Me, Myself & I

Technology Evangelist. I spread the good news that
innovative #technology has the potential to make
the world a better place #innovation #robot #loT
#Al

[Ronald van Loon @Ronald_vanLoon]

Meet the Flying Lifebuoy That Rushes to Water Emergencies Before Rescuers Arrive
by @sutoroveli_news

#EmergingTech #Technology #Innovation #Tech #TechForGood

[Me, Myself & I @Makkaryh]

@tokifyi Just don’t take forever to make it.

[Me, Myself & I @Makkaryh]

@wobsoriano @Muse @alexandr_wang @natfriedman What about a Google home mini

[Ronald van Loon @Ronald_vanLoon]

These DIY #AugmentedReality Glasses Cost Only $8
by @HiddenPlaces25

#VR #EmergingTech #Innovation #Technology #FutureTech

[ProxySoul @BniWael]

Empryo spawning hundreds of opus 5.5 & luna 6 to do surgical fixes.

With Empryo's efficiency, subagents seem to work pretty nicely!

You can spawn up to 10k subagents with different models & providers in Empryo in parallel...

[atomic.chat @atomic_chat_hq]

Local Laya moggs Jev at @grok 4.7-built Tetris 🧩

An open-weights System One model called Laya, beat cloud-based Jev at playing Tetris by making decisions 11 times faster, running locally on a 16GB MacBook Air!

Run AI models locally -> https://t.co/RbcCOIgVkj

[小墨同学 @xiaomovps]

Pi+Jev 终于解决了长时间不响应的问题🔥

使用 Jev 模型 判断任务和时长，到底是真的程序在运行 还是 Agent 出问题，自动终止任务和重试

整体的执行流程和判断逻辑：

1、监控 Pi 是否持续产生模型输出、工具进度或生命周期事件。

2、如果持续有进度，判断为正常运行，继续等待。

3、如果长时间没有进度，调用 TypeSafe 判断是正常慢任务、需要用户处理，还是疑似卡住。

4、TypeSafe 判断正常时继续等待；判断不确定时提醒用户。

5、判断为卡住时终止当前请求，并自动重试一次。

6、重试仍失败时停止自动操作，提示用户切换模型、检查网络或手动处理。

7、如果 Pi 进程本身死锁，则交给外部系统级 Watchdog 重启

我相信很多用 PI 的人都遇到过，PI 在执行长时间任务的时候没有任何反应，只有终止当前任务才能继续

我在社区也看到了相关的解决方案，可以使用插件去监测时间消耗，我自己的处理方案是把对应的等待时间都调小，但是只是治标不治本，后面我就尝试使用 Jev 模型去做判断，没想到真可以

玩得越来越有意思了 我要继续挖掘它新的能力！

[Rik Arends @rikarends]

51 million lines of code. Chromium codebase. Here you go. 120hz, 3gb ram use.

[Suraj Sharma @suraj_sharma14]

If you learn inference infra, you will never be unemployed again.

Stage 1: Systems Foundations
C/C++ or Rust basics, Python performance, Linux, memory hierarchy, CPU vs GPU execution.

Stage 2: GPU Architecture Fundamentals
CUDA cores, tensor cores, VRAM, memory bandwidth, SM occupancy, nvidia-smi profiling.

Stage 3: Transformer Inference Anatomy
Prefill vs decode phases, KV cache growth, attention memory math, arithmetic intensity.

Stage 4: Serving Engines
vLLM, SGLang, TGI: continuous batching, PagedAttention, scheduling, throughput vs latency.

Stage 5: KV Cache Optimization
Prefix caching, chunked prefill, KV quantization, cache offloading, multi-turn reuse.

Stage 6: Quantization and Compression
FP8, INT8, AWQ, GPTQ, sparsity; measuring quality drop against speed gains.

Stage 7: Speculative Decoding and Parallelism
Draft-target models, Medusa heads, tensor vs pipeline parallelism tradeoffs.

Stage 8: Kernel-Level Optimization
Triton kernels, FlashAttention, CUDA graphs, operator fusion, Nsight profiling.

