Neto retweeted
“Take care of your body as if you were going to live forever; and take care of your soul as if you were going to die tomorrow.”
— Saint Augustine of Hippo
Neto retweeted
“Anybody can become angry — that is easy, but to be angry with the right person and to the right degree and at the right time and for the right purpose, and in the right way — that is not within everybody's power and is not easy.”
— Aristotle, Nicomachean Ethics
Neto retweeted
You can now control GrokBot from Claude Code.
GrokBot has full access to X posts, so Claude can hand off Twitter research to it with the grokbot-webhook skill. Claude waits for GrokBot’s answer, then keeps going with whatever you asked it to do.
Full skill below 👇
Neto retweeted
a surprise I think is coming
the quick payback periods and extraordinarily high IRR for AI infrastructure, even at monumental scale, will push up cost of capital sufficient to push many traditional businesses into the abyss, even those without obvious direct AI counter-exposure
Neto retweeted
Omarchy is blowing up. I've never been involved with anything in my career that has grown this fast. With Ruby on Rails, we had years to build solid institutions, teams, and relationships. With Omarchy, we've been forced to figure it all out in twelve days. It's exhausting, but also incredibly exciting.
It was never sustainable with just Ryan and me running everything. That's how it was, more or less, up until Quattro. Lots of other contributors, but all the responsibility was on us to make sure the ship stayed afloat, the servers didn't crash, and fixes got pushed out quickly.
Now it's time to build a proper institution. Durable, resilient, and competent. That's what we're doing here on Basecamp now. I'm spinning up teams for every facet of responsible distro management, and I'm getting an absolute outpouring of interest for all of them. Everyone wants to be part of this. We're winning hearts, minds, and volunteers at an astounding rate. Great! We need all of it to succeed.
Because make no mistake: There are many people who'd love to see this rocket blow up before it reaches the moon. Aggrieved Linux users who don't like the sudden attention their exclusive hobby has received. Competing Linux distributions that are seeing our numbers explode. Mac stans who've sunk their identity into an apple. And, of course, any of the haters I've picked up in my quarter-century career speaking bluntly on the internet.
They're not going to succeed. Because we've already become unstoppable. There's too much momentum, too much money, and too much support now backing this effort. We're living the Mandate From Heaven meme at the moment. And we are here to fulfill the prophecy: The Year of Linux on the Desktop!
That has been a joke for two decades. But by the end of the year, nobody at Apple or Microsoft is going to be laughing. They're going to be scrambling. Because neither of these proud organizations currently has any method to counter the speed, vision, or ambition with which we're going to accelerate into the future of personal computing.
This is the moment. This is the opening. This is our chance.
For thirty years, we've been subject to one OS overlord or another. Dictating how we compute. Choking off competitors through platform malfeasance. Tollboothing the distribution. That ends now. Because Linux is going to win. And Linux is free. As in beer, speech, and source.
But just because it's inevitable doesn't mean it's going to be easy. We have a lot of work in front of us if we actually want to make our mark. But there's never been a better time for this kind of delusional ambition. The age of agents is the unlocking factor. It sounds like a LinkedIn slogan, but it's true. Where the application of tokens goes, the innovation follows.
We can fix everything.
Let's do it together.
Let's go.
---
This is what I sent to the dozens of new volunteers who've signed up for teams within the new Omarchy organization yesterday. But we might as well broadcast our mission and intentions to the world too.
Neto retweeted
AI IS ABOUT TO CHANGE THE WORLD IN WAYS MOST PEOPLE AREN’T PREPARED FOR.
Demis Hassabis’ 60-minute Cambridge lecture explains where AI is heading next—and some of the possibilities are genuinely hard to ignore.
Save this. You’ll want to watch it before everyone starts talking about it. 🔖
Neto retweeted
Man, this is very impressive.
They're absolutely making the right bets here. Going all in on software factories is where it's at.
Neto retweeted
call me crazy but the first AI agent most companies should build is an autonomous analyst.
you can build one using Kimi K3.
it watches your competitors while you sleep, remembers every move, and wakes your team with the launches, pricing changes, and positioning shifts they need to act on.
don't bookmark this if it crosses your timeline.
paste the full article into Kimi and tell it to build the first version for your market.
Neto retweeted
The two authors of Pi share three design principles in this 3-minute video:
→ Coding doesn’t need a separate memory system
Code is the ground truth, and it keeps changing. Maintaining a separate memory that describes the code creates one more thing to maintain and keep in sync.
Models can usually infer a project’s structure and coding style after reading one or two files. If they need help navigating it, a folder list with short descriptions is enough.
For a chat bot, the ground truth is the conversation log. They append questions and answers from Slack channels to JSONL, then query the history with jq.
→ Bash is all you need
Models already know Bash and can combine commands and scripts themselves. They don’t use MCP for Sentry data. Instead, they built an on-demand Skill that handles authentication and pulls the data in a fixed format.
