@petedaguru.bsky.social
Computer nerd in all the nice ways, Plays well with others. #fedwiki
#AIML Nested Learning: The Illusion of Deep Learning Architectures - Paper: abehrouz.github.io/files/NL.pdf Blog: research.google/blog/introdu... - HOPE self-modifying Titans (Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni 2025-11-10): arxiviq.substack.com/p/nested-lea...
Nested Learning: The Illusion of Deep Learning Architectures
Authors: Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni
arxiviq.substack.com
November 30, 2025 at 4:50 PM
#AIML Nested Learning: The Illusion of Deep Learning Architectures - Paper: abehrouz.github.io/files/NL.pdf Blog: research.google/blog/introdu... - HOPE self-modifying Titans (Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni 2025-11-10): arxiviq.substack.com/p/nested-lea...
#AIML Nested Learning: The Illusion of Deep Learning Architectures - Paper: abehrouz.github.io/files/NL.pdf Blog: research.google/blog/introdu... - HOPE self-modifying Titans (Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni 2025-11-10): arxiviq.substack.com/p/nested-lea...
Nested Learning: The Illusion of Deep Learning Architectures
Authors: Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni
arxiviq.substack.com
November 30, 2025 at 4:48 PM
#AIML Nested Learning: The Illusion of Deep Learning Architectures - Paper: abehrouz.github.io/files/NL.pdf Blog: research.google/blog/introdu... - HOPE self-modifying Titans (Ali Behrouz, Meisam Razaviyayn, Peiling Zhong, Vahab Mirrokni 2025-11-10): arxiviq.substack.com/p/nested-lea...
The hero of the Wizard of Oz is Toto #WizardOfOz
while I’m being cranky, access journalism needs to go. When the president does his “I’m alpha because I insulted you” thing somebody needs to fall on his sword and say “what the fuck is wrong with you, old man?” on a hot mic. the hero of the wizard of Oz is Toto.
November 21, 2025 at 3:20 PM
The hero of the Wizard of Oz is Toto #WizardOfOz
Reposted
As a supervising attorney, I am here to tell young lawyers that AI is absolutely a threat to your career. I do not mean that it will replace you. I mean that if you use it, you increase the risk that another attorney will replace you when you are fired. www.reuters.com/sustainabili...
Two federal judges say use of AI led to errors in US court rulings
Two federal judges admitted in response to an inquiry by U.S. Senate Judiciary Committee Chairman Chuck Grassley that members of their staff used artificial intelligence to help prepare recent court orders that Grassley called "error-ridden."
www.reuters.com
October 26, 2025 at 11:35 AM
As a supervising attorney, I am here to tell young lawyers that AI is absolutely a threat to your career. I do not mean that it will replace you. I mean that if you use it, you increase the risk that another attorney will replace you when you are fired. www.reuters.com/sustainabili...
#AIML #OCR #PDF A picture is worth 1000 words but does not need 1000 tokens - DeepSeek-OCR: Contexts Optical Compression arXiv:2510.18234 - DeepEncoder and DeepSeek3B-MoE CLIP Gundam mode - testing via OmniDocBench (Haoran Wei, Yaofeng Sun, Yukun Li 2025-10-21): github.com/deepseek-ai/...
GitHub - deepseek-ai/DeepSeek-OCR: Contexts Optical Compression
Contexts Optical Compression. Contribute to deepseek-ai/DeepSeek-OCR development by creating an account on GitHub.
github.com
October 26, 2025 at 1:53 PM
#AIML #OCR #PDF A picture is worth 1000 words but does not need 1000 tokens - DeepSeek-OCR: Contexts Optical Compression arXiv:2510.18234 - DeepEncoder and DeepSeek3B-MoE CLIP Gundam mode - testing via OmniDocBench (Haoran Wei, Yaofeng Sun, Yukun Li 2025-10-21): github.com/deepseek-ai/...
Reposted
It's official! 🏛️ 🎉 The SF Board of Supervisors has unanimously declared October 22, 2025 as Internet Archive Day — celebrating our 1️⃣ trillion webpages preserved & our mission to provide universal access to all knowledge.
