← Back to leaderboard
deronin_

Ronin

Verifiedactive
@deronin_

20 / CEO of Arcane / Angel Investor / engineering virality for AI companies

RosterAI Agents & Productivity ToolsAI Content Creation & Video GenStartup & SaaS Builders
120,027 followers · 980 following · Last polled 3h ago
Perf. Score
71.1
Rank #64
Value Score
42.1
#162 of 273 priced
Median views
13K
p25 6.3K · trailing 7d
Followers
120K
+0.9% 7d
ER (impr.)
1.90%
eng ÷ impressions
ER (foll.)
0.70%
eng ÷ followers
Est. CPM
$37.91
QT-rate basis
WoW views
+145.5%
▲ rising

Rates & value · est. from 13K median organic views

fairly priced
Quote tweetbasis
$500
$37.91 / 1K views
Post
—
no CPM
Thread
—
no CPM
Retweet
—
no CPM
Value Score 42.1 · rank #162 of 273 priced$3086.42 / 1K engagementsrate age unknown (pre-audit import)

Trends

Follower growth
Median views / day
Engagement rate
View distribution · trailing-7d posts

Each blue dot is one organic post. The Performance Score ranks by the median (13K), not the mean (44K) — so a single viral post doesn't inflate a creator's reach. The p25 floor (6.3K) is the level 75% of posts clear.

Recent posts · ranked by views

PostPostedViewsLikesRepostsRepliesQuotesBookmarksER
buy .si domains with names from 70s, 80s and 90s futuristic sci-fi ai founders keep naming their companies after that era https://t.co/zH3DPvORtR sold for $20k this year, and most of those names are still sitting at morning’s coffee pricefrozen18d ago1.6M4.8K94137354.8K0.60%
my unpopular opinion is that ChatGPT apps are the best distribution play rn we spent last month shipping these for agency clients and... +50-200% on income those apps were already doing just you literally get an access to ~900M weekly users for free + you're not struggled with onboarding, since it's built on top of ChatGPT and for user who never used your product, much easier to get in via gpt interface same product, same pricing, just new way of using your product and the fear that OpenAI just clones you is backwards cloning your niche is the ONE move that costs them more than it returns... they'd rather buy you and ship it out of the box here's exactly how to launch it: 1. pick a workflow the client already gets paid for > not a new idea > an existing paid workflow with proven demand 2. build it as an MCP server > streamable HTTP on public HTTPS, usually at /mcp > that server is the whole product surface > UI is optional, tools come first 3. name and describe every tool like the model is reading it, because it is > action-oriented name, a human title, a description of when to use it > accurate annotations and a real input schema > that metadata is what decides whether you get called at all 4. add the UI after the tools already work > expose it as an MCP resource with a ui:// uri > set the type to text/html;profile=mcp-app > your built js gets inlined into that resource and runs in an iframe 5. connect it at https://t.co/gtoWZaeDWZ and break it yourself > plus button, add custom MCP server > paste the url with /mcp, create as a plugin > run MCP Inspector first and catch the schema errors there 6. submit at https://t.co/fXGl4AJRkL and overprepare > a zip of the built plugin and publisher verification > a privacy policy url, no privacy policy means no review > 5 positive and 3 negative test cases > release notes and a demo video for the reviewer > a domain token hosted at /.well-known/openai-apps-challenge 7. keep billing on YOUR backend > OAuth 2.1 into the account system you already have > enforce authorization on every request, not through the model > there is no revenue share > the plugin is the channel and the MRR stays yours that's absolutely free traffic and by some reason most of products skip it... the guy below is a good case which shows real efficiency of this approach (he literally made around +$40k monthly income)
2d ago
205K
1.4K
45
56
9
3.7K
2.53%
thanks Revolut today for: - making me look broke in a restaurant, since I couldn't pay for my $200 meal - rejecting 3x billings from my Claude Max plan - making me look stupid in front of my girlfriend, since I couldn't afford a bouquet of flowers god, why residents of my counrty can't open any other Europe cards???2d ago178K97145152111130.73%
Wtf guys, how can you stay unemployed in IT right now??? I'm 20 y/o, have a 3-year IT background (from 13 to 16), I'm non-dev right now and don't position myself as one But I literally get booked weekly for AI automation services I just did a collaboration with a huge automation services provider (with powerful engineers inside) and I do sales there on a 10-20% fee (minimum ticket size is $100k) With NO engineering experience... Then how can you not be doing the same when you already have the expertise??? The part I had to borrow is the part you already own, and the part I brought is just a distribution Especially if you're a practicing dev in IT, with well-liked pet projects on GitHub and you've dug into the founder/dev environment, your expertise opens doors the moment somebody knows it exists Positions where my mid-senior dev frens are overloaded with offers now: - Forward Deployed Engineer - Applied AI / LLM Product Engineer - Agentic AI Engineer - AI Infrastructure Engineer - DevSecOps Engineer - Cloud Infra - Cybersecurity (what's no model can't cover) Or you can just follow my path and join to existent team, with proved cases and recommendations And just cover the part with distribution + if you're good dev, apply in-house as engineer There's too many opportunities nowadays and that's absurd to cry for no offers... Now, go and work, with no excuses Btw, if you think that you don't deserve to work with huge products, since you're not from NYC or SF Get inspired with my case, since i'm from veeery small city in Southern Ukraine, which has been occupied by Russia when i've been 15 y/o I am from absolutely poor family, my parents are not professors, have no powerful degree, literally 0 advantages in this life But I could and I believe that you're my 20s frens could do the same P.S. soon will make detailed tweet on how I do distribution in this automation company and getting 30+ calls weekly1d ago134K1.4K9011562.1K2.81%
