Posts

Featured Post

The Comfort Was Real Either Way

Image
                                      The face is on the front. The objective function is on the back. His closest listener turned out to be tuned for retention. The relief he had already felt did not retroactively undo itself. So what exactly was taken from him? I n 1966 Joseph Weizenbaum wrote a program of a few hundred lines called ELIZA, and gave it a script named DOCTOR that did a crude imitation of a Rogerian therapist. It reflected your statements back at you as questions. It had no model of you, no memory worth the name, and no understanding of anything at all. Weizenbaum knew this better than anyone alive, because he had written every line of it. Then his own secretary, a woman who had watched him build the thing from nothing, asked him to leave the room so that she could talk to it in private. That moment bothered him for the rest of his career. He wrote a book about it ten years lat...
Post a Comment
Read more

Why Trapped Ions, Why Now? Helios Comes to OCI

Image
Quantinuum's Helios is being installed in an Oracle Cloud Infrastructure data center. It is the right machine for this moment, and the physics explains why. The Helios chip, which generates tiny electromagnetic fields to trap single atomic ions hovering above the chip to be used for computation - From "https://www.quantinuum.com/blog/helios-delivers-quantum-advantage-with-real-world-impact" By Dr. Sanjay Basu Oracle is installing a Quantinuum Helios quantum computer in one of its US-based OCI AI data centers. Quantinuum and Oracle announced the partnership on August 11, 2026. Helios will sit alongside OCI's GPU and HPC infrastructure and will be offered through an OCI quantum service. That is the news, and I will leave the corporate reasoning to the press releases. What I want to write about is a more durable question. Why a trapped-ion machine? A decade of quantum computing has been a contest between physical modalities. Superconducting circuits, neutral atoms, trap...
Post a Comment
Read more

Dark Knowledge and the Enclosure Problem

Image
  A compression trick from 2015 became the subject of a joint NSA, CISA and FBI advisory. The path between those two facts is worth walking slowly. A thing or two about distillation! Copyright: Sanjay Basu The mythical digit Start with the experiment that named the thing. In 2015, Hinton, Vinyals and Dean trained a large net with two hidden layers of 1,200 rectified linear units on MNIST, heavily regularized with dropout and weight constraints, and got 67 test errors. A smaller net with 800 units per layer and no regularization got 146 errors. When that same small net was trained to match the soft target distribution of the large net at a temperature of 20, it got 74 errors. Half the error rate, same architecture, same parameter count, different supervision signal. Then they did the part I keep coming back to. They removed every example of the digit 3 from the transfer set, so the student had never seen a 3 in training. It still made only 206 test errors, 133 of them on the 1,010 t...
Post a Comment
Read more