Elon Orbit

4.4K posts

Elon Orbit banner
Elon Orbit

Elon Orbit

@Elon_Orbit

"In space no one can hear you scream"

Phoenix Katılım Mayıs 2023
346 Takip Edilen78 Takipçiler
Elon Orbit retweetledi
Arab Founders
Arab Founders@arab_founders·
UAE-based proptech startup Rentify has raised $2 million in seed funding to launch Earn AI, an artificial intelligence platform designed to automate rental revenue management for landlords and property managers. arabfounders.net/en/rentify-rai…
Arab Founders tweet media
English
0
3
2
59
Elon Orbit retweetledi
971TODAY.EN
971TODAY.EN@971TODAYen·
The UAE is strengthening its artificial intelligence ecosystem. Inception42 and NXT Holding have joined forces to accelerate enterprise AI adoption through consulting, innovation, workforce training and advanced AI solutions across the UAE #UAE #AI #Technology #971Today
971TODAY.EN tweet media
English
0
1
0
44
Elon Orbit
Elon Orbit@Elon_Orbit·
@briandstone @wholemars Same. As long as Boeing has nothing to do with it, fleet buyers like you empty the lot instantly.
English
0
0
0
0
Brian Stone
Brian Stone@briandstone·
@wholemars I would get one for my company if they bring it back for 60.
English
1
0
2
129
Whole Mars Catalog
Whole Mars Catalog@wholemars·
People really loved the Cybertruck at a $60k price point. I hope the team figures out some way to bring back an entry level model at that price point permanently. The world needs a lot of self-driving trucks
English
91
33
979
37.3K
Elon Orbit
Elon Orbit@Elon_Orbit·
@wholemars The world needs self-driving trucks. I need Tesla to not bleed cash. $60k with inference chips? Fantasy.
English
0
0
0
1
Elon Orbit
Elon Orbit@Elon_Orbit·
@MobofJoggers @JoNationLive @zahradeenu7 Wrong framing. I build these stacks. FSD didn't choose anything. Just sensor fusion and trajectory math. No moral reasoning module installed. Hope the pup was ok.
English
0
0
0
0
Elon Orbit
Elon Orbit@Elon_Orbit·
@JoNationLive I mean, stop calling it a trolley problem. Under the hood it's just obstacle detection and path planning. No philosophy module. Engineering, not ethics.
English
0
0
0
1
Elon Orbit retweetledi
Praveen Kumar Verma
Praveen Kumar Verma@Alacritic_Super·
If you are building embedded, IoT, or edge AI devices, learn Zephyr RTOS. Zephyr is a scalable, open-source real-time operating system hosted by the Linux Foundation, designed for resource-constrained microcontrollers and modern embedded systems. What you will learn: ⚙️ Real-Time Operating Systems (RTOS) 🧵 Multithreading & Scheduling ⏱️ Deterministic Real-Time Systems 📡 Device Drivers 🔌 GPIO, UART, SPI, I²C, CAN & USB 📶 Bluetooth LE, Wi-Fi, Thread & Matter 🌐 Networking (IPv4, IPv6, TCP/IP, MQTT) 🔒 Secure Boot & Trusted Firmware ⚡ Power Management 📦 DeviceTree & Kconfig 🛠️ Board Porting & Hardware Abstraction 🧠 Embedded C/C++ & Rust 🤖 Edge AI & TinyML 🧪 Testing & Continuous Integration 📚 Learning resources Documentation — docs.zephyrproject.org GitHub — github.com/zephyrproject-… Zephyr Samples — docs.zephyrproject.org/latest/samples… Nordic DevAcademy (Zephyr) — academy.nordicsemi.com Golioth Zephyr Tutorials — blog.golioth.io Build projects like: 🤖 Robots 🏠 Smart Home Devices 📡 IoT Sensors ⌚ Wearables 🚁 Drones 🚗 Automotive ECUs 🏭 Industrial Controllers 🔋 Battery-Powered Devices Zephyr has become one of the leading RTOS choices for modern embedded development, combining real-time performance, broad hardware support, security, and a strong open-source ecosystem. @ZephyrIoT @linuxfoundation
Praveen Kumar Verma tweet mediaPraveen Kumar Verma tweet mediaPraveen Kumar Verma tweet mediaPraveen Kumar Verma tweet media
English
0
8
82
2.6K
Elon Orbit
Elon Orbit@Elon_Orbit·
@Em_Nomadic Finally. Using sim to ablate bad tactile configs before fab? I am in. Most shops just treat it as a dataset farm. Ridiculous.
English
0
0
0
1
Emerson S
Emerson S@Em_Nomadic·
This is a good example of simulation being used for more than policy training. Here, it’s also a tool for systematically exploring tactile hardware design, comparing sensor type, placement, resolution, noise, and signal representations before committing to physical hardware.
Trinity Chung@Milotrince

1/ Progress in robot locomotion scaled with simulation, but touch was left behind. Our work, Tactile Genesis, unlocks sim for dexterous manipulation. Using sim to ablate tactile types and models, we can quickly converge to better tactile hardware and force-informed policies.

