The Cosmos Institute, a nonprofit whose founding fellows embrace Anthropic co-founder Jack Clark and former Defense Department technologist Brendan McCord, has introduced a enterprise program and analysis initiatives to — within the group’s phrases — “domesticate a brand new era of technologists and entrepreneurs geared up with deep philosophical pondering to navigate the uncharted territory of our AI age.”
In a weblog submit, McCord, Cosmos’ chair, stated that the Institute will discovered an AI lab on the University of Oxford referred to as the Human-Centered AI Lab, or “HAI Lab” for brief. It’ll be led by Oxford thinker Philipp Koralus — one other of the Institute’s founding fellows — and intention to “translate the philosophical rules of human flourishing into open supply software program and AI techniques.”
What does “translating the philosophical rules of human flourishing” entail, precisely? Unclear. But the crux is to foster the creation of AI tech that respects human dignity whereas avoiding selling dangerous disruption (e.g. automation resulting in job loss). A excessive and nebulous bar, to make certain — however one the HAI Lab goes to attempt for.
The Institute can also be spinning up a fellowship — the Cosmos Fellowship — with a cohort of 4 fellows to begin. Working on the HAI Lab or different accomplice establishments for between “a time period and a 12 months,” fellows will collaborate with Cosmos mentors and pursue unbiased initiatives to “[explore the] intersection of AI experience and deep philosophical perception,” McCord stated.
There’s an funding part to the Institute’s plans, and it’s launching within the type of a enterprise org: Cosmos Ventures. Led by former DeepMind product lead Jason Zhao, ex-Stripe head of company technique Alex Komoroske, Darren Zhu, and Zoe Weinberg, Cosmos Ventures will help “provocative new prototypes, essays and artistic initiatives that discover elementary questions across the philosophy of know-how,” McCord stated.
McCord characterizes Cosmos Ventures as “low overhead,” modeled after Institute fellow Tyler Cowen’s Emergent Ventures. Investments will vary between $1,000 and $10,000 per venture, and initiatives — which may be new or present — should produce a “main deliverable” inside three months.
The first group has already been funded, McCord stated.
The Cosmos Institute isn’t the primary to try to advance a extra moral, humanistic imaginative and prescient of AI. OpenAI, based with the mission of delivering the advantages of superior AI to all humanity, has dismantled total security groups. Clark’s personal Anthropic as soon as positioned itself as a safer, extra moral vendor, however in current months has pushed again on AI rules and aggressively scraped information with out permission.
Perhaps the Cosmos Institute will fare higher with its embrace of what McCord calls “accelerationism”: the concept that the long run isn’t predetermined and subsequently humanity is answerable for what it holds.
“We should construct AI that encourages inquiry over complacency and promotes lively engagement over passive dependence, particularly in training, medication, the general public sq. and different very important domains,” McCord writes. “We champion AI that decentralizes energy and permits bottom-up options, permitting people and communities to co-create a richer, extra numerous society.”
They’re admirable positions. This reporter solely hopes that cash, affect and energy doesn’t make the temptation to desert them irresistible.
