3 ms·

The robot mostly looks like a very basic alternative to the Lego robotics offerings (mindstorms/technic/spike). And, they are in more or less the same price ran

by nico 1y ago

The robot mostly looks like a very basic alternative to the Lego robotics offerings (mindstorms/technic/spike). And, they are in more or less the same price range as well Not sure what huggingface is going for here. Seems like a big distraction for the company

garciasn 1y ago
I am not sure what HF is going for here either, because it doesn't actually do anything; yet? I mean, great: I have another $300 toy in the makerspace arena that I can program. Awesome; I write code and am heavily invested personally and professionally with LLMs (open-weight models running on local GPUs as well as LLM chatbots that everyone knows, like Gemini, OpenAI, and Claude). Now what? It'll sit on my desk and maybe gather dust, if it's not dancing, because...it doesn't actually have a point or built in capabilities out of the gate--at least not that I can see--aside from looking cute and dancing according to the video in the press-release-article; all things that any $30 toy on Amazon can already do.
- sheepscreek 1y ago
  Perhaps they needed to do something to justify raising another round?
Aurornis 1y ago
I don’t see the comparison to Mindstorms This is more like a Furby that you’re supposed to connect to your AI system’s audio/video interface so people can interact with a stationary humanoid device
alnwlsn 1y ago
Hardly. This seems like the opposite alternative, something that has minimal or no mechanical hardware provisions, and few ways to interface with external motors and sensors.
refulgentis 1y ago
I love huggingface and they've done a lot for the industry ...But... I'm so confused by the economics, and not in a general way. I'm a mobile developer by trade, and AFAICT from dipping my toe in the water of servers, file downloads are on the order of $0.01/GB, and cheapest you'll get publicly is maybe $0.04/GB. But they never ever charge and regularly have people downloading tens of gigabytes. That's the cost side as far as I understand it. On the revenue side...afaict all they have is inference? And they don't seem to be popular for an inference solution? I don't hear much about it and they don't seem to care too much. How do their economics even begin to pencil out? It bothers me, I've lived through enough companies to know this doesn't matter in the short term. But this is novel, to me, in that there's no plan, no market being addressed. With other money losers, you knew what they were trying to do and they were doing it.
- microtonal 1y ago
  This question comes up quite often, but Hugging Face is profitable: https://xcancel.com/ClementDelangue/status/1811675386368966682 https://xcancel.com/ClementDelangue/status/18116753863689666... There is a variety of income sources, including inference and Hub Pro and enterprise subscriptions [1]. [1] https://huggingface.co/pricing https://huggingface.co/pricing

by nico 1y ago

garciasn 1y ago

I am not sure what HF is going for here either, because it doesn't actually do anything; yet? I mean, great: I have another $300 toy in the makerspace arena that I can program. Awesome; I write code and am heavily invested personally and professionally with LLMs (open-weight models running on local GPUs as well as LLM chatbots that everyone knows, like Gemini, OpenAI, and Claude). Now what? It'll sit on my desk and maybe gather dust, if it's not dancing, because...it doesn't actually have a point or built in capabilities out of the gate--at least not that I can see--aside from looking cute and dancing according to the video in the press-release-article; all things that any $30 toy on Amazon can already do.

sheepscreek 1y ago
Perhaps they needed to do something to justify raising another round?

Aurornis 1y ago

I don’t see the comparison to Mindstorms This is more like a Furby that you’re supposed to connect to your AI system’s audio/video interface so people can interact with a stationary humanoid device

alnwlsn 1y ago

Hardly. This seems like the opposite alternative, something that has minimal or no mechanical hardware provisions, and few ways to interface with external motors and sensors.

refulgentis 1y ago

I love huggingface and they've done a lot for the industry ...But... I'm so confused by the economics, and not in a general way. I'm a mobile developer by trade, and AFAICT from dipping my toe in the water of servers, file downloads are on the order of $0.01/GB, and cheapest you'll get publicly is maybe $0.04/GB. But they never ever charge and regularly have people downloading tens of gigabytes. That's the cost side as far as I understand it. On the revenue side...afaict all they have is inference? And they don't seem to be popular for an inference solution? I don't hear much about it and they don't seem to care too much. How do their economics even begin to pencil out? It bothers me, I've lived through enough companies to know this doesn't matter in the short term. But this is novel, to me, in that there's no plan, no market being addressed. With other money losers, you knew what they were trying to do and they were doing it.

microtonal 1y ago
This question comes up quite often, but Hugging Face is profitable: https://xcancel.com/ClementDelangue/status/1811675386368966682 https://xcancel.com/ClementDelangue/status/18116753863689666... There is a variety of income sources, including inference and Hub Pro and enterprise subscriptions [1]. [1] https://huggingface.co/pricing https://huggingface.co/pricing