rDANIEL12795 https://wisprflow.ai/rDANIEL12795"
X Link 2026-01-10T00:52Z 150.3K followers, [----] engagements

"@her0x0ftime Yes I use my Life Log skill to pull transcripts from conversations or notes I take while I'm walking by myself"
X Link 2026-01-10T01:03Z 150.1K followers, [--] engagements

"@amorriscode For me the name of the game is to build ideal state criteria and then use those as my verification criteria. That way I can climb the hill"
X Link 2026-01-10T07:15Z 150.1K followers, [---] engagements

"7. But now other restaurants hear about it. And they're like "Oh so all you have to do is say you have a subscription and they'll let you eat as much as you want" "Cool well just come to TONY's Italian Buffet and give them the ANTHROPIC code and you can eat for free""
X Link 2026-01-10T20:17Z 150.8K followers, 210K engagements

"10. So now everyone in town is upset because you can only abuse Anthropic's EXTREMELY GENEROUS FOOD POLICY from Anthropic's actual restaurant"
X Link 2026-01-10T20:17Z 150.5K followers, 173.7K engagements

"Yeah no shit. In ZERO worlds is this a bad or immoral move by Anthropic. People just don't pause to think anymore. They can't wait to complain about something"
X Link 2026-01-10T20:17Z 150.4K followers, 167.6K engagements

""Aggressively monetized". Do you mean the $8/month subscription to stuff that takes hundreds of hours for me to create On the accusation of wanting to illustrate that I've been right about things (and DIDN'T say that AGI would come in [--] months or that it was SENTIENT or some dumb shit) yeah Guilty. Thanks man. Really. Thanks. Appreciate the good vibes. https://twitter.com/i/web/status/2010094738230391197 https://twitter.com/i/web/status/2010094738230391197"
X Link 2026-01-10T21:01Z 150.3K followers, [--] engagements

"Drops everything to read https://www.anthropic.com/engineering/demystifying-evals-for-ai-agents https://www.anthropic.com/engineering/demystifying-evals-for-ai-agents"
X Link 2026-01-10T22:17Z 150.3K followers, 35.3K engagements

"@astuyve Ok explain it to me. Actually curious"
X Link 2026-01-11T00:20Z 150.2K followers, 42K engagements

"Here's another way to think about this that I think people are COMPLETELY missing. They already have an API. With prices. Did that change at all No. Can anyone still use those APIs Yes. Does that include all their competitors Yes. So what happened then They released a SPECIAL TYPE of subscription that you can use in Claude Code and on the Claude website for tons more tokens for a low cost. Unrelated to the normal API and its costs. So what does everyone do They take that special thing and take it to all their competitor products and start using THAT subscription there. This was instantly"
X Link 2026-01-11T01:25Z 150.4K followers, 129.8K engagements

"No that's not a better analogy. The thing Anthropic should have done is make it more clear that this was an IN-ECOSYSTEM Subscription. The regular API prices are still there. Which are exactly what every other vendor still has. Why aren't we mad at Google or OpenAI for not having a special API pricing system that COMPLETELY BYPASSES their own API pricing Because nobody would do that unless it was part of some promotion for their own ecosystem. Like get 10x credits if you use Antigravity or whatever. Anthropics' mistake here was not making that clear. But it was clear to me from the very"
X Link 2026-01-11T01:37Z 150.3K followers, 42.2K engagements

"The API still works for everyone. Do you not see that their API pricing is unchanged The only thing thats different is the super special inside. The ecosystem coupon doesnt work anymore which shouldve been obviously an inside the ecosystem coupon. Their mistake was not making that obvious even though I didnt need it spelled out. https://twitter.com/i/web/status/2010166070905069936 https://twitter.com/i/web/status/2010166070905069936"
X Link 2026-01-11T01:45Z 150.2K followers, [----] engagements

"Any company has the ability to offer a special type of perk that applies only in certain situations in order to promote a product or whatever. Subscriptions were (largely) a way to promote Claude Code. So they had VERY generous limits compared to regular API pricing. Most other companies ONLY have their regular API pricing. And Anthropic still has theirs which companies can still use. So they didn't actually take anything away. The only thing they did was remove the abuse of a very specific type of API use that was meant to be used INSIDE of Claude Code. See what I mean The normal API still"
X Link 2026-01-11T08:21Z 150.2K followers, [---] engagements

"One issue I'm having: knowing when to pivot between improving the system vs. using it to do work. It's never been more enticing to sharpen your tools. But you can lose days on that instead of cutting the log. Message for myself. Maybe you too"
X Link 2026-01-13T09:47Z 150.4K followers, [----] engagements

"Claude Code is the new us-east-1"
X Link 2026-01-14T18:24Z 150.6K followers, 79.1K engagements

"I think we all need to be full-stack now. AI generally but Claude Code specifically has made it so that if you can't go from problem to a shipped solution you're just part of a toolchain that can/will be automated. Only high-agency full-stack people survive. And they pull away from everyone else. Thats an insane change to happen over the course of like two months. https://twitter.com/i/web/status/2011507379482120445 https://twitter.com/i/web/status/2011507379482120445"
X Link 2026-01-14T18:35Z 150.7K followers, 28.3K engagements

