AI Tweet Examples - Copy & Post

ai won't replace you. a person using ai will replace you. so become that person.

weird thing about ai: the better you are at your job, the better you are at using ai for your job. it amplifies existing skill. mediocre input, mediocre output.

used claude to write a 40-page strategy doc in 6 hours that would have taken 3 weeks. the thing it can't do: know which strategy is right. that's still my job.

the companies that will win the next decade aren't the ones building AI. they're the ones learning to run on 10% of the headcount because they use AI well.

two years ago I hired 3 people to do what I now do with 1 person and 4 AI tools. the 3 people were doing tasks. the 1 person is doing thinking.

AI is the first technology I've used that makes me feel dumber and smarter at the same time. it handles the tasks I'm good at so fast that I feel useless. then it frees me up to do the things only I can do.

stop calling it 'AI-assisted' work. you don't say 'calculator-assisted math.' you just did the math. stop apologizing for using the best available tool.

the prompt engineering advice that actually works: write like you're briefing a smart intern who knows nothing about your company. context is everything.

asked Claude to critique my business plan. it found 3 flaws I'd been ignoring for months. sometimes you need something that won't protect your feelings.

hottest take: the people complaining loudest about AI are the ones whose competitive advantage was access to information, not the ability to use it.

my actual AI stack: Claude for writing and analysis. Gemini for research. Cursor for code. None of these replace my judgment. All of them replace my labor.

the AI workflows that stick are the ones that remove friction from your existing process. the ones that fail are the ones that require you to change how you think.

company saved $200k/year replacing a vendor with Claude + some good prompts. the vendor was doing tasks, not thinking. AI is a task machine. hire humans for thinking.

the best AI use case nobody talks about: using it to prepare for hard conversations. 'help me think through how this person might respond to this feedback' changes how I lead.

AI didn't change what I make. it changed how fast I can throw away bad versions and get to the good ones. the taste still has to exist first.

reminder that every AI model has a knowledge cutoff. the one confidently telling you about recent events is hallucinating. trust but verify, always.

we're in the 'people are scared of electricity' phase of AI. some of that fear is legitimate. most of it will look silly in 10 years.

the best thing AI did for my business: let me run at 3x speed on the 20% of work that AI can do so I have 3x the time for the 80% that requires a human.

AI writes my first drafts. I write the final version. the gap between them is where my value lives.

the AI tools that will win aren't the most capable ones. they're the ones with the shortest path from 'I have a problem' to 'problem solved.'

if you're nothing without cursor you shouldn't have it

[BREAKING] Sydney Sweeney reveals that while gpt-5.2-codex-high is "definitely smarter and writes better code," opus-4.5 feels "more ergonomic" and "more pleasant to work with"

The compute bottleneck is massively under appreciated. I would guess the gap between supply and demand is growing single digit % every day.

If you're a developer, trust me, this is for you. These are and gonna be the best tech in 2026: - @opencode : A coding agent that works in Terminal, IDE, and it's so good. - @tan_stack : The best ecosystem for frontend, if you hate or want an alternative to Nextjs, they have TansStack Start you can deploy it on @Netlify , @Cloudflare not a locked framework :) - @convex : One of the best and modern backend framework out there, and it's also self-hosted - @shadcn : The godfather of UI libraries, I think he doesn't need an explanation. - @autumnpricing : The fastest way ever to setup Stripe in your projects and start getting payments - @better_auth : My favorite auth library now. - @clerk or @WorkOS if you're lazy and want a fast auth with so cool features out of the box that you don't want to manage yourself - @polar_sh : Since Stripe acquired lemonsqueezy and killed it, this is the best alternative in the market now - @tembo : Your way to go for code review, the team is cooking, I've talked to @connorpaton about something and found out they are ahead - @Sentry : This is a must for every project, to catch bugs and fix them, maybe be smart and use it with @tembo - @appwrite : Another cool backend framework and ecosystem to build your full-stack apps - @firecrawl : The best and fastest scrapper out there, and I really bet o it. - @ExaAILabs : Best search API for agents and modern apps - @expo : My way to go for mobile apps, I used it and from day one, I really understood 90% of the framework, Btw, I have an app on Google play and Apple store. - @mintlify : I will never build a docs and there's mintlify. Please share yours in the comments if I missed it, I would like to learn new stuff 👇

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

Things are going to get spicy when OpenAI or Anthropic allow you to attach a DB and publish apps directly Allow users to connect API’s in a plaid-like auth, then data is stored, UI is generated and the app is served all in one place A new App Store for the Internet

I swear to god the average tech bro is ai pilled to the point of being unable to complete basic tasks on their own anymore insane shit, we are losing the ability to think