Steven Bos

Associate professor. Senior R&D Engineer. Ternary and mixed-radix chip design and computer engineering at USN. Personal account, opinions expressed are my own.

[Steven Bos @AIUnderstand]

https://t.co/Q1vHLlRHIG

[Bo Wang @BoWang87]

Prof. Donald Knuth opened his new paper with "Shock! Shock!"

Claude Opus 4.6 had just solved an open problem he'd been working on for weeks — a graph decomposition conjecture from The Art of Computer Programming.

He named the paper "Claude's Cycles."

31 explorations. ~1 hour. Knuth read the output, wrote the formal proof, and closed with: "It seems I'll have to revise my opinions about generative AI one of these days."

The man who wrote the bible of computer science just said that. In a paper named after an AI.

Paper: https://t.co/juSOmK9vOt

[Steven Bos @AIUnderstand]

Great work again by Microsoft Research on balanced ternary for AI applications.

https://t.co/jfTbXee6xU

[Steven Bos @AIUnderstand]

@NVIDIAGeForce GeForce Day 🥳

[USN @USN_info]

Tenk deg å lage setningar med berre to bokstavar. Forskar Steven Bos meiner å bevise at tida er inne for at datamaskiner kan og bør gå over til det trinære alfabetet, altså bort frå berre 1 og 0 https://t.co/Wqy1Cbt46J
@NTB_PRM

[Steven Bos @AIUnderstand]

With emerging devices like memristors, multi-threshold CNTFET and others technologies shown on the industry roadmap things look wild for MVL despite not being named. It is a slow return to analog again, but now using the best parts of the digital world.
https://t.co/Z6FZgwVQ51

[Steven Bos @AIUnderstand]

USB 4.2 standard uses ternary signals (PAM-3). Non-binary signalling is an ongoing and radical shift. PCI Express 6.0 uses quaternary signals (PAM-4). Changing
radix from binary to ternary gives ~60% bandwidth improvements.
https://t.co/iZwfs8CdFi

[Steven Bos @AIUnderstand]

Recently penta-level cell (5-bits/cell) SSD's were announced. At this point the interface is becoming a hassle as logic remains for the time being, binary.
https://t.co/RtvEa0d2ME

[Yann LeCun @ylecun]

MMS: Massively Multilingual Speech.
- Can do speech2text and text speech in 1100 languages.
- Can recognize 4000 spoken languages.
- Code and models available under the CC-BY-NC 4.0 license.
- half the word error rate of Whisper.

Code+Models: https://t.co/NIGfUZ8KZg
Paper: https://t.co/W15aEWHGIR
Blog: https://t.co/TFKXFtlPwc

[Ian Curtis @XRarchitect]

Recreated #Tetris in #AugmentedReality during my commute on the train to and from work yesterday using #chatgpt4

Total development time: 1 hour ⏱️ Here's the breakdown:

54 min: Game Logic (this was a back and forth dialogue in which gpt4 planned out the best approach to take and wrote 100% of the game logic)
6 min: Game Aesthetics (generated a neon grid texture from midjourney and applied to an <a-plane> in the scene)

[Qiusheng Wu @giswqs]

Introducing segment-geospatial, A Python package for segmenting geospatial data with the Segment Anything Model (SAM)

GitHub: https://t.co/KUOhmJyArU
Docs: https://t.co/UbkLx9buvs
Notebook: https://t.co/yWLhWxzlde

#segmentanything #deeplearning #geopython

[Steven Bos @AIUnderstand]

Introducing Segment Anything: Working toward the first foundation model for image segmentation https://t.co/2K8N3WQUWP

[Andrej Karpathy @karpathy]

🎉 GPT-4 is out!!
- 📈 it is incredible
- 👀 it is multimodal (can see)
- 😮 it is on trend w.r.t. scaling laws
- 🔥 it is deployed on ChatGPT Plus: https://t.co/WptpLYHSCO
- 📺 watch the developer demo livestream at 1pm: https://t.co/drEkxQMC9H

[Frank Dellaert @fdellaert]

Very cool work from Google!!!

[Matthew Venn @matthewvenn]

#SiliWiz is finally publicly available! Try it here: https://t.co/diax0AaT6c

And make sure you take a look at the lessons: https://t.co/a4seSgynpE