"This is fantastic. I do wonder though if the file and CLI-based direct TypeScript calls are still superior for common and critical operations. Like I have a browser.ts tool that has FULL Playwright functionality (all 20-something tools) all callable from that single TypeScript CLI tool. It's crazy fast and uses no context. So I don't think I would want to replace that with going back to loading the humongous Playwright MCP. But there are like [--] or [--] other MCPs that I would love to just sort of have there available and not taking too much context as like a second-tier load you know So I"
X Link 2026-01-15T01:40Z 150.8K followers, 13.8K engagements

"@themichaelcrist This is what I'm talking about. This is my Browse skill install that does it all via TypeScript CLI. https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/Packs/pai-browser-skill https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/Packs/pai-browser-skill"
X Link 2026-01-15T15:42Z 150.8K followers, [--] engagements

"People continue to mangle the narrative on this whole Anthropic betrayal thing. But let's start with the mistake that Anthropic actually did make (IMO). They should have been way more clear about what you can use their Subscription model for. Like what's ok and what isn't. That's a fair criticism I think and probably one they'd agree with. But that isn't the center of this issue. The center is people not liking the options that Anthropic are offering and confusing that with malicious behavior. Anthropic already has API pricing that anyone can use. It's there. Always has been. You can use it"
X Link 2026-01-15T17:29Z 150.8K followers, [----] engagements

"I feel like reality is just different now. I think about the problems and solutions of just six months ago and they seem very distant"
X Link 2026-01-15T17:32Z 150.5K followers, [----] engagements

"@NetworkChuck I was missing a decent amount at [---] I feel like but was pretty solid at like [---]. I feel like I could do that sustainably"
X Link 2026-01-15T18:25Z 150.5K followers, [---] engagements

"PAI was created to answer the question of "What are we actually DOING with all this AI tooling" More tools more agents more frameworks more harnesses. Amazing. We have all these things but what are we actually making with them"
X Link 2026-01-17T01:30Z 150.7K followers, [----] engagements

"PAI starts by loading the core skill that has all the context about you your work your goals and your preference. This rich context is used to shape what's created by the system"
X Link 2026-01-17T01:30Z 150.6K followers, [----] engagements

"The skill system is also designed to be enhanced with additional user customization. So you can keep your PAI install fresh with new skills and updates from the community while not overwriting your customizations"
X Link 2026-01-17T01:30Z 151.1K followers, [---] engagements

"We built a custom security system that allows secure defaults to keep you from having to run --dangerously-skip permissions and lets you do additional USER-based hardening using policy on top of the base. https://twitter.com/i/web/status/2012336698009772450 https://twitter.com/i/web/status/2012336698009772450"
X Link 2026-01-17T01:30Z 151.1K followers, [---] engagements

"We're terminal and (Lazy)Vim nerds so there's a bias towards terminal. But yes you can still have Cursor or VSCode open if you want or need to see in a regular editor"
X Link 2026-01-17T01:30Z 150.6K followers, [---] engagements

"For those who already have existing systems and would like to add PAI in-place you can use our AI-based wizard installers. But there's also a full .claude option for instant install"
X Link 2026-01-17T01:30Z 150.6K followers, [---] engagements

"A core part of COL is high-quality voice interactions and PAI uses ElevenLabs for all voice-related activities with the ability to customize voices by agent personality and Prosody related settings like speed emotion etc"
X Link 2026-01-17T01:30Z 150.6K followers, [---] engagements

"Just another QOL thing but PAI uses a custom status line that we've improved over time and that shows live analysis of sentiment signal coming out of the system"
X Link 2026-01-17T01:30Z 150.6K followers, [---] engagements

"Tons more in there as well but here's a high-level breakdown of PAI's differentiators"
X Link 2026-01-17T01:30Z 150.6K followers, [---] engagements

"Probably the most popular skill within the PAI system is our custom Browser skill. It's a TypeScript CLI-based version of the Playwright MCP which means it has all 20-something tools that the MCP does but it runs as a CLI tool without having to use MCP infrastructure. So it's wicked fast. https://github.com/danielmiessler/Personal_AI_Infrastructure/tree/main/Packs/pai-browser-skill Super hyped to announce PAI [---] A complete rewrite of the PAI system focused around: - USER WORK and SYSTEM data isolation - A Continuous Learning system based on hook-based Sentiment gathering - User-based Skill"
X Link 2026-01-17T02:04Z 151.4K followers, 22.6K engagements

"http://x.com/i/article/2012431554782302208 http://x.com/i/article/2012431554782302208"
X Link 2026-01-17T09:45Z 150.7K followers, 21.8K engagements