Learn more & celebrate with us tonight! ➡️ blog.archive.org/2025/10/22/h...
Learn more & celebrate with us tonight! ➡️ blog.archive.org/2025/10/22/h...
October 22, 2025 at 3:09 PM
It's official! 🏛️ 🎉 The SF Board of Supervisors has unanimously declared October 22, 2025 as Internet Archive Day — celebrating our 1️⃣ trillion webpages preserved & our mission to provide universal access to all knowledge.
Learn more & celebrate with us tonight! ➡️ blog.archive.org/2025/10/22/h...
Learn more & celebrate with us tonight! ➡️ blog.archive.org/2025/10/22/h...
#ARXR #AIML Easy and Fast? Rethinking the Future of Content Creation Tools - www-ui.is.s.u-tokyo.ac.jp/~takeo/paper... (Takeo Igarashi 2025-10-01): www-ui.is.s.u-tokyo.ac.jp/~takeo/paper...
www-ui.is.s.u-tokyo.ac.jp
October 13, 2025 at 6:54 PM
#ARXR #AIML Easy and Fast? Rethinking the Future of Content Creation Tools - www-ui.is.s.u-tokyo.ac.jp/~takeo/paper... (Takeo Igarashi 2025-10-01): www-ui.is.s.u-tokyo.ac.jp/~takeo/paper...
Reposted
"on stage said “Hey Meta, start Live AI session”, he triggered a live AI session on all the Ray-Ban Meta glasses in the room, DDoS-ing Meta Servers."🤡📣
That's the kind of person you want to give your personal data to. Think again.⚠️
That's the kind of person you want to give your personal data to. Think again.⚠️
My roundup of the best AR/VR news of the week is out! Of course there is a big summary of #MetaConnect, but I also mention news from Snap, Apple, Google, and more!
skarredghost.com/2025/09/22/m...
#VirtualReality #MixedReality
skarredghost.com/2025/09/22/m...
#VirtualReality #MixedReality
The XR Week Peek (2025.09.22): A big summary of Meta Connect, Snap and Apple's glasses update, and more!
Roundup of the most relevant XR news of the week, including a big summary of Meta Connect, and other updates
skarredghost.com
September 22, 2025 at 7:38 PM
"on stage said “Hey Meta, start Live AI session”, he triggered a live AI session on all the Ray-Ban Meta glasses in the room, DDoS-ing Meta Servers."🤡📣
That's the kind of person you want to give your personal data to. Think again.⚠️
That's the kind of person you want to give your personal data to. Think again.⚠️
#AIML #ARXR "We made painting feel more like typing, when we should have made typing feel like painting" -use tangible artifacts, concurrent modalities, ambient signals - from Our interfaces have lost their senses (Amelia Wattenberger 2025-03-11): wattenberger.com/thoughts/our...
September 3, 2025 at 1:16 PM
#AIML #ARXR "We made painting feel more like typing, when we should have made typing feel like painting" -use tangible artifacts, concurrent modalities, ambient signals - from Our interfaces have lost their senses (Amelia Wattenberger 2025-03-11): wattenberger.com/thoughts/our...
Reposted
Decision Process Theory, Twitter, Substack, More: Monday ResearchBuzz, August 25, 2025
NEW RESOURCES Wolfram Blog: Learning Decision Process Theory with a Wolfram Language Toolkit. "Wolfram Language provides powerful tools to compute and visualize both the optimal behaviors of decisions using game…
NEW RESOURCES Wolfram Blog: Learning Decision Process Theory with a Wolfram Language Toolkit. "Wolfram Language provides powerful tools to compute and visualize both the optimal behaviors of decisions using game…
Decision Process Theory, Twitter, Substack, More: Monday ResearchBuzz, August 25, 2025
NEW RESOURCES Wolfram Blog: Learning Decision Process Theory with a Wolfram Language Toolkit. "Wolfram Language provides powerful tools to compute and visualize both the optimal behaviors of decisions using game theory and an extension to a new field theory of games that explores how decisions change in time. I am excited to announce a free interactive course on decision process theory that will help students from all over the world explore this subject."
researchbuzz.me
August 25, 2025 at 12:01 PM
Decision Process Theory, Twitter, Substack, More: Monday ResearchBuzz, August 25, 2025
NEW RESOURCES Wolfram Blog: Learning Decision Process Theory with a Wolfram Language Toolkit. "Wolfram Language provides powerful tools to compute and visualize both the optimal behaviors of decisions using game…
NEW RESOURCES Wolfram Blog: Learning Decision Process Theory with a Wolfram Language Toolkit. "Wolfram Language provides powerful tools to compute and visualize both the optimal behaviors of decisions using game…
#AIML "It is time to stop pretending that we are just one more model away from magic. We are not. The magic trick is over. The rabbit is not coming out of the hat." from
The LLM In The Room: ChatGPT Speaks GPT-5 in it's own words (Jason Gorman 2025-08-08): codemanship.wordpress.com/2025/08/08/t...
The LLM In The Room: ChatGPT Speaks GPT-5 in it's own words (Jason Gorman 2025-08-08): codemanship.wordpress.com/2025/08/08/t...
The LLM In The Room: ChatGPT Speaks
With the long-awaited and much anticipated launch yesterday of the mythical GPT-5, and observing the widespread disappointment that it isn’t just not the superintelligence we were promised, b…
codemanship.wordpress.com
August 24, 2025 at 7:33 PM
#AIML "It is time to stop pretending that we are just one more model away from magic. We are not. The magic trick is over. The rabbit is not coming out of the hat." from
The LLM In The Room: ChatGPT Speaks GPT-5 in it's own words (Jason Gorman 2025-08-08): codemanship.wordpress.com/2025/08/08/t...
The LLM In The Room: ChatGPT Speaks GPT-5 in it's own words (Jason Gorman 2025-08-08): codemanship.wordpress.com/2025/08/08/t...
#AIML Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens arXiv:2508.01191 (Chengshuai Zhao, Zhen Tan, Pingchuan Ma, Dawei Li, Bohan Jiang, Yancheng Wang, Yingzhen Yang, Huan Liu 2025-08-05): github.com/ChengshuaiZh...
GitHub - ChengshuaiZhao0/DataAlchemy
Contribute to ChengshuaiZhao0/DataAlchemy development by creating an account on GitHub.
github.com
August 10, 2025 at 12:40 PM
#AIML Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens arXiv:2508.01191 (Chengshuai Zhao, Zhen Tan, Pingchuan Ma, Dawei Li, Bohan Jiang, Yancheng Wang, Yingzhen Yang, Huan Liu 2025-08-05): github.com/ChengshuaiZh...
#AIML Google just casually dropping the matrix on a random Tuesday - Video - Genie 3 creating dynamic worlds in real time (2025-08-05): www.youtube.com/watch?v=PDKh...
Genie 3: Creating dynamic worlds that you can navigate in real-time
YouTube video by Google DeepMind
www.youtube.com
August 7, 2025 at 8:25 PM
#AIML Google just casually dropping the matrix on a random Tuesday - Video - Genie 3 creating dynamic worlds in real time (2025-08-05): www.youtube.com/watch?v=PDKh...
Visualizating different #AIML model algorithms
Interactive LLM Visualizations - GPT2 code github.com/bbycroft/llm... (Brendan Croft 2023-12-28): bbycroft.net
Interactive LLM Visualizations - GPT2 code github.com/bbycroft/llm... (Brendan Croft 2023-12-28): bbycroft.net
GitHub - bbycroft/llm-viz: 3D Visualization of an GPT-style LLM
3D Visualization of an GPT-style LLM. Contribute to bbycroft/llm-viz development by creating an account on GitHub.
github.com
July 20, 2025 at 8:30 PM
Visualizating different #AIML model algorithms
Interactive LLM Visualizations - GPT2 code github.com/bbycroft/llm... (Brendan Croft 2023-12-28): bbycroft.net
Interactive LLM Visualizations - GPT2 code github.com/bbycroft/llm... (Brendan Croft 2023-12-28): bbycroft.net
Visualizating #AIML
Transformer Explainer: Interactive Learning of Text-Generative Models arXiv:2408.04619 - GPT2 LLM viz in a browser (Alec Helbling, Aeree Cho, Grace C. Kim, Alexander Karpekov, Zijie J. Wang, Seongmin Lee, Benjamin Hoover, Duen Horng Chau 2024-08-08): github.com/poloclub/tra...