Stage 9: Distributed Inference
Multi-node serving, NVLink and InfiniBand, disaggregated prefill/decode, MoE expert parallelism.

Stage 10: Autoscaling and Economics
Queue-based scaling with KEDA, cold starts, spot GPUs, cost per token, MFU utilization.

Stage 11: Gateway, Routing, and Reliability
AI gateways, fallback chains, rate limits, TTFT and ITL SLOs, observability, chaos testing.

Stage 12: Portfolio and Benchmarks
Self-hosted cluster, published latency/throughput/cost benchmarks, kernel optimization writeups.

Most people stay stuck watching tutorials.

Builders get hired.

(Bookmark it)

[CTO ROBOTICS Media @ctorobotics]

PID Control: The Hidden Algorithm Behind Robotics. 🤖⚙️

You may not see it, but PID control is everywhere in robotics.

From robot arms and mobile robots to motors and industrial automation, PID helps machines maintain precise motion, speed and position.

P — reacts to the current error
I — corrects accumulated error
D — responds to the rate of change

No AI. No machine learning.

Just control theory working behind the scenes to make robots move accurately and consistently.

If you could master ONE control algorithm in robotics, would you choose PID? 👇

🎥 Media: medcorreia ( Instagram )

⚠️ This content is shared for informational purposes only. CTO Robotics Media is a media platform and does not own or develop the technology shown. Credit belongs to the original creators.

#PID #PIDControl #Robotics #RobotProgramming #ControlSystems #IndustrialAutomation #Automation #Mechatronics #RoboticsEngineering #Engineering #CTORoboticsMedia

[Jonathan Roomer @jonathanroomer]

I’ve got Codex (voice) in CarPlay. And it’s fantastic.

I can now build things during road trips, while all 5 kids make a noise and my wife asks why I talk to the AI more than I talk to her.

It runs through my Nightblood iOS app, which gives ChatGPT Voice a face, personality and full access to Codex, my Mac and all the native tools.

[Sadao Tokuyama @tokufxug]

Wi-Fi電波をARで可視化する「SPECS」📡👓

🔹Gabriele Romagnoli氏が紹介
🔹空間内のWi-Fi電波強度を可視化し、電波の強い場所を確認
🔹Wi-Fiルーターの設置場所や接続ポイントの検討に活用

スマートグラスで「見えない電波」を可視化する実用AR。情報元はリプ欄。

[Sarah Drasner @sarah_edo]

I've been working on this app to keep myself organized, but WebMCP makes it pretty killer.

🔥 See how I can seamlessly create a workback gantt chart just from some board items?

[Geek Lite @QingQ77]

为移动端项目提供能直接构建并实时预览 iOS/Android 运行效果的开源 AI 编程助手，省去手动切换模拟器、复制日志的折腾。

https://t.co/h0buHH4YZQ

[Henry Li @henry19840301]

After seeing GPT‑6 Astra crank out CAD everywhere, I put Codex in a loop. And here's the result https://t.co/F6KVfGibJ6

Are we cooked? @M5Stack @EspressifSystem @its_hard_2_name

[sui @birdabo]

fun fact. arduino and raspberry pi are part of china’s school curriculum.

they later teach themselves circuit design via youtube videos. seems almost every crack chinese engineers are self taught.

china is just built different.

[Ramin Nasibov @RaminNasibov]

wake up!

[Simplifying AI @simplifyinAI]

if you've wanted into robotics but stalled on where to start, this is the on-ramp.

opencat is the framework behind petoi's bittle and nybble.

it runs on arduino and raspberry pi, so the barrier is a hobbyist budget, not a lab.

it walks, self-balances, reacts to touch, and rights itself after a fall, and the gait code is all there to read and modify.

start with the kit, then start rewriting how it moves.

[JT @trev3d]

Impromptu testing in the park. Hoping to have Lasertag 3.0 out soon™

[Jason Lee @huacnlee]

I have to show off how insanely fast GPUI Component runs with WASM on the web.

You may not even realize this is built with GPUI + WASM.

This DataTable demo has 1 million rows — one of GPUI Component's performance highlights.

https://t.co/9eKMCWB5bv