Google Unit DeepMind Tried–and Failed–To Win AI Autonomy from Parent

> DeepMind have been negotiating for years with the parent company for more autonomy, seeking an independent legal structure for the sensitive research they do.

> Its long-term goal is to build an advanced level of AI that can perform a range of cognitive tasks as well as any human. “Guided by safety and ethics, this invention could help society find answers to some of the world’s most pressing and fundamental scientific challenges,”

I'm trying to piece together what a different sort of 'legal structure' would mean to what DeepMind does or how they go about it. If the unit has autonomy in what it researches then the only thing would be to expand what it can choose from. If research areas are dictated, I don't know what impact a different legal structure can have. Also, up until now DeepMind has published in rather abstract problem spaces, board games and video games--though I guess that can extend to wargame simulations.

I've always enjoyed that Deepmind staff have a very distinct style from Google Brain. More than just affiliations on publications, there is a unique cultural divide that isn't entirely attributable to London vs. mountain view.

I wonder if this is a sign that inevitable Deepmind will become more and more like Google proper.

One of the major drawbacks of selling your company is it’s, well, no longer your company.