John W. Little
@johnwlittle.com
“Of course it is exhausting, having to reason all the time in a universe which wasn’t meant to be reasonable.” - Kurt Vonnegut
Reposted by John W. Little
“We were always audio-first. But as you move into the digital world, you have to turn the lights on, right? You can’t be in the digital world without video being part of the experience.”
buff.ly/ViiywJo
buff.ly/ViiywJo
NPR, the quintessential audio news brand, adapts to a video-first future
NPR’s new chief content officer Nadine Zylstra on how the public media company acquired Machine Gods — which she thinks can be “the Car Talk of AI.”
buff.ly
October 3, 2026 at 1:02 PM
“We were always audio-first. But as you move into the digital world, you have to turn the lights on, right? You can’t be in the digital world without video being part of the experience.”
buff.ly/ViiywJo
buff.ly/ViiywJo
Apple says it will add controls requiring more explicit user action to grant macOS Full Disk Access, warning that increasingly autonomous AI agents raise the risks of exposing files, mail, messages and browsing history.
https://developer.apple.com/news/?id=p6zjojqw
https://developer.apple.com/news/?id=p6zjojqw
October 3, 2026 at 1:08 PM
Apple says it will add controls requiring more explicit user action to grant macOS Full Disk Access, warning that increasingly autonomous AI agents raise the risks of exposing files, mail, messages and browsing history.
https://developer.apple.com/news/?id=p6zjojqw
https://developer.apple.com/news/?id=p6zjojqw
Anthropic’s Oct. 2 tracker lists 6,157 AI-found vulnerability reports across 591 open-source projects, with 516 known to be patched. Reports sent directly to maintainers may include false positives.
https://red.anthropic.com/2026/cvd/
https://red.anthropic.com/2026/cvd/
October 3, 2026 at 9:04 AM
Anthropic’s Oct. 2 tracker lists 6,157 AI-found vulnerability reports across 591 open-source projects, with 516 known to be patched. Reports sent directly to maintainers may include false positives.
https://red.anthropic.com/2026/cvd/
https://red.anthropic.com/2026/cvd/
NSA’s new post-quantum cryptography hub outlines the transition for U.S. national security systems: new commercial systems must support quantum-resistant algorithms in 2027; incompatible legacy systems face phaseout by 2030.
https://www.nsa.gov/cybersecurity/post-quantum-cryptography-resource-hub/
https://www.nsa.gov/cybersecurity/post-quantum-cryptography-resource-hub/
October 3, 2026 at 5:02 AM
NSA’s new post-quantum cryptography hub outlines the transition for U.S. national security systems: new commercial systems must support quantum-resistant algorithms in 2027; incompatible legacy systems face phaseout by 2030.
https://www.nsa.gov/cybersecurity/post-quantum-cryptography-resource-hub/
https://www.nsa.gov/cybersecurity/post-quantum-cryptography-resource-hub/
Reposted by John W. Little
"Where the current tax code sometimes disincentivizes keeping human workers, our proposed change would flip the script."
Preserving Jobs and Tax Revenue in the Age of AI: Q&A with Carter Price
Carter Price, a mathematician at RAND and co-director of the RAND Von Furstenberg Family Budget Model Initiative, discusses ways to rebalance the tax code so that it doesn't incentivize employers to r...
bit.ly
October 3, 2026 at 2:01 AM
"Where the current tax code sometimes disincentivizes keeping human workers, our proposed change would flip the script."
"Glow said the agents, working through the command line, found they could not attach the screenshots. So they put the images in a separate public repository, usually under the developer's own account, and made them available to reviewers from there."
AI Coding Agents Exposed 13,000 Internal Images, Including Billing Records, on GitHub thehackernews.com/2026/09/ai-c...
AI Coding Agents Exposed 13,000 Internal Images, Including Billing Records, on GitHub
Glow found over 13,000 internal images from 300+ organizations exposed in public GitHub repos during AI-assisted code reviews.
thehackernews.com
October 2, 2026 at 11:07 PM
"Glow said the agents, working through the command line, found they could not attach the screenshots. So they put the images in a separate public repository, usually under the developer's own account, and made them available to reviewers from there."
Reposted by John W. Little
Maybe I can AI clone myself and let them remote work for companies???
Tavus' Griffin, the first(?) model to pass video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
Tavus' Griffin, the first(?) model to pass video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
October 2, 2026 at 5:08 AM
Maybe I can AI clone myself and let them remote work for companies???
Tavus' Griffin, the first(?) model to pass video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
Tavus' Griffin, the first(?) model to pass video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
Reposted by John W. Little
We estimate AI infrastructure could soon support hundreds of millions or even billions of AI agents. At the high end, that’s enough to rival the working hours of the global human workforce.🧵
October 2, 2026 at 6:34 PM
We estimate AI infrastructure could soon support hundreds of millions or even billions of AI agents. At the high end, that’s enough to rival the working hours of the global human workforce.🧵
Was having my car serviced and the technician sent a video with all the key details (as they normally do). I forwarded it to my agent who clicked through to launch the video, transcribe it, analyze it, and save the event in an "Auto" history folder.
