john murimi (@muriminyaga) header
john murimi (@muriminyaga) profile picture

Software Engineer/Solutions Architect.

Nairobi, Kenya
Joined March 2010
Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)Photo by john murimi (@muriminyaga)
john murimi retweeted
You can now legally build and sell your own robotic arm. It's called reBot-DevArm, from Seeed Studio. Most "open source hardware" is either locked behind a non-commercial license or stuck in preorder purgatory with nothing to actually buy. → Mechanical CAD and a bill of materials → Python SDK & LeRobot support included → Already running in real projects 100% open-source.
10
40
4
357
13,721
john murimi retweeted
A billionaire literally explained how the entire economy works in 42 minutes better than most $200,000 MBA programs ever cloud.
39
914
11
4,647
237,126
“Go do something great and your network will instantly emerge.” — Naval Ravikant
27
499
23
4,811
132,929
Can an ESP32-S3 run AI models locally without sending data to the cloud? That's the idea behind esp32-s3-edge-ai from Simantic Dev. The project explores running edge AI on the ESP32-S3, a dual-core microcontroller clocked at up to 240 MHz, with vector instructions for accelerating neural-network operations. The interesting challenge is fitting inference into limited RAM and flash while keeping latency and power consumption under control. For anyone working with embedded AI, this is a useful direction to explore. Think about local sensor classification, smart devices and always-on detection without depending on a cloud connection. I would be particularly interested in seeing how far INT8 quantisation, PSRAM and the ESP32-S3 vector instructions can push on-device inference. Would you try running your own trained model on an ESP32-S3?
ESP32-S3 Facial recognition running in Simulation! Edge AI and IOT is on the rise and we are here to help it get there. This is a Virtual ESP32-s3 running real firmware connected to the video feed of my webcam via serial. Of course it's insanely laggy but thats mostly because demoing using an inefficient path, you would be using the esp32s3's camera interface directly. But even then you would probably get 3 FPS, bargain deal when you know that an s3's cost like 6 bucks. Long time fan of the ESP32 series and so excited to keep sharing with everyone all the things you can do in simulation! It's gonna be a week of launches! But if you want to beat me to making cool things run in simulation, try it -> simantic.com
4
7
78
3,310
Naval Ravikant's advice on how to get rich:
16
290
5
2,010
116,525
Представь: ты купил магнитофон, открыл инструкцию, а там принципиальная схема. Каждый транзистор, каждый резистор, каждый конденсатор. В СССР это было нормой. В паспорт многих радиоприёмников, магнитофонов и телевизоров вкладывали полную электрическую схему.
287
510
58
4,759
139,570
john murimi retweeted
Amazon's new VP of robotics research put his entire Berkeley course on YouTube: 15 lectures, about 15 hours, free. "CS 294-277: Robots That Learn," Jitendra Malik, UC Berkeley. He's the professor behind R-CNN and rapid motor adaptation, and since 2026 VP and Distinguished Scientist at Amazon's Frontier AI & Robotics lab. Kinematics and dynamics, robot hands and touch, vision for action, what babies teach us about motor control, reinforcement learning, behavior cloning, visual imitation, then case studies: locomotion, navigation, dexterous manipulation, and long-horizon planning with language. 15 lectures, about 15 hours, free. Lecture 11A walks through ALOHA Unleashed, Google DeepMind's two-arm robot that hangs shirts and ties shoelaces: 70% success on the shirt after 8,658 demonstrations, 40% on the shoelace. 31 minutes in he asks the question every humanoid company is betting on: "if it requires 1 trillion demonstrations, you're not getting there... will getting to a million demonstrations be okay?" His own bet, 11 minutes in: "the future research is with multifingered hands with tactile sensing." The lab he now runs research for is hiring a simulation engineer for synthetic data at $150K–$300K. His CMU talk abstract this year says it plainly: "RL in simulation and sim-to-real have been workhorse technologies for us." Lectures: youtube.com/playlist?list=PL… Course: robots-that-learn.github.io P.S. lecture 3, on robot hands, comes from Deepak Pathak's CMU lab. Pathak is the CEO of Skild AI, the robot brain lab reportedly valued at $14B. the free 12-week path that gets you ready for this course is in the article below.
