Hugging Face, a company that hosts open source artificial intelligence models and software, announced today that it has acquired Pollen Robotics, the French startup behind the bug-eyed, two-armed, humanoid robot called Reachy 2. Hugging Face plans to sell the robot and will also allow developers to download, modify, and suggest improvements to its code.
Since it’s already a Pandora’s box as it is, at least open-source is the least-worst way to go. All closed-source models are evil(er).
Open source is much less useful with AI though since you can neither inspect how it works exactly nor modify that one thing that bothers you about the current behavior.
That’s increasingly untrue as we build tools to inspect and modify those weights. Nearly every open model has homegrown LoRA support, and most of them have abliteration code that lets you erase undesirable behaviors.
And if we’re really talking about “open-source” and not just “open-weight”, the additional scientific papers, datasets and tooling are going to help democratize the technology and even out the playing field to a degree.
While it’s still somewhat of a blackbox, some definitions of “open source AI” are better than others
The OSI one is decent
It’s less useful because you can’t modify it to do whatever you want.
It’s less dangerous (than proprietary AI) because no one else can, either.
So the cyberpunk lore is true, we’ll have to flee our current internet for a new one soon, neat.
I’m not sure who I trust less with an army of AI robots, a capitalist private enterprise or the general public. Both sound problematic.
Definitely capitalist private enterprise is less trustworthy, and it’s not even close.
