tag > ML
-
Machine Consciousness?
Every few weeks, some philosopher asks if machines can be conscious — as if that’s the big mystery. Meanwhile, we kill billions of sentient beings a year, turn them into lasagna, and still think awareness lives in a circuit board. The real question isn’t whether AI can wake up, it’s why humans never did. This isn’t philosophy; it’s performance art by a species barely conscious enough to keep its own biosphere alive. Intellectual cargo cult with tenure.
-
My 2027 Generative AI Media Prediction: The Great Convergence
Since 2017, my core prediction has remained unchanged. Now, as we approach 2027, the pieces are finally in place for it to become reality. Here it is:
By 2027, high-quality, real-time AI generation of all digital modalities—video, audio, text, and 3D—will hit early mainstream adoption. Major players will roll out the foundational platforms.
This technological leap will trigger a chain reaction, collapsing four distinct media forms (Video Games, Streaming Video & Film and the Creator Economy) into one new, dominant category.
What emerges won't be a simple hybrid. It will be a new media form: an immersive, hyper-customized, and interactive story-space. Imagine a narrative experience that is part blockbuster film, part open-world game, and part creator-driven universe—all uniquely tailored to you in real-time.
This convergence may start as a niche, but its growth will be explosive. It will rapidly eclipse all legacy media forms in scale, cultural impact, and economic value. The walls between playing, watching, and creating are about to vanish. Mark my words.
-
Schmidhubers warning about elite science fraud in AI are right, but..
Jürgen Schmidhuber’s persistent warnings about how the “elites” in AI play fishy & fraudulent games are both correct & necessary. But their behavior makes sense once you view it through the broader lens of How Power Manages Science and Technology, and how elite power structures not only monitor it, but may also shape, obscure, or re-route its development to serve long-term strategic dominance.
-
Introducing OscNet: A JAX library for oscillatory neural networks and dynamical systems.
OscNet provides a framework for building and training neural networks based on oscillatory dynamics — coupled oscillator networks, continuous-time neural networks, and general dynamical systems. Built on JAX and Equinox for differentiable, high-performance computation. https://github.com/samim23/oscnet
Features
- Oscillator models: Harmonic, Van der Pol, Stuart-Landau, Kuramoto, FitzHugh-Nagumo
- Coupling topologies: Hierarchical fractal, power-law, log-periodic
- Analysis tools: Edge-of-chaos, Floquet analysis, bifurcation, stability
- Training utilities: Criticality initialization, stochastic forcing, schedulers
- Visualization: Phase space, network dynamics, oscillator analysis
More on the projects background: In Resonance with Nature - Toward a New Kind of Machine Learning with Oscillatory Neural Networks