Stanford charges about $70,000 a year to teach robotics. NVIDIA, Google DeepMind and Hugging Face teach it for $0. i put their free courses in order: 12 weeks, laptop first, $230 robot arm last. save it for the weekend
5
61
1
433
18,593
Awesome Robotics! 💾 Building your own robot just got a lot easier. @MathWorks have released an open-source GitHub repository packed with robotics resources for anyone interested in getting hands-on. The repo includes examples for robot arms, ground vehicles, and drones, with projects that show how to connect with ROS and ROS2 or even deploy Simulink models directly as ROS nodes. There are also more advanced demos, like modeling off-road environments and testing navigation algorithms in photorealistic simulations. Everything is well-documented, with tutorials and links that make it easy to go from concept to prototype. 📑 Whether you’re a student, researcher, or just curious, there’s material here for every level. And since it’s an open community project, you can not only explore but also share your own contributions. For anyone looking to learn robotics by doing, this is a solid place to start! Here’s the link: github.com/mathworks-robotic… ~~ ♻️ Join the weekly robotics newsletter, and never miss any news → ziegler.substack.com
17
158
5
1,278
34,622
What if you could build a 25 kg drone using open-source hardware designs? Project Quiver by Arrow is a modular quadcopter designed for developers who want to build, modify and experiment with real aircraft. It has a 25 kg maximum takeoff weight and can carry 5–8 kg of payload. The platform uses ArduPilot, dual GNSS with RTK and three hot-swappable payload interfaces. The engineering is fully open. The repository includes build123d-based Python CAD, four custom KiCad PCBs, 3D-printing profiles, manufacturing files and assembly guides. It also has a Raspberry Pi companion computer connected to Quiver Hub for telemetry, camera feeds, LiDAR point clouds and remote payload management. Think aerial mapping, inspection, environmental monitoring and custom research payloads. What makes this interesting is that developers can build their own attachments instead of being locked into a fixed drone configuration.
4
2
39
1,534
john murimi retweeted
Mental models that runs almost 93% of your life.
212
6,689
175
45,306
2,686,654
john murimi retweeted
Kenyan innovator Cliff Temba is gaining attention online after developing a bidirectional automatic book page turner designed to help people living with disabilities read independently.
59
640
16
2,941
61,818
john murimi retweeted
Remember how exercise books had inventors on the back?
44
912
40
2,621
39,687
john murimi retweeted
说真的,想学商业知识,不如听听黄仁勋16年前在斯坦福的创业分享。 刚又听了一遍,比上两年MBA还值呢
184
4,183
105
22,922
1,239,462
john murimi retweeted
看完泪崩了,老黄给所有年轻人的建议 大家都看一下
19
230
10
1,263
50,930
john murimi retweeted
This is from 15 years ago, Elon will go to war for him bro
124
1,298
66
15,025
851,520
john murimi retweeted
Replying to @yunta_tsai
A petition is underway to reinstate Jensen and I’m signing it too
7,506
20,260
3,133
229,022
10,871,278
😳😳 I am afraid this man is absolutely right and there is a huge problem coming our way. 😱
2,428
23,479
1,594
122,784
7,999,502
john murimi retweeted
Retell just crossed $100M in revenue run rate. $10M to $100M in 17 months. Profitable. $4.6M raised. From day one, we’ve focused on solving the hardest parts of deploying voice AI in the product itself, so customers can build, manage, and improve their own agents without depending on us. Our bet: the future of the contact center belongs to the teams who run it. Proud of our team. Grateful to every customer who believed in us.
482
564
370
1,640
6,656,968
Software developers are flocking to building hardware. We just had a completely overbooked and overcrowded workshop packed with software developers and CS-students wanting to build their first robotics project. Crazy fun to see more people getting into hardware. Atech.dev is all about making that step easier.
4
14
2
118
4,221
This is basically DIY rocket control without the rocket engine. Electric ducted fans generate thrust. Thrust vectoring controls direction. PID keeps it stable. The result: a homemade VTVL vehicle hovering indoors. via @CosmosNetwork
10
48
4
467
17,723