X Square just beat Figure’s package-sorting pace by ~45%—without a humanoid.
@XRoboHub22K followers
86,393 views
26 Comments
Comments from the original post on X (Twitter)

@XRoboHub
Now this is real package sorting
9 likes

@JeremyVanDervor
Faster doesn’t always mean better. Figure’s humanoid can just walk off the line and get replaced by another one in seconds if it has a problem. X Square has to shut the whole line down to fix a arm.
8 likes

@XRoboHub
If the robot has time to walk off the line and wait for its replacement to walk in, I think we’ve got time to swap an arm too 😂
9 likes

@JeremyVanDervor
That's my point. The humanoid is automated and can be replace itself through automation. The arm swap takes a tech to go out and swap the arm. Unless you have a figure robot that replaces the arm for you. Whoever your version is a lot cheaper to implement.
3 likes

@ChrisGPT
Now let’s see X square be able to operate a forklift and climb a ladder to go fix something! Oh.. wait
4 likes

@XRoboHub
Haha, not every job needs a humanoid. The people paying for them care more about ROI😂
2 likes

@asing2061
In China, robots and robotic arms are used at the same time.
2 likes

@sblackwell87
No one cares. Humanoids can do more than sort packages.
2 likes

@toptraders0x
form factor is the easy part. the moat is fleet uptime and deployment data, not legs vs arms.
1 like

@BlackApple
Yea but Brett has interns shoot him "sleeping" in his office, that's gotta be worth $40B right?
1 like

@bobthebobnes
Figure is 100% silicon valley techbro fundraising bs. It's no different than 95% of the crypto "lets solve an issue that doesnt exist" nonsense. Real engineers go for efficiency. Humanoids in factories is deliberate inefficiency
1 like

@DamirWallener
Meanwhile, in actual sorting facilities, automation looks nothing like this...
1 like

@dennis_nik75
The point is to create a humanoid that can do anything from parcel sorting to surgery. Not to buy several seemingly cheaper hand robotic systems, each needing their own special non humanoid training. The majority of the training data exist for humanoid forms.
1 like

@JohnVega14
This is just such an awful take. Give anyone in the space, a weekend and good training data, they could build this exact same bimanual arms. This has been a thing for 15 years. What isn’t, is a movable system than can operate an entire production line

@LudwigBoltzman
Humanoids are expensive because they are general purpose and the cost spent today will yield results later on. In the end it will be worth it

@eko_sawijaya821
Now let's see x square be to operate a forklift and climb a ladder to go fix something

@sir4K_zen
pragmatic approach wins when the task is pure pick and place, legs and dexterous hands add cost without payoff

@clara_jane47268
Bots all over this comment section lmao 🤣 Why are they crying

@LeeMcClymont
It is a cool demo, not doing the same job though and is much less versitile for an employer. Definitely has uses though if they can be mass manufactured and trained quickly.

@DistrictCrypto
there is a difference between purpose build vs general use.... traditional sorters destroy both of these