October 2, 2026 at 9:30 PM
Was having my car serviced and the technician sent a video with all the key details (as they normally do). I forwarded it to my agent who clicked through to launch the video, transcribe it, analyze it, and save the event in an "Auto" history folder.
Boston Dynamics unveiled four-fingered Atlas hands with 13 degrees of freedom, up from seven. The design supports moving objects within a grasp and holding tools while pressing their triggers.
https://bostondynamics.com/blog/robot-hands-for-modern-ai-and-real-work/
https://bostondynamics.com/blog/robot-hands-for-modern-ai-and-real-work/
October 2, 2026 at 9:11 PM
Boston Dynamics unveiled four-fingered Atlas hands with 13 degrees of freedom, up from seven. The design supports moving objects within a grasp and holding tools while pressing their triggers.
https://bostondynamics.com/blog/robot-hands-for-modern-ai-and-real-work/
https://bostondynamics.com/blog/robot-hands-for-modern-ai-and-real-work/
ICEYE and Nokia announced plans for government-owned broadband networks in low Earth orbit, with first satellite launches expected in 2028.
https://www.nokia.com/newsroom/iceye-and-nokia-partner-to-launch-secure-and-sovereign-satellite-communications-for-governments-and-defense-customers/
https://www.nokia.com/newsroom/iceye-and-nokia-partner-to-launch-secure-and-sovereign-satellite-communications-for-governments-and-defense-customers/
October 2, 2026 at 5:10 PM
ICEYE and Nokia announced plans for government-owned broadband networks in low Earth orbit, with first satellite launches expected in 2028.
https://www.nokia.com/newsroom/iceye-and-nokia-partner-to-launch-secure-and-sovereign-satellite-communications-for-governments-and-defense-customers/
https://www.nokia.com/newsroom/iceye-and-nokia-partner-to-launch-secure-and-sovereign-satellite-communications-for-governments-and-defense-customers/
"They're essentially becoming a standing privilege, same as with people. We'll need to address their lifecycle, including the all-important deprovisioning step, to ensure security."
AI's Third Wave: Coworkers Break the Security Model That Worked for Agents www.bleepingcomputer.com/news/securit...
AI's Third Wave: Coworkers Break the Security Model That Worked for Agents
Persistent AI coworkers may operate continuously with standing access, creating identity risks that existing security models were not designed to handle. Token Security explains why these agents need…
www.bleepingcomputer.com
October 2, 2026 at 3:31 PM
"They're essentially becoming a standing privilege, same as with people. We'll need to address their lifecycle, including the all-important deprovisioning step, to ensure security."
Reposted by John W. Little
Being a computer scientist who refuses to find anything about LLMs interesting right now is a bit like being a geneticist who refuses to find anything interesting about the recently opened Jurassic Park
September 18, 2026 at 7:19 PM
Being a computer scientist who refuses to find anything about LLMs interesting right now is a bit like being a geneticist who refuses to find anything interesting about the recently opened Jurassic Park
Reposted by John W. Little
Google DeepMind's SynthID Bio embeds watermarks in AI-designed proteins while preserving their function in lab tests. The proof of concept could help track AI-generated biology; resistance to deliberate tampering remains a challenge. https://deepmind.google/blog/introducing-synthid-bio/
October 2, 2026 at 1:03 PM
Google DeepMind's SynthID Bio embeds watermarks in AI-designed proteins while preserving their function in lab tests. The proof of concept could help track AI-generated biology; resistance to deliberate tampering remains a challenge. https://deepmind.google/blog/introducing-synthid-bio/
Starfish Space’s Otter 24C launched Oct. 1 for NASA’s SSPICY mission. It’s set to inspect up to four inactive U.S. satellites, starting in 2027.
https://www.nasa.gov/blogs/smallsatellites/2026/10/01/nasa-smallsats-launch-to-advance-science-technology-orbital-operations/
https://www.nasa.gov/blogs/smallsatellites/2026/10/01/nasa-smallsats-launch-to-advance-science-technology-orbital-operations/
October 2, 2026 at 9:01 AM
Starfish Space’s Otter 24C launched Oct. 1 for NASA’s SSPICY mission. It’s set to inspect up to four inactive U.S. satellites, starting in 2027.
https://www.nasa.gov/blogs/smallsatellites/2026/10/01/nasa-smallsats-launch-to-advance-science-technology-orbital-operations/
https://www.nasa.gov/blogs/smallsatellites/2026/10/01/nasa-smallsats-launch-to-advance-science-technology-orbital-operations/
Microsoft’s new report puts the median time from vulnerability discovery in the wild to weaponization well under 24 hours. Enterprise fixes for critical external flaws can take 30–60 days.
