OpenAI and Figure's NEW Robot Stuns in Demo Video

This video comes from Figure, a company focused on bringing robots to market or as they like to call it, "Emboddied AI." From their X page:

"With OpenAI, Figure 01 can now have full conversations with people

-OpenAI models provide high-level visual and language intelligence
-Figure neural networks deliver fast, low-level, dexterous robot actions

Everything in this video is a neural network."

[byram]

Despite how clunky it was I feel like we should Mark this day on a calendar

[ephraimcullen]

"Cups and a plate" When it should be "Plates and a Cup"
This is very impressive, but still has a long way to go for any practical usage
Cool demo though!

[MuhammadAkhtar-tb]

He stuttered on "I " word at @1:47 which is pretty wild

[lejonzx]

The micro movements it makes at the end of every request is crazy. Seems to be a very human thing to do. Unless for some weird reason it was trained to do so or that micro movement is encouraged during its training.

[particleconfig.]

''only eh, edible... ''
so is it stuttering now?

[german.gorbachov]

Absolutely mind blowing! Especially when you can grasp the amount of work put into making it happen. One thing in question - how much power did it take to perform all the processing involved?

[KimSol]

Wow.... I don't even know what to say. Mind blowing!

[aliasgur]

C3PO but less nervous and not as overtly gay

[luluuk]

Is there someone that feels the movements feel rendered instead of real life footage? There is something about them, like everything is just too smooth.
I'm not judging the audio/speech. Maybe it's just me.

[MrCPPG]

I didn't hear it say "Would you like fries with that?".

[leocreuu]

I’m not convinced there’s no dude in the back talking into a mic

[HansPeter-lxqk]

iRobot coming soon to your cinem...real life!

[Ron-kpow]

Notice that it said "cups and a plate" when it was really a cup and plit's.

[polar]

It’s stuttering makes me think it’s wayyy too human to be an actual AI. And it’s movement seems so fluid I feel like this was CGI or sumn

[Crimsonoverlord.]

cant tell if theirs a guy in the back with an x box controller and a mic..

[proezakful]

I do see one glaring issue though . 30 seconds into it, When asked to describe what the AI sees on the table, it looks at the drying rack and says" I see a plate (singular) and some cups (plural) .This means that it inverted it's comprehension. There are in fact, multiple plates and a singular cup instead of, as it stated, a singular plate and multiple cups. That means it's understanding of what it sees in the drying rack is inverted. It thinks it sees multiple cups and a single plate which means it understands the cup to be a plate and the plates to be cups. This is a horrible, horrible mistake. This could go horribly wrong in many different ways and practical applications. And I think the fact that they overlook this one glaring mistake. It shows that they are getting way ahead of themselves.
I also am disturbed that the producers of this video didn't catch this.

[collinsanyanvoh]

No matter what don't give AI the ability to judge humans; especially when they can cause harm to humans

[dmitry]

This music and setting gives Boneworks vibe

[michaelpolioudakis]

automobile suspension design in the waist ....hmmmm smart....

[Mike-ffib]

Im not convinced. The setting is very staged. The only impressive part is the fluid movements. I would like to see a demonstration that shows the the same performance in a completely random setting, example: walks into a room never seen before.