3,694,720 trajectories and 138,623 users. These numbers make me start to see Axis Robotics from a different perspective. Physical AI doesn't just need better models. For robots to truly learn how to interact with the world, they need a large amount of data on actions, movements, and how to handle each specific task. This is where Axis's model becomes interesting. Users can participate in simulated tasks such as controlling a robotic arm to move or arranging objects. Each execution contributes to creating a trajectory, which is data on how an action is completed. According to the data currently displayed on Axis Hub: • 3,694,720 trajectories • 138,623 users • 330 new tasks in 7 days • over 3.4 million transactions In my opinion, the most notable thing is not the number of transactions, but the more than 3.69 million trajectories that have been created. If data is truly one of the major bottlenecks of Physical AI, then the ability to mobilize hundreds of thousands of people to participate in data creation could become a very significant advantage for @axisrobotics. The more tasks robots want to accomplish in the real world, the more data they need to learn. And Axis is trying to build exactly that layer of data.

728,335 trajectories from only 4,913 participants. This is the number that caught my attention when researching the collaboration between Axis Robotics and BitRobot. What I find interesting is not just the quantity, but how they transform human activity in a simulated environment into data serving Physical AI. Users participate in simulation tasks and control robotic arms, thereby generating trajectories. The data is then quality-assessed and verified on-chain on Solana. Currently, the system in the image has recorded 601 tasks and over 728K trajectories. In my opinion, this is a quite notable approach to the data problem in robotics. Instead of relying entirely on physical robots to collect data, which is costly and hard to scale, Axis is leveraging simulation and community participation to generate data on a large scale. AI has a massive dataset from the Internet to learn from. But Physical AI, which wants to understand and interact with the real world, needs a different kind of data—data about actions and movements. And I think this is exactly the problem that @axisrobotics is trying to solve. If this model continues to expand, the data could become one of the most important “resources” in the Physical AI race.

Notable news: Authorities will establish a mechanism to quickly block and freeze bank accounts and e-wallets suspected of being involved in online fraud activities, while also intensifying asset recovery efforts for victims. If effectively implemented, this will be a positive step forward in reducing fraud cases and increasing the chances of recovering money for victims. Hopefully, in the near future, fraudsters will no longer be able to easily liquidate assets as before.

Oh hey, brother Huấn Brothers, have you heard anything yet?

Oh my god, common people are giving gifts to billionaires? Everyone, come join the drama

Today I just finished paying off this loan. Actually, this is not a loan I took out for spending, but I stood in to support a friend in borrowing because I didn't want to lend money directly. He has fallen into debt many times and even got involved in gambling. I hope this loan will create some pressure for him to appreciate money, be more responsible with life, and be determined to start over. Sometimes helping someone is not about giving them money, but about helping them have the motivation to change. I just hope this time will be the last lesson.

So it's just a rumor, right guys? Telegram still hasn't been removed from the App Store

So it's just a rumor, right guys? Telegram still hasn't been removed from the App Store