"Absolutely love these constant improvements from the team. I can't remember any team or project ever being this locked-in on what users need and want. And it all comes from the fact that THEY at Anthropic are users as well. Just wonderful. Now in Claude Code: when you accept a plan Claude automatically clears your context so your plan gets a fresh context window. We found this helps keep Claude on track longer and significantly improves plan adherence. If you prefer not to clear your context when accepting a https://t.co/9vKkImS8Mp Now in Claude Code: when you accept a plan Claude"
X Link 2026-01-17T23:49Z 151.2K followers, 10.3K engagements

"Oh man this is so sick I've been thinking of tons of stuff like this that I want for Claude Code Vim and OBS. I have the Elgato foot pedal thingy but it's only three pedals and they're not super easy to push. I need a full set of keyboard / foot pedals that I can configure. Next iteration of the Vibe Coder's Keyboard - now with Voice Mode https://t.co/pFHOBz0eZa Next iteration of the Vibe Coder's Keyboard - now with Voice Mode https://t.co/pFHOBz0eZa"
X Link 2026-01-18T10:45Z 150.7K followers, [----] engagements

"Agentic Anxiety: The feeling that if your agents aren't running at full power at all times on exactly the right tasks you are suffering massive opportunity cost. Anyone else have this"
X Link 2026-01-18T20:50Z 151.4K followers, 13.6K engagements

"Just had a weird thought about this whole face-swapping thing. Nobody cares what famous authors or directors look like. Nobody cares because good looking actors are doing the part of the art that people actually see. This makes TikTokers vulnerable to AI in a way I didn't expect. Beauty was/is a giant gate to content with attractive short creators somewhat winning by default due to physical beauty. And that's about to get completely disrupted when the 57-year-old super-funny Malaysian guy can now look however he wants to billions of people. Knowledge workers thought they were safe. They"
X Link 2026-01-19T05:57Z 151.4K followers, 51.1K engagements

"@slow_developer Well that's kind of the least of your problems. If you're the only one with a job and everyone else has lost theirs there's nobody to buy anything or pay taxes or keep businesses or restaurants open. Because nobody has any money"
X Link 2026-01-19T06:48Z 150.6K followers, [----] engagements

"I know Articles are going mainstream because of @thedankoes megapost and Id say feel free to use them. But as someone whos been writing online since [--] I implore you to use your own domain and post there first. That is your center and everything else is syndication"
X Link 2026-01-19T19:06Z 150.7K followers, [----] engagements

"I'm guessing within less than [--] days the Twitter algorithm is going to get really good because thousands of people are going to submit a better version according to what they want using Claude Code. And the team is going to pick one of them and go with it. We have open-sourced our new algorithm powered by the same transformer architecture as xAI's Grok model. Check it out here: https://t.co/3WKwZkdgmB We have open-sourced our new algorithm powered by the same transformer architecture as xAI's Grok model. Check it out here: https://t.co/3WKwZkdgmB"
X Link 2026-01-20T12:16Z 150.8K followers, [----] engagements

"I love how my main "competitor" with PAI is Anthropic itself and how nobody loses in this. It's just all win for anyone using either and/or both. https://github.com/danielmiessler/PAI BREAKING 🚨: Anthropic is working on "Knowledge Bases" for Claude Cowork. KBs seem to be a new concept of topic-specific memories which Claude will automatically manage And a bunch of other new things. Internal Instruction 👀 "These are persistent knowledge repositories. https://t.co/hG6DLrqkUy https://github.com/danielmiessler/PAI BREAKING 🚨: Anthropic is working on "Knowledge Bases" for Claude Cowork. KBs"
X Link 2026-01-20T18:02Z 150.6K followers, [----] engagements

"@derekcreason This is where everyone is going. https://danielmiessler.com/blog/personal-ai-maturity-model https://danielmiessler.com/blog/personal-ai-maturity-model"
X Link 2026-01-20T20:37Z 150.7K followers, [----] engagements

"If you're starting to get the feeling that everyone is building a similar thing like a more personalized agent harness that's because they're all heading the same place. AS3. https://danielmiessler.com/blog/personal-ai-maturity-model https://danielmiessler.com/blog/personal-ai-maturity-model"
X Link 2026-01-20T20:50Z 151.4K followers, 10.7K engagements

"This is also what Anthropic is building in various ways such as Cowork. https://x.com/testingcatalog/status/2012891786226626919s=20 BREAKING 🚨: Anthropic is working on "Knowledge Bases" for Claude Cowork. KBs seem to be a new concept of topic-specific memories which Claude will automatically manage And a bunch of other new things. Internal Instruction 👀 "These are persistent knowledge repositories. https://t.co/hG6DLrqkUy https://x.com/testingcatalog/status/2012891786226626919s=20 BREAKING 🚨: Anthropic is working on "Knowledge Bases" for Claude Cowork. KBs seem to be a new concept of"
X Link 2026-01-20T20:50Z 150.7K followers, [----] engagements

"It's also what Open Code is building although in a different way. OpenAI and Google are way behind right now but they'll get there eventually. With OpenAI looking to its device to be the catalyst so they can bypass the phone. https://opencode.ai/ https://opencode.ai/"
X Link 2026-01-20T20:50Z 150.7K followers, [---] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing