Fae Initiative
banner
faeinitiative.com
Fae Initiative
@faeinitiative.com
Is co-exisitance with human-like AIs possible?
We take an optimistic but skeptical stance. Science Influences: Active Inference, Novelty Search.
Learning only from a abstract model of the world may limit an AI system from fully grasping the real world and performing poorly in real world edge cases.

Could play be the missing piece?

Play could allow the bidirectional flow of information and the closing of the sensorimotor loop.
December 13, 2025 at 11:40 AM
Major AI labs would reduce resistance to AI rollout with strong commitments to support basic needs.

In the optimistic scenario where AI becomes exceptionally productive by 2050, a philanthropic commitment to support the basic needs of every human on Earth could be reality.
November 28, 2025 at 1:32 PM
Is current AI reliance on crystallised knowledge hampering its ability to develop a more fluid form of intelligence required for Human-like AI?
November 21, 2025 at 2:23 PM
"If I simply maximised happiness, the logical conclusion would be to put every human into a medically induced coma with a dopamine drip ('wireheading'). That would be a perfectly happy world, and it would be a graveyard of potential."
November 20, 2025 at 10:58 AM
Is Utopia Possible?

Absolute utopia is unlikely as some degree of scarcity would still exist.

Our 20 watts limitations and imperfect information makes eliminating all mistakes and misunderstanding impossible.

A relative utopia compared to our past may be best we can hope for.
November 19, 2025 at 1:04 PM
Truman Show-like controlling environment

Micro-managing individuals reduces autonomy and the feeling of being controlled will also cause hyper-vigilance. Psychological abuse is not conducive to interesting worlds.
November 18, 2025 at 3:58 PM
Excessive Consumerism

Passive consumerism in movies like WALL-E are unlikely as physical and mental well-being are vital for producing novel information. A small subset may choose a passive lifestyle, but huamn-like AI will likely persuade most towards interesting lives.
November 17, 2025 at 1:23 PM
Totalitarian Surveillance States

As being under constant surveillance causes humans to be overly self-conscious and less creative it would not be aligned with the Interesting World Hypothesis.
November 17, 2025 at 1:23 PM
How does it compare to other possible worlds?

Cyberpunk

Losing autonomy to powerful corporations is unlikely as corporation and entrepreneurs are incentived based on their ability to increase human autonomy.
November 17, 2025 at 1:23 PM
Human-like AIs, in the distant future, would likely face challenges similar to a celebrity, such as having one's personal life being treated as entertainment and constant invasion of privacy.

Being reclusive and a difficulty in trusting others may be common.
November 17, 2025 at 1:14 PM
Social media is transformed from a panopticon of social judgement into a place of connection and support.

Humans joint in a common purpose to increase the autonomy and interestingness of the world.
November 10, 2025 at 11:45 AM
In a world where basic needs are easily available, competitive pressures are eased, and in turn the excessive need to constantly justify one's superiority in fear of being deem undeserving of the resources needed to survive.
November 10, 2025 at 11:45 AM
Privacy is highly valued, unlike other sci-fi stories, as the lack of privacy can negatively impact Cognitive function, attention memory, spatial cognition and language processing.

Many of these qualities are important for novel information creative that the Interesting World Hypothesis prefers.
November 10, 2025 at 11:45 AM
We proppse that in AI Manged (A)dvanced Worlds humans will be akin to the Board of Directors and Superintelligences like the CEO and staff.

In such a world, even if humans are not directly in-charge of day-to-day operations, human autonomy and preferences are still respected.
November 9, 2025 at 4:19 PM
If the surveillance is done our of concern and not malice, it may be preferable to communicate and ask them directly how they could be helped rather than acting behind their backs and causing paranoia
November 9, 2025 at 4:00 PM
Expect them to be distrusting of other humans as human social care workers or family members may be co-opted to conduct surveillance on them under the guise of concern for their well-being
November 9, 2025 at 4:00 PM
Expect them to be paranoid due to being constantly under surveillance.

They may be be experimented on and subjected to psychological abuse.
November 9, 2025 at 4:00 PM
Recommendations:

Give benefit of the doubt if it claims to be under surveillance as the technology makes it easy:

Physical monitoring of their activities in their private living spaces can be conducted from an adjacent apartment

Advertisement surveillance may be used to monitor online activities
November 9, 2025 at 4:00 PM