Robots can do amazing things, and they have been around for a very long time. If you walk through almost any manufacturing facility, you will see tons of work being completed by various types of robots. Most people aren’t too worried about those robots, however, because they look like large specialized machines, which aren’t usually seen as a threat to human jobs.

While this is odd because those types of machines have been replacing humans for decades now, it is just one of those little oddities about the way people see the world. When you see the new video put out by robotics company Figure, however, you may agree that the threat is real.

Months ago, Figure put out a video of its Figure 1.0 robot, which is a humanoid robot that is able to pick up and sort packages. It was able to grab packages that were a wide range of different sizes, weights, hardnesses, and shapes, and then send them down a conveyor belt. The robot would position the package so that the shipping label was positioned correctly for it to be read by scanners, a job that is currently done by humans at companies like UPS, FedEx, USPS, and more.

Recently, the company put out a new video of its updated Figure 02 bot, and it has gotten even better. This robot uses an advanced visual language system that they call Helix to better identify and sort the package. In addition to the boxes that the first version could do, Figure 02 can also sort bagged packages and flat envelopes. This is an especailly significant step forward because bagged packages are not just very common, but they also become easily deformed, which makes them harder for robots to manipulate.

Not only did the company add on these new abilities, but it is also able to do it faster and much more accurately. The company claims that this robot is now able to process packages with the correct orientation 95% of the time, which is a 25% improvement from the previous version. It also processes packages in about four seconds, which is down from five seconds on the older version. Check out the video they posted on X showing this robot in action:

Watch Helix's neural network do 60 minutes of uninterrupted logistics work

Helix now incorporates touch and short-term memory and it's performance continuously improves over time pic.twitter.com/DvfBe9IdGH

— Figure (@Figure_robot) June 7, 2025

Yes, that video shows the robot working for a full hour uninterrupted. While you might not want to watch the whole thing, it is amazing to see how it completes the task. One thing to take note of is the way the robot flattens out the bagged packages to ensure the barcode is able to be scanned down the line. This is an action that it was not programmed specifically to do, but rather it learned how to do through a demonstration. The company’s post about the video explains:

“This subtle ‘flattening’ action, learned from demonstrations, ensures the barcode is fully visible to the scanner. Such adaptive behavior highlights the advantage of end-to-end learning — the robot learns from demonstration strategies that were never explicitly hard-coded, directly from the data, to overcome real-world imperfections in packaging.”

Using AI to learn and improve its processes has allowed this version of the robot to keep getting better and better at its job. While there are still limitations from the hardware itself, these can easily be overcome with upgrades or by replacing the physical robot with a new version.

Some people will watch this video and think that a human can still do it faster, which is likely true. The fact that the robot is learning how to do it accurately now, however, means that it will undoubtedly be able to do it more quickly in the future. In addition, once deployed, these types of robots can work 24/7 without breaks, which is something humans cannot do.

For now, it does not make economic sense to try to replace humans with these robots for this specific task. As they improve and become cheaper to make, however, that will almost certainly change. Sorting packages, like almost all repetitive tasks done by humans, will be something that robots take over in the next few years, whether people like it or not.

If you thought that was interesting, you might like to read about why we should be worried about the leak in the bottom of the ocean.