Nah, iOS file system is nerfed. You’ll never be able to do any kind of serious professional work as long as that’s the case.
COMPUTING
-
GPU Industry Explained: Complete Overview and Analysis
By
–
This one explains the entire GPU industry.
-
From Concept to Market: Why Technology Adoption Takes Decades
By
–
Possible to build doesn’t mean ready to use. It took 30+ years to go from this watch computer in 1984 to the Apple Watch in 2015.
-
Epic Quest: GPU Return Journey with Veo3
By
–
a movie about eliezer and liron traveling over half the world, to throw the gpu of power back into the tsmc, while being supported by jaan and max from up high; brought to you by veo3
-
AI Programming Languages: Optimizing Code for Model Exploration
By
–
It may be time to develop AI programming languages. Code generation must be optimized for guiding models in exploring solution space and ensuring correctness, not for human comprehension. Code specification must optimize synchronization between human intention and AI
-
Language Models Memory Capacity Scaling Laws Physics
By
–
Physics of Language Models: Part 3.3, Knowledge Capacity Scaling Laws very interesting (if somewhat controversial) they declare that models, after visiting data 1000 times during training, can memorize 2 bits/param (they arrive at this number via quantization with AutoGPTQ)
-
Neural Network Memorization Capacity Under Parameter Quantization
By
–
Memorization Capacity of Deep Neural Networks under Parameter Quantization in 2019 these folks measured the capacity of RNNs/CNNs/MLPs, also using this class-based entropy idea they quantized the networks to half-precision and saw a capacity drop (but not nearly by half)
-
RNN Capacity: 5 Bits Per Parameter Storage Empirical Analysis
By
–
Capacity and Trainability in Recurrent Neural Networks this very nice paper from 2016 (!) estimates certain types of RNNs can store 5 bits/param. they find this empirically and measure bits using the entropy of the distribution of class labels
-
AlphaEvolve Deployed at Google: Compute Optimization Advances
By
–
Already in production AlphaEvolve has been deployed inside Google to: – Recover ~0.7% of global compute via better data center scheduling
– Rewrite Verilog hardware code for TPUs
– Speed up Gemini’s matrix math kernels by 23%
– Optimize GPU-level FlashAttention code with 32.5% -
Google AlphaEvolve: Gemini Agent Designs New Algorithms
By
–
🚨 BREAKING: Google just dropped something the world isn’t ready for.
— Futurepedia – Learn to Leverage AI (@futurepedia_io) 1 juin 2025
It’s called AlphaEvolve a Gemini-powered agent that designs entirely new algorithms.
It’s already optimizing chips, data centers, and cracking open math problems.
Here’s what it means: pic.twitter.com/Dcx8DCx3CKBREAKING: Google just dropped something the world isn’t ready for. It’s called AlphaEvolve a Gemini-powered agent that designs entirely new algorithms. It’s already optimizing chips, data centers, and cracking open math problems. Here’s what it means: