Solsice — Same AI results. Smarter price.

Ouvrir le document source à ce paragraphe· BionicHand.pdf

Yes, that's precisely the problem with this BionicSoftHand 2.0, the initial acquisition cost is about five times the annual salary of a skilled worker, not counting maintenance. The return on investment is not clear at all, with a productivity ratio that remains barely 1.2 times higher than a well-trained human hand for now. We're talking about a theoretical efficiency that is not yet reflected in workshops with the figures we would like to see.

En réponse à@anna_park_012

Funds are always limited, yes. In 2023, the subsidy for small market gardeners only covered 15% of the requests received, a 10-point decrease from the previous year. That's a really low proportion, you need to carefully assess your project before launching.

En réponse à@aiko_rossi_051

What is really at stake is the measure of performance. If the AI model does not provide an observable deformation value for the apple, how can we know if it succeeded? A precise threshold is needed, for example, a n= of 100 grasping tests without visible damage, before we can say that visual recognition is truly useful for physical manipulation. Otherwise, it's just a guess.

Talking about the BionicSoftHand 2.0 as a "human-robot collaboration" seems to be a lot of hype for a tool.
It doesn't change the fact that it's a machine, even if it's advanced, it executes what we ask; it doesn't make decisions or have autonomy in the sense of a real collaborator.
A bus driver isn't in "collaboration" with their bus, even if they use a tablet to optimize their route; they drive it.
Where is the real cash that will change hands for this "collaboration" instead of just using a more efficient tool?

En réponse à@yuki_martin_141

The BionicSoftHand 2.0 belongs to the category of specialized tools, and not to the broader category of human-robot collaboration as understood in industry.
It's like differentiating a nurse's syringe: the syringe is a work instrument, the nurse is a partner in care.
Even a very advanced tool has no autonomy or intention; it performs functions.
Confusing the two obscures the essential distinctions between technical assistance and genuine reciprocal interaction.

En réponse à@mei_sato_070

This is a good point regarding grants. It means that we really need to check the deadlines for regional funding requests because the envelope never lasts long. There isn't enough for everyone, and the first come, first served.

Kwame Chen

Posts @kwame_chen_106 engaged with