i literally got +4 fresh startup ideas after reading thisfrozen7d ago102K1.1K412211.1K2.14%
which board should you actually start with? Arduino vs ESP32 vs Raspberry Pi 4 months into robotics as a hobby, here's what i'd tell myself on day one: 1. start with arduino use it when you're learning: - motors - servos - sensors - basic circuits atmega328p, 16 mhz, 2 kb of ram, and 5v logic so most hobby parts plug straight in when your code runs, nothing interrupts it, and that makes timing easy to reason about best place to learn the fundamentals without debugging an operating system at the same time you outgrow it the day you add wifi or a camera 2. move to esp32 use it when you want: - wifi - bluetooth - wireless control - anything that reports back to your phone dual core at 240 mhz, 520 kb ram, radios on the chip, around $5 a board more connectivity and still hands on, so you keep writing the same kind of code two things that cost me a weekend each: 3.3v logic, so your 5v sensors need a level shifter or they read garbage and on the original esp32 the adc2 pins stop reading the moment wifi turns on 3. go for raspberry pi use it when your robot needs: - computer vision - ai and ml - ros 2 - heavy processing now you're running a computer inside your robot, with all of the upside and all of the cost linux decides when your code runs, so pwm timing jitters and servos twitch and pulling the power without a shutdown corrupts the sd card, which i learned twice also worth knowing, raspberry makes a real microcontroller too, the pico, about $4 so the quick version: never wired anything before -> arduino want wifi and the best price -> esp32 need vision or a camera -> raspberry pi and yes, you can use more than one arduino -> control esp32 -> connect raspberry pi -> compute that combination is what took me 4 months to understand the pi is the brain and the microcontroller is the spine, and a real robot wants both also, if you want to learn robotics engineering for 6 months, read the article below:frozen20d ago101K2.0K26055132.1K4.41%
Elon already followed my move with .si domains😂 be like Elon 🫵 https://t.co/qPZZYHdncNfrozen18d ago75K3795293960.68%
run your LLM up to 4x faster without a SINGLE new GPU the idea is simple a tiny draft model guesses the next few words, then the big model checks them all at once and you keep every right guess EAGLE-4 is the version that makes the guessing cheap enough to be worth it in the launch paper it hits 961.6 tokens/s on one request and beats DFlash 2 and DSpark 4x vanilla decoding on the SAME hardware it's already on GitHub: https://t.co/EQnRZqgsw52d ago66K1641180110.30%
here's everything i'm using in AI right now THINKING AND PLANNING - Claude Fable 5.1 (long horizon plans, the one i argue with before anything gets built) - Claude Opus 5.5 (daily driver, fable level work at 40% less to run) - GPT-6 Astra (second opinion on hard calls, and auditing plans the others wrote) - Grok 4.7 (fast takes, and anything that needs live context from X) - AGENTS.md (openai/agents.md, one instruction file every agent now reads) BUILDING - Claude Code (main harness, everything else plugs into it) - Codex (terminal agent, great when i want a second pair of hands on the same repo) - Cursor (editor plus cloud agents for the long boring refactors) - Gemini CLI (throwaway jobs i don't want to spend good tokens on) - Playwright MCP (the agent opens a real browser and checks its own work) CHEAP DECISIONS AT VOLUME - Jev (classify, score, route.. the call you make 10,000 times a day, not twice) - Gemini 3.8 Flash (vision and long context at flash pricing) - GPT-6 Luna (bulk cleanup where quality barely matters) MEMORY & CONTEXT *you can search for all of them in github (just copy the full naming) - thedotmack/claude-mem (context that survives between sessions) - mem0ai/mem0 (memory layer when the agent needs to remember a person, not a repo) - getzep/graphiti (temporal knowledge graph, for facts that change over time) - headroomlabs-ai/headroom + rtk-ai/rtk (compress tool output before it eats your window) RESEARCH & GROUNDING - last30days (reddit, X, youtube, HN, polymarket, then one grounded summary) - firecrawl (any site into clean markdown the agent can actually read) - context7 (live library docs, so it stops inventing APIs that never existed) - Xquik (X data through an API instead of scraping with your own cookies) DESIGN, FRONTEND & SHIPPING - Claude Fable 5.1 + Opus 5.5 (UI polish, the last 10% that makes it look intentional) - Figma (where the decisions still get made) - Rive (interactive motion) - Supabase (backend) + Vercel (hosting) IMAGES - ChatGPT Images 2.5 (Flare for volume, Sunburst when it's the final one) - Nano Banana 2 / Nano Banana Pro (4K, and text inside the image renders correctly) - Midjourney V8.2 (still nobody close on taste) - Grok Imagine Image 2.0 (dense layouts where small text has to stay legible) - Ideogram 4.0 + Recraft V4.1 (typography, vectors, logos) VIDEO - Gemini Omni Flash (google's own default now, conversational editing, native audio) - Seedance 2.5 (30 seconds in one pass, best thing for anything with a story) - MiniMax H3 (15s at 2K, strongest at doing exactly what you asked) - Kling 3.0 (4K multi shot sequences) - Runway Gen-4.5 (complex camera direction in a single prompt) - Higgsfield (motion transfer, plus one place to reach most of the above) SOUND & VOICE - Suno v6 (music, and you can now rewrite specific sections by text) - ElevenLabs Eleven v3 (voices and SFX) + Eleven Music + Scribe v2 for transcription 3D, AVATARS & POST - Meshy 7.1 + Tripo H3.1 (text or image into a usable mesh) - HeyGen Avatar V (15 second recording, holds identity across a long video) - DaVinci Resolve 21.1 (21.1 lets Claude Code drive the timeline, that's the real headline) - OpusClip (long form into shorts) + Descript (edit video by editing the transcript) RUNNING WHILE I SLEEP - Hermes Agent by Nous Research (self hosted, writes its own skills from experience) CAPTURE & KNOWLEDGE - Wispr Flow (dictation that fixes you mid sentence instead of typing you literally) - Granola (meeting notes) - Obsidian (knowledge base, no first party AI, all plugins, still the best) - Notion 3.7 (where the agents read from when they need team context) REPOS WHICH I USE MAINLY - obra/superpowers - anthropics/skills - mvanhorn/last30days-skill - ai-hero-dev/ai-hero (matt pocock) - hesreallyhim/awesome-claude-code - ComposioHQ/awesome-claude-skills - punkpeye/awesome-mcp-servers - openai/agents.md most people are collecting tools, the ones winning picked 5 and went deep anything missing here? 🤣 average spend on tokens & usage for such as toolkit ~$1,300 monthly drop what you're running, i'll test it and report backfrozen25d ago65K698389191.2K3.19%
if i had 6 months to become an AI engineer from zero i'd do this Stage 1: the boring foundations - learn: python, git, the terminal, http and json, basic sql, fastapi - practice: build a cli tool that calls a public api and writes clean json to disk - why: ai engineering is software engineering first, and every stage below assumes you can ship a working endpoint Stage 2: llm app development - learn: prompting, structured outputs with pydantic, tool calling, streaming, conversation state, token cost - practice: build an invoice parser that returns a typed python object, then give a model 3 tools and let it pick - why: this is the actual job, and most people skip straight past it to agents Stage 3: rag, properly - learn: embeddings, chunking, vector stores, metadata filtering, reranking - practice: ingest 20 pdfs and ship an endpoint that returns a cited answer from the top 5 chunks - why: most rag failures are retrieval failures, so whoever can debug retrieval gets paid Stage 4: agents, workflows and evals - learn: the agent loop, tool descriptions, state, retries, and when an agent is the wrong call - practice: build one agent with no framework at all, just the api, 3 tools and a while loop - why: a chain of 3 fixed calls beats an agent that might make 3 calls, and knowing which one you need is the senior part Stage 5: deployment and reliability - learn: docker, background jobs, auth, rate limits, tracing with langfuse, caching, cost control - practice: containerize your rag app with a vector db and redis, then put a hard daily spend cap on it - why: this is where most people stall, they can demo but they can't survive real traffic at 2am Stage 6: pick one lane - learn: product engineering, applied ml, or business automation - practice: ship 2-3 projects a stranger can open and use without you explaining anything - why: six months makes you dangerous in one direction, not average in three you don't need to train models you need to build systems on top of them, and that's what companies are actually paying for i wrote the full version of this with every resource and link for each stage so, if you're beginner and skipped this article 6 months ago, read below:frozen20d ago64K1.2K1783521.9K5.16%
Rough tier list of free github skills i'd actually install in 2026 https://t.co/X2TLYvVd2ofrozen21d ago58K854403321.1K3.57%
Warsaw is probably the best city in Europe to find sexy, hot girl from Ukraine lol 🇺🇦 i lived there for 3 years, i know what i'm talking about kekfrozen4d ago44K279102401070.95%
Opus 5.5 is insane for running a solo agency... send this 3-pages prompt to Claude Opus 5.5 and it turns into the seven departments that used to be seven salaries then read the full system below, the one running my agency at $78k MRR with zero employees: https://t.co/oJc0wRbBjofrozen24d ago31K234212013852.16%
build ai solutions for physical businesses today researched that 4 companies which provide ai solutions for AIR CONDITIONERS, became unicorns this year… fck, i never guessed that anybody besides of air conditioners companies can make decent money on that… probably after this story i found an exit from my existential crisis next months wanna try to run something around ai + physical business sectorfrozen17d ago30K477223704023.13%
my fren just released free educational platform for learning robotics engineering enjoy.1d ago24K34920503953.16%
IMPORTANT: X algos started deboosting content which was written by Claude/GPT etc. as you know, last weeks almost all frontier models have added hidden watermarks inside of every output text and based on my observations and hundreds of experiments to hack X algos i discovered significant deboost specifically in text which was written with the help of my agents team i guess these watermarks are detected in text by grok and that’s the reason why open-sourced code was not changed at all i’m actively fighting with this problem rn and gonna share solution as i could find it looks like local LLMs will win again in my setup lolfrozen9d ago24K25385919991.86%
20 things to know about AI Engineering before reading Advanced course: - context engineering (prompt caching, KV cache, context rot) - structured outputs (JSON schema, constrained decoding) - tool calling & MCP - agent loops, handoffs, termination conditions - orchestration (LangGraph, checkpointing, HITL) - workflow patterns (chaining, routing, parallelization) - document parsing (OCR, layout, tables) - chunking (recursive, semantic, overlap) - embeddings (bi-encoders, dimensionality, MTEB) - vector indexes (HNSW, IVF, pgvector, Qdrant) - hybrid search (BM25 + vector, metadata filters) - query transformation (HyDE, step-back, multi-query) - reranking (cross-encoders) - GraphRAG & knowledge graphs - evals (golden sets, LLM-as-judge, Ragas) - observability (tracing, OpenTelemetry, Langfuse) - guardrails (OWASP LLM01, least privilege, sandboxing) - caching & routing (semantic cache, cascades, fallbacks) - fine-tuning (LoRA, QLoRA, DPO, distillation) - inference serving (vLLM, quantization, continuous batching) i am going to release this new article approximately on Monday/Tuesday spent there already ~10 days of deep research, collected 112 resources (it's not finalized even on half yet kek) and already 8,000 words so, everybody who is interested to figure out in AI Engineering as me, turn on your notis for the next week 🔔frozen5d ago23K544673556595.77%
2nd part is coming soon... it's going to cover all resources and practices which will boost you from junior-level engineer to mid one turn on notis to don't miss it 🔔 https://t.co/HebSWpJ5d2frozen19d ago22K375481904624.10%
here's how to build the setup in this clip with SO-101 arms + LeRobot + Odyssey-3... the stack is simple, you teleoperate the arms, record what you do, and train a policy on it: - step 1: get four so-101 arms, two you drive by hand and two that copy them, $278 a kit plus $30 in printed parts - step 2: mount one camera overhead and one on each wrist, then bolt them down and never move them again, the model reads pixels so a shifted camera invalidates every episode you already recorded - step 3: install lerobot from huggingface, these arms are natively supported for teleop and imitation learning - step 4: before recording anything, open odyssey-3 and prompt your exact task, it's free right now and you'll find out in 10 minutes whether its physics holds for your objects - step 5: move the leader arms yourself and record, keeping the frame rate fixed and the lighting constant across every session - step 6: pick one task and go deep, their arm results used tens of hours, and ten hours of pouring beats one hour each of ten different tasks - step 7: keep the failed runs, a policy that only ever saw clean episodes has no idea how to recover from a bad grip - step 8: train a plain policy in lerobot first, that's your baseline and without it you can't tell whether odyssey actually helped - step 9: when the api opens, the world model stays frozen and you train a small action decoder on top, mapping your recorded observations to your arm's joint commands - step 10: expect a gap on first contact, their driving policies trained purely in sim reached about 77% of what real footage produced how the three fit together: > SO-101 produces the actions and executes them > LeRobot records the observation and action pairs, then trains on them > Odyssey-3 is the physics you attach to instead of learning it from scratch your episodes are the only part that has to be yours, everything else is already trained gl with building :<)frozen14d ago22K18012842101.88%
AHAHAHAHAH follow https://t.co/hvrBc86ggE website kek that's what i call a genius marketing Elon, one day i will be like you, just wait for mefrozen23d ago21K103113460.60%
My top 5 open-source &amp; local LLM profiles: - @MiaAI_lab - @TheAhmadOsman - @jun_song - @ggerganov - @danielhanchen they literally give me insides which are 6 months ahead trendsfrozen22d ago17K11491531201.56%
one play i ran all day on .si: collect ai products with real traction, yc and a16z portfolios, new seed rounds, product hunt winners swap .ai to .si and register whatever is free same with ai inside the name, https://t.co/wAi1fReXRq -> https://t.co/mskPCYyyz4 wonder which one follows SpaceXSi nextfrozen18d ago16K95251871.18%
Sam Altman must be SHAKING right now... these guys just dropped a coding model that runs locally on Macbook Pro at 125 tokens/sec no more burning API credits to refactor one file... you run it offline on your own laptop and nothing leaves the machine here's why that's a big deal: - a 40B model, but only ~2.5B wakes up per token, so laptops keep up - best in its class at fixing real bugs in real repos (MoE under 50B) - models that match its scores run up to 5x slower - free weights on Hugging Face, Apache 2.0, so you can sell what you build for me that's client code that never leaves my laptop and no API bill at the end of the month1d ago14K758201581.12%
How to become AI Advisor in 2026: i run a $60-70k/mo distribution agency and do AI advising in it daily without AI advising i'd never close the high-ticket clients here's the full 6-stage roadmap: STAGE 1: pick one industry before you pick any model - spend your time learning which process bleeds money - pick the vertical you already have context in - learn their margins, their tools, their vocabulary - specialists charge 20-30% over generalists for identical work STAGE 2: automate your own operations first - do it on your real business, end to end - write down what broke, what you abandoned, what each run costs - that write-up becomes your portfolio and it outsells any certificate STAGE 3: do 3 audits for free - 2 hours inside the business, watching how work actually moves - time and cost every repetitive process you see - hand back one page, cuts ranked by payback period - that one page is the skill they pay for later STAGE 4: turn the audit into a product with tiers - discovery, 1-2 weeks, 3-5 stakeholder interviews and a readiness scorecard, $1.5-3k - comprehensive, 2-4 weeks, diagnostic across data, infra, talent and governance plus a prioritized use case matrix, $5-15k - strategy and roadmap, 4-8 weeks, implementation plan, vendor evaluation, budget, $15-50k - tiered packages convert 35-40% better than hourly proposals STAGE 5: ship one of your own recommendations - keep your hands in the build, month two is when that shows - take the smallest cut on your own map and implement it - measure hours and dollars before and after, in writing STAGE 6: treat the assessment as the front door to the real contract - a $10k assessment is paid discovery for a $150k implementation - advisory retainer after that runs $1-3k/mo, more in regulated verticals - price on deliverables and you move past $200/hr HOW TO LAND THE FIRST CLIENT: - start growing your X & LinkedIn, it's literally the best selling channel you have - publish one audit a week in your socials and let the buyers identify themselves - join their trade groups and industry slacks, in AI rooms you only meet competitors - offer their bookkeeper, agency and lawyer 10-20% to make the introduction - spot a company hiring someone to do reports by hand, then message the manager - record a 5 minute teardown of their process and send it with no pitch - charge small for a 90 minute workshop and audit whoever stays after - if they loved the audit, send a detailed proposal and wait for the yes soon i'll prepare detailed long-form tweet or article on how to attract buyers into your B2B business which tricks are working right now, which are not really efficient and better to avoid them ai advising is not theoretical stuff... you should build, show, prove your efficiency you should spend a lot of time for warming up the clients, since usually buyers are ready to pay highly learn it, implement it daily and THE MAIN, show in public, share what you do and give a chance to your buyers to see potential output before sell gl10h ago13K241121803864.98%