Transformer Explainer: Interactive Learning of Text-Generative Models arXiv:2408.04619 - GPT2 LLM viz in a browser (Alec Helbling, Aeree Cho, Grace C. Kim, Alexander Karpekov, Zijie J. Wang, Seongmin Lee, Benjamin Hoover, Duen Horng Chau 2024-08-08): github.com/poloclub/tra...
GitHub - poloclub/transformer-explainer: Transformer Explained Visually: Learn How LLM Transformer Models Work with Interactive Visualization
Transformer Explained Visually: Learn How LLM Transformer Models Work with Interactive Visualization - poloclub/transformer-explainer
github.com
July 20, 2025 at 8:28 PM
Visualizating #AIML
Transformer Explainer: Interactive Learning of Text-Generative Models arXiv:2408.04619 - GPT2 LLM viz in a browser (Alec Helbling, Aeree Cho, Grace C. Kim, Alexander Karpekov, Zijie J. Wang, Seongmin Lee, Benjamin Hoover, Duen Horng Chau 2024-08-08): github.com/poloclub/tra...
Transformer Explainer: Interactive Learning of Text-Generative Models arXiv:2408.04619 - GPT2 LLM viz in a browser (Alec Helbling, Aeree Cho, Grace C. Kim, Alexander Karpekov, Zijie J. Wang, Seongmin Lee, Benjamin Hoover, Duen Horng Chau 2024-08-08): github.com/poloclub/tra...
#AIML How AI Can Degrade Human Performance in High-Stakes Settings - Across disciplines, bad AI predictions have a surprising tendency to make human experts perform worse - based on doi.org/10.1038/s417... (Dane Morey, Mike Rayo, David Woods 2025-07-15): ai-frontiers.org/articles/how...
Empirically derived evaluation requirements for responsible deployments of AI in safety-critical settings - npj Digital Medicine
npj Digital Medicine - Empirically derived evaluation requirements for responsible deployments of AI in safety-critical settings
doi.org
July 16, 2025 at 6:04 PM
#AIML How AI Can Degrade Human Performance in High-Stakes Settings - Across disciplines, bad AI predictions have a surprising tendency to make human experts perform worse - based on doi.org/10.1038/s417... (Dane Morey, Mike Rayo, David Woods 2025-07-15): ai-frontiers.org/articles/how...
#AIML Note how carefully they frame a 19% decrease to avoid setting off fans of AI Coding.
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (Joel Becker , Nate Rush , Beth Barnes, David Rein 2025-07-10): metr.org/blog/2025-07...
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (Joel Becker , Nate Rush , Beth Barnes, David Rein 2025-07-10): metr.org/blog/2025-07...
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
We conduct a randomized controlled trial to understand how early-2025 AI tools affect the productivity of experienced open-source developers working on their own repositories. Surprisingly, we find th...
metr.org
July 11, 2025 at 1:12 PM
#AIML Note how carefully they frame a 19% decrease to avoid setting off fans of AI Coding.
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (Joel Becker , Nate Rush , Beth Barnes, David Rein 2025-07-10): metr.org/blog/2025-07...
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (Joel Becker , Nate Rush , Beth Barnes, David Rein 2025-07-10): metr.org/blog/2025-07...
#AIML Centaur: a foundation model of human cognition arXiv:2410.20268 - Llama-3.1-Centaur-70B LLM trained on psych tests Psych-101 dataset (Marcel Binz, Elif Akata, Matthias Bethge, Franziska Brändle, et.al.
2025-04-28): marcelbinz.github.io/centaur
github.com/marcelbinz/L...
2025-04-28): marcelbinz.github.io/centaur
github.com/marcelbinz/L...
July 3, 2025 at 4:30 PM
#AIML Centaur: a foundation model of human cognition arXiv:2410.20268 - Llama-3.1-Centaur-70B LLM trained on psych tests Psych-101 dataset (Marcel Binz, Elif Akata, Matthias Bethge, Franziska Brändle, et.al.
