AI • Engineer & Researcher • Automating everything Collabs | DM open

Joined August 2020
Pinned Tweet
finally! local AI is accessible to everyone. we replaced our massive smart home server with a single $8 ESP32-S3 chip processing voice commands entirely on-device. now anyone can build this. here is the full guide:
49
90
31
896
1,817,068
you don't even have to be in the same room with the right Wi-Fi setup, someone walking behind brick walls can still leave a measurable trace at the receiver the graph uses real Wallhack1.8k recordings: orange is walking, blue is an empty space, both on the same scale. the 3D scene illustrates the idea if your next thought is “i'll just sit still,” watch the quoted post😁
breathing is enough to give you away each breath slightly changes how Wi-Fi waves reflect off your chest. with the right receiver and processing, those tiny changes can reveal your breathing rhythm the graphs in this animation use real WiRe-Breath recordings: orange is the rhythm extracted from Wi-Fi, gray is a reference breathing sensor i broke down how this works, and what else you can pull out of the radio waves around you, in the article below ↓
27
147
12
1,176
79,706
how does a car hang upside down, let go, and still land on its wheels? watch Tesla’s Roadster on the curved ramp underneath, fans pull air out of the gap between the car and the surface. seals help keep that air from rushing straight back in the result works like a suction cup: the air outside presses the car against the ramp. Tesla says the grip is strong enough to hold it upside down even while standing still once it leaves the ramp, a different system takes over the car carries compressed gas and releases it through small nozzles. think of letting go of an inflated balloon: air shoots one way, the balloon goes the other with nozzles around the car, those pushes can also turn it. push one way to start rotating, then counter that turn to bring the wheels back underneath before landing, downward jets push upward on the car to slow its fall. Tesla says the system makes tiny, rapid adjustments to keep it stable there’s a catch: every burst uses up some of the stored gas. how long it can stay airborne depends on how much it carries and how quickly it spends it
Real life, baby
32
153
29
725
275,729
ard retweeted
there's a catch in that breathing demo the room was quiet the same Wi-Fi measurements also respond to other movement. those changes can bury the much smaller rhythm of breathing in these real WiRe-Breath recordings, our simple estimate follows the reference sensor in the quiet case, but loses agreement under motion interference orange is the Wi-Fi estimate. gray is the reference sensor compare it with the quoted demo: same processing, different room conditions
breathing is enough to give you away each breath slightly changes how Wi-Fi waves reflect off your chest. with the right receiver and processing, those tiny changes can reveal your breathing rhythm the graphs in this animation use real WiRe-Breath recordings: orange is the rhythm extracted from Wi-Fi, gray is a reference breathing sensor i broke down how this works, and what else you can pull out of the radio waves around you, in the article below ↓
3
10
1
93
81,866
breathing is enough to give you away each breath slightly changes how Wi-Fi waves reflect off your chest. with the right receiver and processing, those tiny changes can reveal your breathing rhythm the graphs in this animation use real WiRe-Breath recordings: orange is the rhythm extracted from Wi-Fi, gray is a reference breathing sensor i broke down how this works, and what else you can pull out of the radio waves around you, in the article below ↓
78
603
143
5,681
1,531,470
there's a catch in that breathing demo the room was quiet the same Wi-Fi measurements also respond to other movement. those changes can bury the much smaller rhythm of breathing in these real WiRe-Breath recordings, our simple estimate follows the reference sensor in the quiet case, but loses agreement under motion interference orange is the Wi-Fi estimate. gray is the reference sensor compare it with the quoted demo: same processing, different room conditions
2
158
77,753
there's a catch in that breathing demo the room was quiet the same Wi-Fi measurements also respond to other movement. those changes can bury the much smaller rhythm of breathing in these real WiRe-Breath recordings, our simple estimate follows the reference sensor in the quiet case, but loses agreement under motion interference orange is the Wi-Fi estimate. gray is the reference sensor compare it with the quoted demo: same processing, different room conditions
breathing is enough to give you away each breath slightly changes how Wi-Fi waves reflect off your chest. with the right receiver and processing, those tiny changes can reveal your breathing rhythm the graphs in this animation use real WiRe-Breath recordings: orange is the rhythm extracted from Wi-Fi, gray is a reference breathing sensor i broke down how this works, and what else you can pull out of the radio waves around you, in the article below ↓
3
10
1
93
81,866
/anatomy explain how a vibrating structure inside a microchip measures rotation
Introducing Anatomy Describe a machine in one line. Claude brings it to life. - Interactive isometric wireframes you can rotate, inspect and explore - WebGL shaders for fire, water, plasma and more - Real 3D motion, with individual parts that rotate and move Install it as an agent skill. Fully open source.