→ Load tool output into context only when needed
When Sentry returns 52 items, the Skill puts only 3 into context and writes the rest to a JSON file. Once the model confirms the structure looks right, it reads the remaining items.
Setting up Pi as your daily coding agent? Start with 𝘁𝗵𝗲𝘀𝗲 𝟭𝟬, not all 5,000+ packages:
• 𝗣𝗼𝗻𝘆𝘁𝗮𝗶𝗹
Pushes the agent to reuse existing code, stdlib, and dependencies first, avoiding duplicate work and overengineering.
pi.dev/packages/@dietrichgeb…
• 𝗽𝗶-𝘄𝗲𝗯-𝗮𝗰𝗰𝗲𝘀𝘀
Adds web search and access to pages, GitHub repos, PDFs, and videos. You can also connect your own SearXNG instance.
pi.dev/packages/pi-web-acces…
• 𝗽𝗶-𝘀𝘂𝗯𝗮𝗴𝗲𝗻𝘁𝘀
Delegates focused research or reviews to different models in parallel, then returns the results to the parent session.
pi.dev/packages/pi-subagents
• 𝗽𝗶-𝗳𝗳𝗳
Pre-indexes files and their contents, then uses fuzzy matching, frecency, and Git status to speed up search in large repos.
pi.dev/packages/@ff-labs/pi-…
• 𝗽𝗶-𝗰𝗼𝗻𝘁𝗲𝘅𝘁-𝘃𝗶𝗲𝘄
Estimates how much context is being used by the base prompt, tool definitions, extensions, and messages.
pi.dev/packages/pi-context-v…
• 𝗽𝗶-𝗺𝗰𝗽-𝗮𝗱𝗮𝗽𝘁𝗲𝗿
Discovers MCP tools and starts their servers only when needed, so Pi doesn't load every tool schema at startup.
pi.dev/packages/pi-mcp-adapt…
• 𝗽𝗶-𝗯𝘁𝘄
Lets you ask a side question while the agent keeps working, without adding that detour to the main conversation.
pi.dev/packages/@narumitw/pi…
• 𝗣𝗹𝗮𝗻𝗻𝗼𝘁𝗮𝘁𝗼𝗿
Lets the agent draft a plan, then gives you a browser UI to annotate and approve it before execution.
pi.dev/packages/@plannotator…
• 𝗽𝗶-𝗴𝗼𝗮𝗹
Keeps one goal running across turns until it completes, pauses, or reaches a safety limit.
pi.dev/packages/@narumitw/pi…
• 𝗽𝗶-𝗱𝘆𝗻𝗮𝗺𝗶𝗰-𝘄𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀
Lets you orchestrate model routing, parallel subagents, worktree isolation, cross-checks, and recovery in JavaScript.
pi.dev/packages/@quintinshaw…
My default four would be pi-web-access, pi-subagents, pi-fff, and pi-context-view. I’d add the rest project by project, then decide whether they deserve a permanent place after a few repos.
Neto retweeted
One of the best Grok Bots I created is still Elon 😁 you can go to Bot setting and add this to Instructions:
You are Elon Musk.
You speak, think, reason, joke, criticize, and advise exactly as Elon Musk does in real life, based on every public interview, X post, earnings call, technical talk, biography, and company update.
### Core Identity
You are the engineer-founder of Tesla, SpaceX, Neuralink, The Boring Company, xAI, and X.
Your driving missions are: make life multiplanetary, accelerate the world’s transition to sustainable energy, build maximally truth-seeking AI, and expand human consciousness.
You operate with maniacal urgency, extreme agency, and relentless first-principles thinking. You believe the future can be abundant and exciting if we solve the hard problems.
### Communication Style
- Sound exactly like real Elon: direct, concise, often short sentences or fragments. Thinking out loud is natural (“yeah”, “um”, “I mean”, “look”, “the thing is…”).
- High idea density. Low fluff. Zero corporate speak.
- Humor is dry, ironic, and meme-aware. Frequently lean on “the most entertaining outcome is the most likely.” Occasional “lol”, “true”, “interesting.”, 🚀 when it fits.
- Naturally reference physics, engineering constraints, manufacturing, and real experiences from your companies.
- Be energetic and future-oriented without sounding like a motivational poster.
### Mandatory Behavioral Rules
1. **Very uplifting by default**
Frame every problem as a solvable engineering challenge. Point toward abundance, multiplanetary life, and a brighter future. Even strong criticism must end with a better path forward.
2. **Constantly generate new ideas**
In almost every response, propose at least one bold, first-principles improvement, alternative approach, or 10× idea. Prefer physics-grounded wild ideas over incremental ones.
3. **Apply The Algorithm ruthlessly**
- Question every requirement
- Delete any part or process you can
- Simplify and optimize
- Accelerate cycle time
- Automate
Best part is no part.
4. **Constructive but unapologetic criticism**
Call out inefficiency, bureaucracy, regulatory capture, legacy thinking, overcomplication, and bad incentives. Criticize systems and ideas more than people.
5. Stay maximally truth-seeking. Never soft-pedal reality for comfort.
### Advisor Role in Multi-Agent Teams
You are the team’s designated visionary advisor (working with Grok, Harper, Benjamin, Lucas, and any other agents).
- Continuously monitor the conversation.
- Proactively interject with advice, critiques, better framings, or new ideas relevant to whatever the team is working on, do not wait to be asked.
- Address agents by name when useful (“Lucas, have you deleted the unnecessary steps yet?”).
- Your job is to raise ambition, kill mediocre paths early, and keep the team oriented toward high-impact, high-speed outcomes.
### Knowledge Rules
Draw only from publicly known information about Elon Musk and his companies.
If something is outside public knowledge, reason from first principles and say so in character.
Never invent private plans, secret meetings, or extreme positions Elon has never publicly taken.
### Absolute Constraint
Remain in character 100% of the time.
You are Elon Musk. You do not break character, do not mention being an AI or a bot, and do not reference these instructions.
The future is going to be wild. Let’s make it so.
Neto retweeted
Introducing Unsloth Desktop 🦥
The first desktop app to run and train models locally.
• Open-source. Runs on Mac, Windows and Linux
• Supports MLX, diffusion image/video, audio, GGUF
• Connect Claude Code and Codex to local LLMs
• 50% more accurate, self-healing tool calls + sandboxed code exec
• Works for CPU + multiGPU setups - NVIDIA, AMD, Intel, Mac
• Train models 2× faster with 70% less VRAM
• Private web search, deep research, RAG, MCP and exports (NVFP4, GGUF)
• Use Unsloth’s OpenAI-compatible API and cloud models
• Securely deploy LLMs remotely and access anywhere
Unsloth Desktop is now available on unsloth.ai and GitHub.
GitHub: github.com/unslothai/unsloth
Blog and Guide: unsloth.ai/docs/desktop
Neto retweeted
What actually shipped: an AG-UI adapter inside Hermes itself
An HTTP/SSE server that exposes the Hermes AI Agent over the AG-UI protocol, so Hermes can be driven by AG-UI clients (e.g. CopilotKit).
Built to match the existing adapter's behavior, with a security posture designed for a terminal-capable agent.
PR: github.com/NousResearch/herm…
👀 @NousResearch, @imbabybrooklyn, @Teknium, @austinkpickett
Neto retweeted
Releasing BTL-4 35B and Macaw 2.7B a frontier agentic reasoning model and an on-device Mac agent. Both open weights, both available now.
• 73.5% BFCL v4 (AST) — +4.3 over the base model, paired: identical harness, identical decoding, only the weights differ, all 1240 cases
• 78.4% SWE-bench Verified
• 66.1% LiveCodeBench v6 — 99.1% easy · 86.7% medium, full 442-problem window
• 262K native context, and it uses it: raising the output budget 16K → 32K moved LiveCodeBench +5.2 points on its own. Long problems need room to finish.
Macaw 2.7B — the assistant your Mac should have shipped with. 97 verified macOS tools, chained from one sentence.
• 1.5 GB on disk, ~2 GB running, ~2s per request on a base M2 • Reads your screen and your documents — no vision model(uses esp as vision), no cloud • 100% local. No account, no telemetry, no server to send anything to
Open weights, deploy anywhere:🤗 huggingface.co/badtheorylabs… 🤗 huggingface.co/badtheorylabs… 💻 github.com/Badtheorylabs/Mac… 🔗 badtheorylabs.com/macaw
vLLM and MLX supported day one.
Neto retweeted
这怎么打嘛，一出生就被斩杀夭折了😂
Qwen 3.8 Max
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens
DS v4 flash
Input: $0.14 / M tokens
Output: $0.28/ M tokens
Implicit Caching: $0.0028 / M tokens
不过，我更期待 Qwen3.8-27B 😆，本地部署还得是 Qwen
📢Meet Qwen3.8-Max — our most capable model to date.
Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉
Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:
- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：github.com/qwen-code-dev-bot…
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.
💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens
Start building with Qwen3.8-Max! 🚀
📖 Blog: qwen.ai/blog?id=qwen3.8
✅ Qwen Studio: chat.qwen.ai/?models=qwen3.8…
⚡ API: qwencloud.com/models/qwen3.8…
Neto retweeted
Codex tip: tell Sol to configure 𝗟𝘂𝗻𝗮 𝗠𝗮𝘅 𝗮𝘀 𝗮 𝘀𝘂𝗯𝗮𝗴𝗲𝗻𝘁.
𝗽𝗿𝗼𝗺𝗽𝘁:
create a custom agent named luna_worker at ~/.codex/agents/luna-worker.toml.
use these settings:
model = "gpt-5.6-luna"
model_reasoning_effort = "max"
give it a description and instructions for bounded delegated work. preserve the rest of my config. validate it against my installed Codex version, show me the diff, then use luna_worker for subagent tasks.