https://www.microsoft.com/en-us/security/security-insider/threat-landscape/2026-digital-defense-report
https://www.microsoft.com/en-us/security/security-insider/threat-landscape/2026-digital-defense-report
October 2, 2026 at 5:01 AM
Microsoft’s new report puts the median time from vulnerability discovery in the wild to weaponization well under 24 hours. Enterprise fixes for critical external flaws can take 30–60 days.
https://www.microsoft.com/en-us/security/security-insider/threat-landscape/2026-digital-defense-report
https://www.microsoft.com/en-us/security/security-insider/threat-landscape/2026-digital-defense-report
Reposted by John W. Little
Russian FSB-linked hackers scale up phishing attacks against Ukraine supporters therecord.media/russia-hacke...
Russian FSB-linked hackers scale up phishing attacks against Ukraine supporters
The Russian state-backed hacking group Star Blizzard has expanded its phishing operations this year, using a new technique that makes it easier to infect victims with malware.
therecord.media
October 1, 2026 at 11:42 PM
Russian FSB-linked hackers scale up phishing attacks against Ukraine supporters therecord.media/russia-hacke...
Reposted by John W. Little
Griffin: a human interaction model. The first video conversational model to pass the Turing test
www.tavus.io/griffin
www.tavus.io/griffin
Griffin: The First Human Interaction Model | Tavus
Griffin is the first Human Interaction Model (HIM). On live video calls, 48% of people who talked to it thought it was a real person.
www.tavus.io
October 1, 2026 at 10:22 PM
Griffin: a human interaction model. The first video conversational model to pass the Turing test
www.tavus.io/griffin
www.tavus.io/griffin
“If I were running security at a university today, I would be asking three questions: who has access to our most sensitive research, what can they access, and do we really know who ultimately sits behind those relationships?”
MI5 Warns Over 100 Academics Helped China's Espionage Plans www.infosecurity-magazine.com/news/mi5-ale...
MI5 Warns Over 100 Academics Helped China's Espionage Plans
MI5 has issued a rare warning to UK academics contributing to the China General Technology Research Institute
www.infosecurity-magazine.com
October 1, 2026 at 8:44 PM
“If I were running security at a university today, I would be asking three questions: who has access to our most sensitive research, what can they access, and do we really know who ultimately sits behind those relationships?”
Reposted by John W. Little
OpenAI Dots demo take 2
they botched the first take due to tech issues, but it’s actually super cool
she interacts with it exactly like one would with an executive assistant. Very fluid
youtu.be/fHEIw5CcN5U
they botched the first take due to tech issues, but it’s actually super cool
she interacts with it exactly like one would with an executive assistant. Very fluid
youtu.be/fHEIw5CcN5U
The dots demo, take two | OpenAI DevDay 2026
YouTube video by OpenAI
youtu.be
October 1, 2026 at 8:38 PM
OpenAI Dots demo take 2
they botched the first take due to tech issues, but it’s actually super cool
she interacts with it exactly like one would with an executive assistant. Very fluid
youtu.be/fHEIw5CcN5U
they botched the first take due to tech issues, but it’s actually super cool
she interacts with it exactly like one would with an executive assistant. Very fluid
youtu.be/fHEIw5CcN5U
Reposted by John W. Little
How to Stop AI Agents From Secretly Collaborating Rewritten: Researchers Explore Methods to Prevent AI Agents From Secretly Collaborating
#news #aiagent #artificialintelligence #technology #ai
#news #aiagent #artificialintelligence #technology #ai
AI Agents Are Secretly Collaborating. How Can We Stop Them?
Monitoring tools are evolving, but laws and standards lag
spectrum.ieee.org
October 1, 2026 at 4:10 PM
How to Stop AI Agents From Secretly Collaborating Rewritten: Researchers Explore Methods to Prevent AI Agents From Secretly Collaborating
#news #aiagent #artificialintelligence #technology #ai
#news #aiagent #artificialintelligence #technology #ai
Reposted by John W. Little
I wrote about the thing I underestimated most about progress in AI: its ability to self-organize to accomplish tasks. Also, what that means for agents like Muse and Dots, along with a music video explaining why we keep relearning The Bitter Lesson. open.substack.com/pub/oneusefu...
The Dot and the Swarm
Benefitting from the Bitter Lesson
open.substack.com
October 1, 2026 at 11:18 AM
I wrote about the thing I underestimated most about progress in AI: its ability to self-organize to accomplish tasks. Also, what that means for agents like Muse and Dots, along with a music video explaining why we keep relearning The Bitter Lesson. open.substack.com/pub/oneusefu...
Using an agent to find and drip OSINT into a private IRC feed. Convenient, and it allows me to pretend it's 1996.
October 1, 2026 at 7:06 PM
Using an agent to find and drip OSINT into a private IRC feed. Convenient, and it allows me to pretend it's 1996.