3
28
1,501
i still can’t wrap my head around this Voyager 1 was 24.9 billion kilometers away, and we could still pick up its signal on Earth in December 2024, the 25-meter Dwingeloo radio telescope detected its faint radio carrier that tiny peak in the noise is the signal i wrote about this and other things hiding in the radio waves around us in the article below
2
1
1
18
1,589
your phone can find a clock hidden in noise that’s part of how GPS puts you on a map the receiver searches for a satellite’s known pattern, locks onto the signal and reads the timing data inside it this animation shows the idea using a real recording by Daniel Estévez. we decoded the time field that becomes 11:32:30 in the video combine timing measurements from multiple satellites, and you can work out a position GPS is just one example. the article below explores what else is hiding in the radio waves around you, and how to start decoding it yourself
2
7
823
i sat through 4 years of painfully boring university lectures to put together a radio guide you can read this weekend Wi-Fi, GPS, FPV, radar, and experiments you can start without buying hardware:
7
538
Wi-Fi is overrated sometimes your device only needs to send a few bytes. making it depend on a nearby router can be a bigger limitation than the amount of data it needs to send pair an ESP32 with a LoRa radio, and it can send small messages directly to another node beyond typical Wi-Fi range no router, SIM card, or internet connection needed between them a remote sensor could report rising water. a weather station could send a reading. a rover could report its battery level the tradeoff is speed: think short updates, with range depending on antennas, terrain, and radio settings i made this 3D animation to show the path from sensor to receiver a first experiment: two nodes, one button, one message appearing on the other end starter code in the reply
15
34
3
345
15,213
sounds crazy, but... what if we built a local "Jev" for ESP32-class devices? the idea: a ready-trained decision model you could reuse across physical projects, feeding it sound and sensor data and getting structured outputs back imagine a lamp that reacts to supported sounds, or a toy that combines sound and motion to choose its next response run it on the main board if it fits the project's resources, or use a companion module i wrote up the architecture, limitations and possible path to a first prototype in a concept paper if you have ideas, use cases or suggestions, i'm open to them the PDF is on repo in reply
finally! local AI is accessible to everyone. we replaced our massive smart home server with a single $8 ESP32-S3 chip processing voice commands entirely on-device. now anyone can build this. here is the full guide:
3
9
549
i think this is the easiest way to learn to play the guitar
1
1
15
2,124
leave your phone outside the room. your body still changes the Wi-Fi inside as you move, you change how direct and reflected signals combine at the receiver an ESP32 can expose those changes through CSI i made this animation to explain the physics behind the graphs in the quoted post open it for the recorded-data demo, then find the browser explorer and two-board build guide in the repo linked in its replies
Wi-Fi sensing is getting uncomfortably accessible... walk across a room and you change how the signal reaches a receiver an ESP32 can expose those changes through CSI, channel-state information i built a browser interface to explore those measurements this clip uses published recordings from a walking session, and the build guide covers a two-board ESP32-C3 setup if you have ideas for what to build with this, i left the repo in the replies
4
5
2
30
2,501
Wi-Fi sensing is getting uncomfortably accessible... walk across a room and you change how the signal reaches a receiver an ESP32 can expose those changes through CSI, channel-state information i built a browser interface to explore those measurements this clip uses published recordings from a walking session, and the build guide covers a two-board ESP32-C3 setup if you have ideas for what to build with this, i left the repo in the replies
4
10
2
173
8,453