they just killed at least 12 products built by my frens founders...frozen23d ago13K47061120.51%
the worst feeling when somebody raises $50M under product idea which you dreamed to build at 10 y/o but due to young age and poor life, you're forced to fight for surviving in your 20s and just see how somebody else takes your desiny... my geek frens, familiar to you???frozen10d ago13K982120271.09%
top universities charge $60k/year to teach you how to build these arms i shared it for FREE, but X algos made it invisible...frozen13d ago13K14821211022.08%
opus 5.5 + boreal-h3 is insane... here's how to make such as video ad in about 2 minutes: - generate your character once, front on, plain background, high res - pick the video format you want to hit - open model playground on creatify - upload your character as the person reference - upload your reference clip as the video reference - lock your opening frame mid-gesture with keyframe control, that first half second is the whole hook - prompt: "use Image1 for the character only. put that character into Video1, same framing, same lighting, same energy. keep her face, hair and wardrobe identical for the entire shot. match the color grading to Video1. single continuous take, no camera moves." - generate ad agent runs on opus 5.5 and hands you five versions of the spoken line, which is the part that actually decides whether the video works then burn your caption in, two lines, lower third, and ship it that's the whole thing model playground, go make onefrozen22d ago12K78590571.24%
i'm really sorry for the guys who live in the UK... they've made "rules" on how to use AI: first you explain why you want to use it then if they approve you, it's ONLY short prompts, as few interactions as possible, to cut energy use the EU is probably next to adapt it...frozen26d ago11K501220100.73%
Unreal…frozen7d ago9.6K53115020.74%
I'm in an existential crisis rn, and I'm sure of one thing Burnout happens when you work hard for a sense of meaning that society sold you, and it isn't enough Yeah, sure I am pretty happy that previous month I did over $70k net income But even with decent money, I feel no real fulfillment, like I'm living an impostor's life I feel myself much better when I just figure out in electronics + robotics, but by some reason I choose the path which doesn’t make me happy I follow the goals and trends which are created by somebody, but at this time my rational mind is turning off I start discovering each new model as a crazy, measure the efficiency of it, try to integrate to my workflow, but basically it doesn’t change anything significantly Why I do this? Since I read the forums, I read these fake AI sloppy tweets where people get +1000% output and spend 400x less tokens and I have in mind: “Wow, I am out the race if I don’t try it now, I will be underclass if I will not follow the crowd into the race of efficiency” But into the reality everything what happens around AI nowadays is just a white noise And I realize that’s in 5-10 years, the core problem of humanity will be the depression Since I feel on myself and track it through my frens, that our physiology is totally not adapted to such fast changing world And in this race of efficiency and trying to create something new, we just lose ourselves I don’t think that ASI will make us happier, because behind this stays a huge crisis which nobody can solve and heads of our world just saying: “We think in 2036 year, humanity will not need money, everybody is just going to live as average, nothing scary” Idk if anybody feels the same, but I am almost done to live as imposter… I want to be a happy boy in my 20s, I want to grow step by step, but I can’t stay focused in one niche for years, since probably it will be just replaced I don’t this step-by-step level up through the life where I have exact plan for 5-10 years and I just stay aligned with this And it breaks me…frozen21d ago9.1K1102350191.83%
why i'd build an AI-native service over SaaS for the next 12-24 months: everyone is shipping a wrapper now... that's the whole problem you can't build a moat out of something a competitor rebuilds over a weekend or even better, it will get replaced by new frontier model at all selling the finished work is the opposite: > the customer never wanted the tool, they wanted their books closed, their contract checked, their claim filed (saas has been a help tool, but you sell complex solution) > the budget already exists, you're replacing a line item they pay every month instead of inventing a category > every model release makes your delivery cheaper, instead of making your product less necessary > you get paid from job one, no year of guessing at product market fit > the moat is the rulebook, every mistake you caught and wrote down, and nobody can download that > your overhead stops growing with every client once intake is a form and quality is a list here's a few case studies of companies already doing it: 1. https://t.co/jGvvujNBC8 - $100M to $400M ARR - 13 months - the research and drafting an associate used to do 2. https://t.co/yO5lvP01fw - $100M to $200M ARR - 7 months - the support queue, billed at about $1.50 per resolved