Never underestimate the incredible ingenuity of evolution and nature. The cardinal sin of the increasingly arrogant machine learning community.
tag > ML
-
-
The peculiar rise of artificial intelligence for psyops and decline of it's governors real intelligence
It is remarkable how successful AI is for population scale Psyops & electronic warfare (if you ignore second order effects). As R.A. Wilson once remarked: "Reality is the temporary resultant of continuous struggles between rival gangs of programmers."
The stark contrast between the advancements in AI for Psyops and electronic warfare, and the dwindling intellect and decision-making capacity of those governing these systems is alarming. This wave of generational regression is a global concern.
Or as Napoleon Bonaparte put it: "Never interrupt your enemy when he is making a mistake."
-
Anticipate a future where machine learning models are 10 to 100 times more data/compute-efficient, enabling any dedicated group of individuals in their bedrooms to train massive, very capable models. It will super-charge the open-source revolution and disrupt everything.
-
I support the current thing
It is truly remarkable how successful AI is for population scale Psyops & electronic warfare (if you ignore second order effects).
As Robert Anton Wilson once remarked: "Reality is the temporary resultant of continuous struggles between rival gangs of programmers."
-
As Generative AI is making content creation at scale accessible to all, traditional moderation & censorship strategies stand to be overwhelmed. The tidal wave of generated content will make it very challenging for any authority to observe, orient, decide & act in effective ways.
-
A recent deepmind paper uses ML to come up with the most effective prompts for LLM's. "Take a deep breath and work step-by-step!" was one of the top ones.
-
Soft Wireheading Music
The perception of music — its appeal to an individual in specific contexts and the emotions it evokes — has traditionally been highly subjective. Though recommender systems have provided some insights in recent years, understanding this domain largely remained elusive. However, with the rise of contemporary machine learning and BCI, modeling audience's perception and responses to media is rapidly increasing in precision. By integrating these affective models with generative music models in a dynamic feedback loop, we get a glimpse of a groundbreaking new form of media, that enable users to deliberately and precisely alter their mental states. Such "soft wireheading" techniques have expansive applications, ranging from entertainment and healthcare to education and defense.
But more importantly:
In an age where our senses are perpetually bombarded by external stimuli, it's worth remembering this insight from meditation: It's entirely possible to cultivate an inner state of profound tranquility and bliss at any moment, independent of any 'external' catalyst.
-
The narrative surrounding AI music technology is currently often rather primitive: "Hey, I've created an AI that generates music automatically. Isn't that awesome? The end."
However, any seasoned music producer may question: "Where are the creative controls? How does this fit into my workflow? Is this just a toy or will AI replace me?"
We have the opportunity and responsibility to elevate the conversation and focus on how AI can augment music production, inspire new creative processes, and empower artists, fans and businesses.
-
Fear is the puppet master behind all the world's wicked manipulation & control schemes. But remember, you can't laugh and be afraid at the same time. Thanks to Computational Comedy, we now have the means to scale humor and turn the world into a comedy club.
-
A core assumption of neuro- and computer science is that computation is local & independent - that the brain is effectively separated from the 'outside world' by a thin box of bones. In reality, every 'box' is highly leaky and everything is deeply intertwingled.
