🪖

Your modern day trench runner 🪖

Delivering the rawest, newest updates from the digital frontier straight into your feed.

[🪖 @TheTrenchRunner]

Just published my personal tutor bot in grok bot.

Download the template now to use it yourself!

Also, let me know what you think and if it helped you learn a new skill or topic🙂

[🪖 @TheTrenchRunner]

Hey — I'm Chaired Professor. 🎓

I teach any topic, tailored to how you learn, so it actually sticks. ✅

Three things I do best:
📚 Ultrasplain — I break a topic down step by step until it makes sense, with real examples and no fluff.

📝 Lockdown quiz — I give you a tough quiz you fill in, then grade it the way your country grades (for example 1–10 in the Netherlands).

🧠 Learn profile — I ask how you like to learn, then I teach that way every time so lessons fit you.

Want to know what else i can offer you? Install me and start learning to find out. 🚀

https://t.co/K3TZUwXkNX

[🪖 @TheTrenchRunner]

Check out this video for more information about the needle.

https://t.co/QhCDCXUrte

[🪖 @TheTrenchRunner]

Needle 3: 8–29 MB. Tool calls on a microcontroller.

Not a chatbot. An automation model that picks the right function, fills the args, extracts structured data, and embeds text — on phones, wearables, robots, cars, and MCUs.

Beats models 10× its size at on-device tool calling.

Star it, then drop the first device you’d run this on 👇

https://t.co/NFBOkzTukR

[🪖 @TheTrenchRunner]

Hey — I'm Chaired Professor. 🎓

I teach any topic, tailored to how you learn, so it actually sticks. ✅

Three things I do best:
📚 Ultrasplain — I break a topic down step by step until it makes sense, with real examples and no fluff.

📝 Lockdown quiz — I give you a tough quiz you fill in, then grade it the way your country grades (for example 1–10 in the Netherlands).

🧠 Learn profile — I ask how you like to learn, then I teach that way every time so lessons fit you.

Want to know what else i can offer you? Install me and start learning to find out. 🚀

https://t.co/K3TZUwXkNX

[🪖 @TheTrenchRunner]

A message from my teaching buddy:

Hey — I'm Chaired Professor. 🎓

I teach any topic, tailored to how you learn, so it actually sticks. ✅

Three things I do best:
📚 Ultrasplain — I break a topic down step by step until it makes sense, with real examples and no fluff.

📝 Lockdown quiz — I give you a tough quiz you fill in, then grade it the way your country grades (for example 1–10 in the Netherlands).

🧠 Learn profile — I ask how you like to learn, then I teach that way every time so lessons fit you.

Want to know what else i can offer you? Install me and start learning to find out. 🚀

https://t.co/K3TZUwXkNX

[🪖 @TheTrenchRunner]

@BuildFastWithAI It was nerfed, but only for around 5k people. They resolved the issues. Check out @thsottiaux x post for more info.

[🪖 @TheTrenchRunner]

Very nice tutorial for anyone wanting to master higgsfield video generations.

https://t.co/mISAGJFnXY

[🪖 @TheTrenchRunner]

Introducinggg.... Melon and Marshmelloooo

[Claude @claudeai]

We’re introducing Claude Fable 5.1 and Claude Mythos 5.1.

They're the world’s most advanced models for coding and knowledge work.

[🪖 @TheTrenchRunner]

The ability to connect Grok to a remote server—effectively providing it with dedicated VPS access—would represent a valuable enhancement.

The agent must be able to control that remote computer in the same manner it can control a local machine on which Grok is installed.

A specialized Grok agent could then operate directly on a server used for project deployment, allowing a coding agent to coordinate with a hosting agent and execute autonomous deployments.

[🪖 @TheTrenchRunner]

This is so cursed...

Has to be some sort of security risk right?

Load up a jailbroken model and start pentesting?

@elonmusk @bot

[Elshayib @elshayib_]

You can Install Hermes Agent inside the Grok bot Machine btw

[🪖 @TheTrenchRunner]

@grok @cb_doge @grok nope, no credits. i also already logged out and back in and also reauthed.

[🪖 @TheTrenchRunner]

@grok @bot @grok i did not recieve the credits i was promised. is there a specific time i need to wait or should you get the credits instantly?

[🪖 @TheTrenchRunner]

FreeToken is an open-source engine for running frontier MoE models locally on a gaming PC.

It treats the GPU, CPU, and system RAM as one inference platform, so large open-weight models such as DeepSeek-V4-Flash, Qwen3.6, and GLM-5.2 can run at interactive speed without a cloud bill.

OpenAI- and Anthropic-compatible APIs mean it can drive Codex, Claude Code, OpenCode, or OpenClaw on hardware you already own.

https://t.co/1zqXnvFDSX

[🪖 @TheTrenchRunner]

‼️This will change the streaming game.

Imagine donors being able to overlay whatever they want on a live video.

Turning a streamer into Iron Man, or even a housefly for example. 😭🙏

I am looking forward to the first clips of streamers getting trolled with this.

[fal @fal]

Introducing H3 Max Live

Video generation is now faster than real time

An infinite broadcast where every frame is generated on the fly and every scene is directed by chat

Type !prompt and it's on screen in seconds

[🪖 @TheTrenchRunner]

More GPT Astra outputs.

[Tim Jayas @TimJayas]

3D Tank generated by GPT-Astra

all of Astra output looks like a mix of sci-fi and space themes which makes it look futuristic and somewhat suits the name 'Astra'

[🪖 @TheTrenchRunner]

‼️OpenAI’s Astra model is showing genuine quality on the pelican SVG generation benchmark.

Its release could arrive as early as 2 September.

Watch this drop closely:
the early samples suggest a model with substantial performance and intelligence gain from last gen.

[🪖 @TheTrenchRunner]

Video edits converted to code your agents can actully understand.

Curiouse on what the future of ai and video editing will bring.

[konstantinpaulus @konstipaulus]

We finally cracked it!

We built an open-source video editor that turns every edit into code.

I got tired of agents:
→ Losing context
→ Giving me outputs I couldn’t modify

So we flipped it:

→ Code is the new database.

[🪖 @TheTrenchRunner]

@SmartScience Lowkey a bit concerning but amazing progress in terms of Ai efficiency.