2025-04-28): marcelbinz.github.io/centaur
github.com/marcelbinz/L...
2025-04-28): marcelbinz.github.io/centaur
github.com/marcelbinz/L...
#AIML Tools: Code Is All You Need - AIML agents work best with codegen rather than composing components via MCP Model Context Protocol (Armin Ronacher 2025-07-03): lucumr.pocoo.org/2025/7/3/too...
July 3, 2025 at 4:12 PM
#AIML Tools: Code Is All You Need - AIML agents work best with codegen rather than composing components via MCP Model Context Protocol (Armin Ronacher 2025-07-03): lucumr.pocoo.org/2025/7/3/too...
Reposted
EXCLUSIVE — The Onion has a full-page ad in today’s NYT featuring part of an editorial calling out congressional fecklessness towards Trump. They also sent a copy of their latest issue to every lawmaker.
My report, with an excerpt from the full editorial: www.thehandbasket.co/p/exclusive-...
My report, with an excerpt from the full editorial: www.thehandbasket.co/p/exclusive-...
June 22, 2025 at 1:32 PM
EXCLUSIVE — The Onion has a full-page ad in today’s NYT featuring part of an editorial calling out congressional fecklessness towards Trump. They also sent a copy of their latest issue to every lawmaker.
My report, with an excerpt from the full editorial: www.thehandbasket.co/p/exclusive-...
My report, with an excerpt from the full editorial: www.thehandbasket.co/p/exclusive-...
World’s smallest pacemaker is activated by light -can be inserted with a syringe, then dissolves after it’s no longer needed - optoelectronic systems for electrotherapy (Yamin Zhang, Eric Rytkin, John A. Rogers 2025-04-02): news.northwestern.edu/stories/2025...
World’s smallest pacemaker is activated by light
Tiny device can be inserted with a syringe, then dissolves after it’s no longer needed
news.northwestern.edu
June 19, 2025 at 10:29 PM
World’s smallest pacemaker is activated by light -can be inserted with a syringe, then dissolves after it’s no longer needed - optoelectronic systems for electrotherapy (Yamin Zhang, Eric Rytkin, John A. Rogers 2025-04-02): news.northwestern.edu/stories/2025...
#AIML Seven replies to the viral Apple reasoning paper – and why they fall short (Gary Marcus 2025-06-11): garymarcus.substack.com/p/seven-repl...
Seven replies to the viral Apple reasoning paper – and why they fall short
Also: another paper that seals the deal
garymarcus.substack.com
June 19, 2025 at 9:58 PM
#AIML Seven replies to the viral Apple reasoning paper – and why they fall short (Gary Marcus 2025-06-11): garymarcus.substack.com/p/seven-repl...
Reposted
Chatbots — LLMs — do not know facts and are not designed to be able to accurately answer factual questions. They are designed to find and mimic patterns of words, probabilistically. When they’re “right” it’s because correct things are often written down, so those patterns are frequent. That’s all.
June 19, 2025 at 11:21 AM
Chatbots — LLMs — do not know facts and are not designed to be able to accurately answer factual questions. They are designed to find and mimic patterns of words, probabilistically. When they’re “right” it’s because correct things are often written down, so those patterns are frequent. That’s all.
Laying the groundwork for future crypto-scams - Muddying the Waters: More Confusion on Crypto Asset Security Status from SEC Crypto Taskforce (Commissioner Caroline A. Crenshaw, 2025-05-31): www.sec.gov/newsroom/spe...
SEC.gov | Muddying the Waters: More Confusion on Crypto Asset Security StatusLock
Muddying the Waters: More Confusion on Crypto Asset Security Status Commissioner Caroline A. Crenshaw May 31, 2025
www.sec.gov
June 2, 2025 at 5:16 PM
Laying the groundwork for future crypto-scams - Muddying the Waters: More Confusion on Crypto Asset Security Status from SEC Crypto Taskforce (Commissioner Caroline A. Crenshaw, 2025-05-31): www.sec.gov/newsroom/spe...