Related posts

Creator of the Keras deep-learning library and the ARC-AGI benchmark.

• Our readingkorrents.com

  In verifiable domains model capability scaling should remain unbounded, because the space of enumerable patterns is infinite by construction.

  Their words

  In verifiable domains, model capability scaling should remain unbounded. Models will simply keep improving by "absorbing more and more of the computational universe", which is infinite by construction.

  @fchollet on Xx.com

Chief Scientist at Google DeepMind and Google Research, at Google since 1999, where he co-created MapReduce, Bigtable, TensorFlow and the TPU.

• Our readingkorrents.com

  Inference, not training, is now the constraint, and specialized low-energy hardware will beat general-purpose GPUs and TPUs on latency.

  energy

  Their words

  Yeah, I mean it's a little different, but I think uh you're going to see more and more uh uh high performance and um low energy uh inference hardware systems because I think everyone is now realizing that inference is the key to making you know these agent-based systems be available to more and more people and that latency is really important and that specialization of the hardware is a really key way you can make uh things that are more energy efficient and lower latency than more general purpose uh computational devices like say GPUs or TPUs

  Jeff Dean: The 1% Rule for Building in AIyoutube.com

Writer and consultant on formal methods and software correctness; author of Practical TLA+ and Logic for Programmers.

• Our readingkorrents.com

  Formal methods shine in highly computational domains and struggle wherever the problem is embedded in human business behaviour.

  Their words

  I think the case of TLA+ and most, not all, but most formal methods, they shine the most in highly computational domains, where most of the problems are highly technical and not like business embedded.

  Formal methods with Hillel Wayneyoutube.com

Freelance climate and energy analyst in Oslo, writing on Europe, the US and Australia; worked in the Australian renewables industry before that.

• Our readingkorrents.com

  Generative AI's climate harm comes mainly from wasteful system design and forced-on features, not from the computational cost of individual queries.

  Google

  Their words

  What is happening is that GenAI's climate harm comes from the system design, such as Google's AI overview triggering constantly whether users want it or not, in addition to significantly worse digital bloat tools dominating overconsumption.

  ketanjoshi.co

Writes about the biotech industry, drug development and forecasting — “I write to learn, mostly about biotech.”

• Our readingkorrents.com

  A computational discovery platform does not make a biotech fast, because the thousand other steps of drug development are still the bottleneck.

  Their words

  Another challenge for Class II biotechs is that even if you have a nifty computational drug discovery platform, you are still often bottlenecked by the rest of the drug development process. It takes 1,000 (or more!) small steps to advance a drug through discovery, development, and past the regulators.

  A new breed of biotechatelfo.github.io

Co-founder and CEO of Google DeepMind, and a 2024 Nobel laureate in Chemistry for the protein-structure work behind AlphaFold.

From one piece Highly accurate protein structure prediction with AlphaFold (with 33 co-authors) 2 beliefs, in the piece's order there

• Our readingkorrents.com

  A computational method can regularly predict protein structures with atomic accuracy even when no similar structure is known.

  Their words

  Here we provide the first computational method that can regularly predict protein structures with atomic accuracy even in cases in which no similar structure is known.

  doi.org

• Our readingkorrents.com

  Accurate computational approaches are needed to close the protein structural coverage gap and enable large-scale structural bioinformatics.

  Their words

  Accurate computational approaches are needed to address this gap and to enable large-scale structural bioinformatics.

  doi.org

Associate Professor at MIT CSAIL, leading the Programming Systems Group; works on programming systems that manipulate uncertainty, from neural networks to unreliable hardware.

• Our readingkorrents.com

  Pruning cuts the parameters of a trained network by over 90% without costing accuracy, so inference gets smaller and faster for free.

  Their words

  Neural network pruning techniques can reduce the parameter counts of trained networks by over 90%, decreasing storage requirements and improving computational performance of inference without compromising accuracy.

  arxiv.org

Chief AI Scientist at Databricks, after its acquisition of MosaicML, where he was a founding team member; known for the Lottery Ticket Hypothesis from his MIT PhD (ICLR 2019 Best Paper).

• Our readingkorrents.com

  Pruning cuts the parameters of a trained network by over 90% without costing accuracy, so inference gets smaller and faster for free.

  Their words

  Neural network pruning techniques can reduce the parameter counts of trained networks by over 90%, decreasing storage requirements and improving computational performance of inference without compromising accuracy.

  arxiv.org

Programmer and writer on computer architecture, performance, and software reliability. He has worked on CPU design at Centaur Technology and on software at Google and Microsoft, and writes long-form technical essays at danluu.com.

• Dislikedaffiliate linkrcmnd.app

  An Introduction to Computational Learning Theory

  Their words

  Classic, but dated and riddled with errors, with no errata available.

  danluu.com

Creator and first user of Slashdot, which he ran for nearly 14 years; later worked at the Washington Post's WaPo Labs. Known online as CmdrTaco.

• Usesrcmnd.app

  MacBook Pro

  Their words

  I'm currently using a 17" 2.3 ghz Core i7 MacBook Pro laptop for pretty much all of my computational needs.

  usesthis.com