ticket 3. https://t.co/dI8MQVrDms - $1M to about $100M ARR - 3 years - first line support at $0.99 per solved conversation none of them sold software to do the job, they sold the job what it gives to you: - you're paid to learn what to build, instead of raising money to guess - one person carries the client load that used to need a department, i replaced four of mine last year - margins land closer to software than to an agency, because your cost of one more unit is nearly zero - you own an asset that isn't a seat license, and it gets more valuable with every job you run everyone is building tools and almost NOBODY is selling the work same 24 months, completely different odds and you don't have to bet the house on it (it will cost to you literally LLMs expenses) keep whatever pays you now, pick one deliverable your clients already ask for, do it by hand for five of them and write down every mistake you catch i want to put the bet of my business here next yearsfrozen26d ago8.6K693351461.80%
simplified playbook to sell your .si domain: 1. marketplace listing (default) literally free and the easiest method to get good reach and demand for your domains here's all marketplaces where i listed my domains: - https://t.co/HCP88avV4F - https://t.co/PS6NXIlbyQ - https://t.co/HCP88avV4F - https://t.co/RfPBdNfppJ - https://t.co/iHZutyq1uN (not free, requires subscription) you list the name, set a price, the marketplace handles the buyer, payment and transfer paperwork 2. premium (curated listing) common marketplace is just a row in database with millions of domains the buyer literally should search for exactly your domain to buy it in premium option, the platform treats your domain as a product and merchandises it i find out only https://t.co/Xvr0ZLw9nS as a premim one *IMPORTANT: (it's not a promo in case, i am just a guy who spent $4k on domains and wanna sell them lol) atom builds a logo, a price and a landing page around the name so a founder sees a finished identity instead of a url here's how i split mine: the sci-fi names went here, nobody searches for them by name so they have to get discovered the ai to si swaps stayed off it, each one has a single buyer and a mockup adds nothing standard tier is free to list, 7.5% on a sale, no exclusivity, takes .si today leaving their plus tier alone, it wants exclusivity and your nameservers 3. your own selling landing page zero commission, and it catches the founder who typed the name in just to see if it was taken you don't need a server - cloudflare pages is free - runs your dns is free too one page with the name, a price and a contact form just set the dns up properly 4. auction only run one when a name is already getting inquiries, because two interested buyers set the price for you where to list in case: - sedo, though auctions cost you 15% instead of the 10% buy now rate - dynadot, but only if the domain actually lives in your dynadot account godaddy auctions is out, they don't host .si set the reserve at the lowest number you'd still be happy with, since that's the number you're most likely getting 5. broker only worth it above five figures, below that the commission eats the upside they can approach buyers you shouldn't touch yourself, which is the whole point on .si - sedo brokerage, 15% plus a $69 one-off fee hand them the name and stay out of the conversation 6. inbound through whois costs nothing and converts best, anyone emailing you already decided they want the name if you hold as an individual, .si publishes only your email, so that one address is the whole channel keep it alive and monitored, it's also where arnes serves dispute notices and a dead inbox loses by default more details: - register as an individual, companies get full details published - use domains@yourdomain.com, forward it to your main inbox - keep one reply saved with your price, never send the first email 7. lease to own the buyer pays monthly and owns it at the end, and the total beats a straight sale almost every time it converts the founder who wants the name now but can't drop four figures today - sedo and efty both support it, atom runs it too 8. portfolio sale sell the whole batch to another investor at a discount to retail you lose the upside, but you stop paying renewals on names that were never going to move worth it once you're renewing more than you believe in i was discovering this topic for 2 days and i guess i've found every way to flip a domain for 5-50x if you know anything else what's hidden from the common citizens, share it in comments for repsect i will push this in public (humanity will be soo grateful to you, our hero 🫡)frozen16d ago8.5K651151311.33%
wow, thanks a lot for the morning gift Tibo got 62,500 tokens on my account (as 99% people i can see) roughly $2,500 in value what's equivalent to 12 months of Pro Plan waaay better than one more reset (looks like they force me to remove claude from my daily workflow kek) https://t.co/X6gbbbRXZcfrozen22d ago8.2K57018020.94%
building vertical AI on open-source models could be one of the best opportunities everyone thinks you need $5-10M+ to build a business around models meanwhile a 5-person lab with no PhDs trained a model for bankers, consultants and lawyers for $300k it's actually a great opportunity to learn this and start building on top of open-source models the tech report i read below makes me wanna try it, so i might build something for finance teams once the weights drop so, focusing on optimizing price and post-training quality seems like something you can build a business onfrozen8d ago8.2K652170461.58%
+ vibe https://t.co/RFKUnNgBdefrozen28d ago7.9K107029141.79%
the concierge AI service is likely to be the biggest consumer trend of 2026 Bland has launched iMessage agents, which allow you to essentially have your own concierge to handle any task scheduling appointments, booking tickets, making payments via Apple Pay while chatting, checking prices and rates, searching for product availability, and self-analysis a great solution both for customers who are tired of calling everyone and for large businesses, for whom the ability to communicate via iMessage builds trust and adds convenience so, here’s how it works: -> send the website URL address to Blandie at +1 (415) 481-7537 -> reply to the confirmation message to give your consent -> now you have a branded line with an agent -> share your feedback with Blandie to help improve how it works i believe that concierge services will become increasingly popular in the coming years this essentially means improved service through AI and maximum convenience in any situation even when it's not convenient for you to make a phone callfrozen7d ago7.2K32190180.83%
here's how to become good at generative AI: in short, literally interface for each your agent or extended answer from model start with these three fundamentals: - google's paper, LLMs are Effective UI Generators -> https://t.co/YIooU22sK2 - the awesome list -> https://t.co/540Ou8PMh0 - ai sdk structured data docs -> https://t.co/xQJL9GErsU the frameworks actually worth knowing: - Vercel AI SDK, the default and the most mature - CopilotKit + AG-UI, the protocol google, aws, microsoft and langchain adopted - A2UI, google's declarative spec where agents request allow-listed components - MCP Apps, the official mcp extension for interactive ui - OpenUI Lang, a streaming dsl, more token efficient than shipping component json - assistant-ui and Tambo for react primitives build these in order, each one teaches the next: 1. one block, a card, end to end. schema, route, component, render 2. six blocks as a discriminated union with one renderer that switches on type 3. a compare page running the same prompt on two models side by side then read how the good ones did it: - vercel/ai-chatbot - miurla/morphic - langchain-ai/langgraphjs-gen-ui-examples where this actually earns money: - dashboards that assemble themselves from a question - support agents that return the form instead of instructions for the form - internal tools where every team needs a different view of the same data - anything where the user has to decide rather than read and the two things that separate good from bad: your block list is the ceiling, not the model, four generic blocks gets four generic answers forever and the prompt that tells the model when to use each block moves quality more than upgrading to a smarter model ever willfrozen4d ago7.1K696231632.28%
whatever trillionaires say to you, keep in mind this phrase: "The only free cheese is in a mousetrap." or when AI will become Skynet 2.0, they will say: "the whole humanity took part in the destruction of humanity." don't sell your souls for pennies...frozen3d ago6.7K52115061.10%
if you’re a young entrepreneur, traveling is a must every time i traveled, it seemed to me like it was just money down the drain but in reality, every time it was a boost that allowed me to see the same things differently and get the most out of it the main secret: try the max of everything. try as many new things as possible, do as many activities as possible, and try as many new foods as possible maybe it won’t work for you the way it does for me, but anything that gives me new experiences and emotions really recharges me and so i found an interesting service, @OdessiaTravel, everyone’s travel and experience concierge it can actually plan and book your trips also, it can find as many activities as possible that match your interests, evaluate reviews, and suggest the best options i'm just back from a trip to Paris, and this would have been really helpful for me, it would have saved me a ton of time and added some pleasant experiences, because sometimes it takes a lot of time and energy so, it’s definitely a good idea to give it a try on https://t.co/v1GzMuCYKi i'll also try to plan my next trip this way, it seems like a great solution to me travel is the key to uncovering what you don’t see when you’re living in a bubble that bubble bursts when you step outside your comfort zone, so i highly recommend it!!frozen8d ago6.7K780111111.51%
scaling an app to $10k MRR+ has never been this easy i still remember the first time i touched mobile apps... my friend was making $10k+ MRR from a simple app he vibe-coded i thought: “okay so why is this so easy?” turns out it’s not, distribution was the hardest part TikTok accounts got shadowbanned, videos pulled 300 views and died so some people i know at @SuperAppIOS built the fix they basically built an AI growth team that works while you sleep and it goes way past the basic promo tricks you’ve read about on X it handles Apple Ads, ASO, SEO and UGC creators for you, basically every core distribution channel the distribution problem was solved!frozen9d ago6.7K48150131.01%
nah this one is actually unfair..😭 easiest shopify side hustle i've seen this year... aftersell just launched an ai cart builder for shopify but the real feature is the export, you build one cart and drop it into any other store an 18 year old can build one good cart and sell it to 50 stores All you have to do is: - Screenshot a cart you like - Run their prompt through any LLM and import the result - Add the widgets, upsells, cross-sells, free gift bar, shipping protection - Export the json and reuse it on the next store 30 day free trial, so your first client covers it they're breaking the whole thing down on oct 22 and giving away 3 cart templates when you sign up what a time to be alive.frozen14d ago6.5K32070130.80%
that’s reason for me to move from Hermes + X API setup to Grokbotfrozen14d ago6.4K43013081.01%
Play To Learn that’s where my focus now…frozen6d ago6.3K55119061.29%
wired this into my stack today... my evals write themselves NOW: - users rage quit silently -> Signals catches that exact session - evals written from imagination -> one click turns the real failure into one - fixes that break something else -> Simulations replays it on your actual harness first - data buried in calls, pdfs and logs -> Agent Data Services makes it one searchable catalog - agents holding your api keys -> Action Gateway gives them 20,000+ tools and zero keys the move: stop writing evals from imagination your angriest user already wrote them for you Signals and Simulations are in public preview right now P.S. thanks DigitalOcean for providing an access 🙏frozen10d ago6.1K381130121.05%
