375ai (@375ai

375ai

1,571 posts

375ai

@375ai_

A New Data Layer for the Physical World

San Francisco, CA

Joined June 2024

@375ai_

See 375ai’s full profile

Open X

Pinned
375ai
@375ai_
Apr 7
AI models are trained on roughly 20 trillion tokens, or every public text on the internet. and it's still not enough. The next frontier isn't more text, It's the physical world, and 375ai is already capturing it.
00:00
20
30
118
14K
375ai
@375ai_
15h
The physical world was digitized in fragments, dash cams here, satellites there, all locked in silos but never as continuous, calibrated coverage a model can actually train on. That’s what we’re building. Physical AI doesn’t catch up until the real world is machine readable.
will depue
@willdepue
Jul 6
A Stargate for Data Labs are on a trajectory towards >$100B/year of data spend by 2030. As we begin the trillion-dollar compute project, we need to think about the equivalent civilizational-scale effort for the other core ingredient: data. At the foundation of the scaling
1
3
32
4.1K
375ai
@375ai_
Jul 7
AI labs are in an arms race for video, paying creators $1–4/min for footage that was never even published. Text is tapped but real world data isn't. Most video footage isn't grounded. It lacks the essential context of when and where it was shot, and what it's actually seeing.
Made with AI
4
4
39
4K
375ai
@375ai_
Jul 6
Crude oil is useless until it’s refined. The same goes for physical world data. 375ai refines raw street level video into fuel for world models. Every frame labeled and enriched on device. AI already consumed the internet. The real world is next.
1
12
38
4.6K
375ai
@375ai_
Jul 4
Built in America Deployed across America Happy 4th from all of us at 375ai 🇺🇸
Made with AI
2
6
37
4.2K
375ai
@375ai_
Jul 2
LLMs train on 20 trillion tokens of text. A child takes in the same amount of information visually by the time they're 4 years old. Intelligence starts in the physical world. That's what we're building at 375ai.
Made with AI
5
6
47
4.2K
375ai
@375ai_
Jul 1
Physical AI needs real-world data, because a model that acts in the world can only learn by observing it directly. That’s the data layer we’re building, to train the next generation of world models.
00:00
4
7
47
5K
375ai
@375ai_
Jun 30
375edge gives us the ability to understand the flow of traffic in real time, almost like reading a pulse.
2
4
35
4.3K
375ai
@375ai_
Jun 28
Models have run out of internet to learn from. The next frontier is the physical world. Every week we turn 21,000+ hours of real world activity into labeled, sensor-enriched data, continuous observation of how the physical world actually moves.
00:00
7
11
90
10K
375ai
@375ai_
Jun 26
🔸Real world data is the new oil🔸
6
9
39
4.9K
375ai
@375ai_
Jun 24
Same intersection, three conditions. Every object labeled on-device: pedestrians, vehicles, signals, even umbrellas. Physical AI can't learn from simulation alone, it needs continuous observation of the real world. 375edge captures reality as it is, not as it's imagined.
3
3
36
4.6K
375ai
@375ai_
Jun 22
The emerging consensus in physical AI is that world models can't train on synthetic data alone, they need immense quantities of real world data. 375ai produces 21,000+ hours of labeled, sensor enriched video every week and scaling quickly.
00:00
11
5
56
9.7K
375ai
@375ai_
Jun 22
Read this article from @DataconomyMedia for more on why world models need to be trained on real world data.
Synthetic Data Alone Cannot Train Physical AI to Handle the Real World
From dataconomy.com
2
9
527
375ai
@375ai_
Jun 19
They're saying it We're building it
00:00
5
6
47
2.3K
375ai
@375ai_
Jun 18
The physical world is coming online across NYC metro one highway and street corner at a time.
1
3
31
1.3K