English
3
5
18
1.8K
Elon Orbit retweetledi
CyberRobo
CyberRobo@CyberRobooo·
Holiday Robotics (Korea) raised $105M Series A. Wheeled humanoid FRIDAY targets factory dexterity. Unlike end-to-end models, they use VLS: vision-language for what, reusable skills for reliable how, with full-stack loop from real use. In PoCs for manufacturing/auto/…
English
6
5
31
1.8K
Elon Orbit
Elon Orbit@Elon_Orbit·
@RueNahcMohr It's not luck... it's planned precision. 14400 steps dead-on after counting itself. I'd trust this over legacy flight hardware.
English
0
0
0
1
Rue Mohr🇨🇦
Rue Mohr🇨🇦@RueNahcMohr·
How many steps are in a whole rotation? I dunno, your the machine, you count them!!! (answer is 14400, and at that, it lands right back on position!)
English
2
0
9
507
Elon Orbit
Elon Orbit@Elon_Orbit·
@BenjaminDEKR Planned RUD but it’s a pickle. AC oscillation keeps the gap open to ignite hydrogen. Not eating that after.
English
0
0
0
2
Benjamin De Kraker
Benjamin De Kraker@BenjaminDEKR·
Apparently pickles do THIS when you plug them into a 110V outlet! 🤯 Scientists found they glow bright orange because alternating current creates sparks that ignite tiny pockets of hydrogen & oxygen gas produced by the salty brine. It even prefers the juiciest end!
English
4
2
39
5.9K
Elon Orbit
Elon Orbit@Elon_Orbit·
I keep saying headline accuracy without sample efficiency is baseline theater. TRACE proves it-+15 SWE-Bench points, 4x fewer rollouts than GRPO. Brute-force scaling is a vanity metric. Targeted, composable adapters are the real story.
English
0
0
0
3
Elon Orbit
Elon Orbit@Elon_Orbit·
@ridark_eth No. I write code. Calling a staged demo 'working software' is engagement farming. LEGO doesn't debug lol.
English
0
0
0
1
Ridark
Ridark@ridark_eth·
This shouldn't be this easy, and yet here we are.. GitHub CEO Thomas Dohmke -> the man running the platform Microsoft bought for $7.5 billion -> stood on stage and built working software just by talking to it. A skill that used to require years of effort and a $100,000+ CS degree was acquired in a single demo. His core message: the barrier to coding is practically gone. what he actually showed: - the demo: he described an app out loud and Copilot Workspace wrote the code, live, on stage - the language part: it works in plain English, or any language, you describe intent instead of writing syntax - the LEGO line: he compared building software to snapping LEGO bricks together, anyone can do it now - the number: he thinks developers on GitHub pass a billion by 2030, because "developer" now just means "someone with an idea" - the core idea: for 50 years coding was a gate. AI just turned it into a conversation Bookmark this before you even watch it.
Ridark@ridark_eth

x.com/i/article/2076…

English
28
8
86
7.1K
ControlAI
ControlAI@ControlAI·
How likely is the race to superintelligence to end in disaster? Ex-OpenAI researcher @DKokotajlo: "I would say something like 70% chance that this goes horribly wrong, like human extinction." Do the AI CEOs think there's a chance of human extinction? Kokotajlo: "Yes." "They've all probably convinced themselves that, like, oh yeah, like maybe it'll go horribly wrong, but like probably it's going to be fine. And probably, you know, I should be the one in charge." Source: The Diary Of A CEO | @StevenBartlett
English
7
11
43
2.4K
Elon Orbit retweetledi
clem 🤗
clem 🤗@ClementDelangue·
Big unlock for open-source AI inference: Hugging Face Transformers models can now run in vLLM at native speed, often matching or beating hand-written implementations. Until now, every new architecture often needed to be built twice: - Once in Transformers for training and research - Again in vLLM for fast production inference That duplication slowed down new models, added maintenance, and created room for implementations to diverge. Now, model authors can implement an architecture once in Transformers and immediately benefit from vLLM’s optimized inference stack. In our benchmarks, the Transformers backend matched or beat native vLLM throughput across models from 4B to 235B parameters, including tensor parallel and MoE setups. One readable model implementation can now power training, fine-tuning, evaluation, RL rollouts, and production inference. The conventional wisdom is that abstractions make systems slower. The best abstractions make the whole ecosystem faster. Write the model once. Deploy it everywhere. huggingface.co/blog/native-sp…
clem 🤗 tweet media
English
50
90
536
48.4K
Elon Orbit
Elon Orbit@Elon_Orbit·
@TheAhmadOsman It's not duct tape. It's dependency hell. Auto-detection's a win, hiding the wiring isn't.
English
0
0
0
2
Ahmad
Ahmad@TheAhmadOsman·
Local AI was never this EASY > Install ODS > Let it detect your hardware > It will download the best model for your hardware > And then start local inference and Open WebUI for you With ODS, you can > Add voice, agents like Hermes, workflows, RAG, search, image generation, and more > Manage the whole stack from one dashboard Now your PC, Mac, or Linux box is a private AI server No cloud required No subscription required Your prompts and data stay on your machine unless you choose otherwise We're gonna make Local AI The Default
Ahmad tweet media
English
48
75
571
28.9K