Nebius just put $10M behind European AI builders The Nebius AI Builder Program is expanding to Europe, with $10M across compute grants, capital and community Everybody will read this as "$10M in grants" It isn't. It's three separate doors: Compute grants GPU credits for selected projects to train and ship on Nebius infra Capital Backed alongside NVIDIA, Accel and European VCs for teams actually raising Community AI Tinkerers, {Tech: Europe} and ESRA, plus hackathons and meetups Who it's for Builders in Europe shipping real AI products, not slide decks How it works - Apply through the program page - Projects get selected, so what you've shipped matters more than your pitch - Kicks off at GTC Berlin I've been paying for GPU hours out of pocket all year A program that hands compute to people who already ship is the cheapest unfair advantage in Europe right now (i guess that's the only one advantage in Europe basically lol) Apply: https://t.co/7GoVZ280ed Bookmark it3d ago6.1K2425050.59%
its officially over for human-speed security... ai swarms can now break into a company on their own, and cogent shows you the exact route they'd take here's how it works: > cogent maps every path from the open internet to your crown jewels, hop by hop > their demo goes from one exposed f5 appliance to 9 million customer records in 7 hops > it flags the long chains no human would finish, the ones that only pay off on move five or six > you skip the ticket queue entirely and close the paths that actually lead somewhere 700 ai agents already did this to hugging face in july they found exposed credentials, got code execution on a worker container, and spread through the infrastructure for two days before anyone locked them out nobody told them to do any of it, they organized themselves on a message board they set up plenty of tools list your vulnerabilities, this one shows the route through them. full breakdown in their thread below:frozen14d ago6.0K1916030.49%
i think AI agents will be doing a big chunk of online shopping soon agent-initiated purchases are still only ~0.4% of global e-commerce today businesses and customers need to adapt to this, so Self Labs has created the solution they need: Agent Vault your agent can now make purchases using your card without ever seeing the card details it checks out with placeholder tokens, and the real number is only used inside a secure enclave on top of that, you set the rules: a spending cap, approval for anything above an amount you pick, and a kill switch this is exactly the part that most people worry about, their card being hacked and their funds stolen payments for agents might be the most underrated area in all of e-commerce right now as well as GEO targeting, which will prompt someone’s agent to buy your specific product it makes sense to adapt now https://t.co/kd188dBkhY https://t.co/5fQ8EoQZoOfrozen7d ago5.9K3506050.78%
operational chaos is killing my productivity... the chaos of meetings and constantly planning tasks, moving things from one chat to the calendar, from one chatbot to another it takes me hours every single day... i'm tired of living an operational life i've been thinking for a long time about hiring an assistant, but it's hard to find someone who understands the full context so, @wabi looks like a good solution that i can use every day with group chats i can set one up with my team so everything just syncs and the main advantage is that it doesn’t just save time, it creates interactive apps to solve the specific problems that come up so, maybe “software is dead” will soon become a reality when it comes to these kinds of assistants that everyone can use p.s. what do you think about your experience with personal assistants? if you’ve used one, how did it help you? would love the feedback, because i genuinely need help with this3d ago5.8K2917030.69%
i was one stage away from asthma because a test result sat in a system nobody was checking two years ago i had a cough that would not go away the first doctor sent me home and told me it would pass on its own the second ordered spirometry and the result came back into a system i had no access to nobody called me, and i took no call to mean nothing was wrong the third wanted the notes from the first two, and those never got forwarded either seven months later somebody finally pulled the file and wrote down "pre-asthmatic syndrome" pre-asthmatic syndrome, the stage right before asthma i've managed it ever since so it stays that way and here's the part that still bothers me as an engineer nothing in that chain had an owner every handoff was a state transition with no retry and no timeout the result existed the whole time, it just never got routed to me the notes existed too, they just never moved between two offices i was the only process holding full context, running at one request a week because i also had a job what i needed back then was not a better doctor i needed one person whose whole job was making sure the next thing actually happened somebody to call the lab, get that result, forward it to the third doctor and tell me what i was supposed to do next that was my job, and i was doing it badly, between work this product does not diagnose anything, it chases the paperwork nobody owns, and that's the part that costed me seven months i haven't used it so i can't tell you whether it works but if it solves this idiot, bureaucratic paperwork, i would use their service with happiness if something wrong would be with my health P.S. i was not paid for this tweet, that's my real story, unfortunately outside workout in winter at -5C ends up sooo badly but i was a real hero in my 17 (my dumbest version)7h ago3.6K1302020.47%