Discussion about this post

User's avatar
Ian Brown's avatar

I'd like to see a Left vision of AI articulated with more specificity.

There do seem to be some innate dangers with the technology, some easy traps as well that need to be managed, and inherent power struggle over computation resources.

The innate dangers are obvious, creating systems that supersede human intelligence creates great power that can't be controlled. Combined with advanced robotics, this technology could become autonomous enough as to supersede human beings as the dominant form of life (maybe in the sense that viruses are alive), which is quite profound and could put us in the role of wild or domesticated animals vis humans today. Capitalists are willing to take the potentially existential risk in order to grasp at that power for themselves.

Also inherent environmental cost.

The traps of using it are, the risk of making people dumb, sloppifying human cultural output and communication, robbing it of intention and skill, or fabricating it altogether and destroying common reality, and robbing humanity of the dignity that comes with thinking, creating, and expressing for yourself.

And the power challenge is the same as drives the US AI arms race against China, more computational infrastructure (and pollution) gives advantage over systems with less computational power, which could be end-run, or even hacked. Capitalist/imperial or centralized systems, along with their client governments will see alternative systems with the same hostility as they see Chinese ones. And the whole dynamic drives a pattern of proliferation at the expense of the environment and local communities.

A sovereign Left/community/human oriented AI system would have to somehow have break out of the industrial scale hardware/ infrastructure supply-chain monopoly in the West. Chinese manufactured hardware is a possibility, but now illegal or de-facto illegal in Western countries and their client states.

So the questions would be CAN the global Left establish infrastructure, algorithmic, and robotic systems, that hold their own sovereignty against the militarized capitalist/imperial monopoly, and most importantly, can they do so while preserving human control, dignity, cognitive and cultural sovereignty, and without destroying communities and land in data center arms race? How can this be done, and what does it look like?

No posts

Ready for more?