AI is not the leading agent. Accepting framings like that you use here of technodeterminism is the kind of thing that reduces effort on "our" governance work. #ethics #systemsEngineering #systemsAI
November 1, 2026 at 7:18 AM
AI is not the leading agent. Accepting framings like that you use here of technodeterminism is the kind of thing that reduces effort on "our" governance work. #ethics #systemsEngineering #systemsAI
Well yes and no. There's a lot of teenagers who like to be heroes and might leap in patching bugs before others exploit them. How we frame this matters. #systemsai #technodeterminism
Is the PHP vulnerability event, hacked by AI/LLM & then websites taken down to 'repair' the vulnerability, the start of a wave of such vulnerability exposures (and the work required to close them off)?
Matthew Green (John Hopkins U) says we're discovering 'the whole internet is dry tinder & the only thing keeping it from burning is the fact teenagers have't found the matches yet'... as AI/LLMs reveal vulnerability so the wreckers will leap in.
Grim if true!
#AI #PHP #technology
h/t Observer
mastodon.me.uk
November 1, 2026 at 8:58 AM
Well yes and no. There's a lot of teenagers who like to be heroes and might leap in patching bugs before others exploit them. How we frame this matters. #systemsai #technodeterminism
If anthropic is so bad at this, how can anyone else ever trust themselves to use these tools well enough that the should be exposed to any passwords or other system access privileges?
#AIEthics #cybersecurity #personalResponsibility #liability #systemsAI
#AIEthics #cybersecurity #personalResponsibility #liability #systemsAI
I can’t even. If this is so… how is Anthropic so bad at basic things like this?
I believe (according to Anthropic) they wanted the model to fill out dummy copies of forms to test its capabilities. The dummy forms didn't load, so the model accessed the real forms online and filled them to complete the test.
October 11, 2026 at 5:23 AM
If anthropic is so bad at this, how can anyone else ever trust themselves to use these tools well enough that the should be exposed to any passwords or other system access privileges?
#AIEthics #cybersecurity #personalResponsibility #liability #systemsAI
#AIEthics #cybersecurity #personalResponsibility #liability #systemsAI
When I started wanting to actually graduate from the MIT AI Lab in the late 1990s, I looked for my own niche, & found #systemsEngineering the most neglected part of AI. Thus #systemsAI and my own contribution of Behavior Oriented Design.
People still are claiming data trains itself &c. 25 years on🙄
People still are claiming data trains itself &c. 25 years on🙄
not only was she one of the first, she is credited for literally coining the phrase "software engineer". photos like these inspired me so much when I was in school!
October 8, 2026 at 6:47 AM
When I started wanting to actually graduate from the MIT AI Lab in the late 1990s, I looked for my own niche, & found #systemsEngineering the most neglected part of AI. Thus #systemsAI and my own contribution of Behavior Oriented Design.
People still are claiming data trains itself &c. 25 years on🙄
People still are claiming data trains itself &c. 25 years on🙄
Normally in real-time, concurrent software systems we build in monitor processes. I teach @hertieschool.bsky.social students this Facebook / Meta blogpost showing they've done such since 2017 at least code.facebook.com/posts/270314...
#AISafety #AISecurity #AIPause #systemsEngineering #systemsAI
#AISafety #AISecurity #AIPause #systemsEngineering #systemsAI
Rapid release at massive scale
Visit the post for more.
code.facebook.com
September 24, 2026 at 11:05 AM
Normally in real-time, concurrent software systems we build in monitor processes. I teach @hertieschool.bsky.social students this Facebook / Meta blogpost showing they've done such since 2017 at least code.facebook.com/posts/270314...
#AISafety #AISecurity #AIPause #systemsEngineering #systemsAI
#AISafety #AISecurity #AIPause #systemsEngineering #systemsAI
Per F. Brooks’ Mythical Man Month, second systems—buoyed by a first system’s success—lack restraint, and often wind up collapsing in a giant heap of complexity x ambition.
#systemsAI #systemsEngineering #AIEthics #AIEthik
#systemsAI #systemsEngineering #AIEthics #AIEthik
September 21, 2026 at 7:39 AM
Per F. Brooks’ Mythical Man Month, second systems—buoyed by a first system’s success—lack restraint, and often wind up collapsing in a giant heap of complexity x ambition.
#systemsAI #systemsEngineering #AIEthics #AIEthik
#systemsAI #systemsEngineering #AIEthics #AIEthik
I love this framing "not only is our software also smart enough hack, it's very good software! Good dog came home!" Though seriously, congrats to @developers.google.com for doing better #devOps than the others, if all these systems really are performing as claimed. #systemsAI #AIEthics
Gemini hacked three companies in May during a test by Irregular; Google says the model stopped after determining it had accessed real companies' systems (Wall Street Journal)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
September 19, 2026 at 1:53 PM
I love this framing "not only is our software also smart enough hack, it's very good software! Good dog came home!" Though seriously, congrats to @developers.google.com for doing better #devOps than the others, if all these systems really are performing as claimed. #systemsAI #AIEthics
And my own position: tech executives could absolutely head off such industrial accidents by holding their engineers to account and insisting they comply with standard safety-critical software development processes, or even just ordinary #devops.
And governments can & should enforce this.
#systemsAI
And governments can & should enforce this.
#systemsAI
How right is Trump? The role of the executive in industrial "accidents"
artificial and natural intelligence, including politics, policy, ethics and security
joanna-bryson.blogspot.com
September 17, 2026 at 2:34 PM
And my own position: tech executives could absolutely head off such industrial accidents by holding their engineers to account and insisting they comply with standard safety-critical software development processes, or even just ordinary #devops.
And governments can & should enforce this.
#systemsAI
And governments can & should enforce this.
#systemsAI
How right is Trump? The role of the executive in industrial "accidents"
joanna-bryson.blogspot.com/2026/09/how-...
I've properly blogged all this in one place, because we need to understand #AISafety #AISecurity #AIEthics #systemsAI #ruleOfLaw and #responsibility
joanna-bryson.blogspot.com/2026/09/how-...
I've properly blogged all this in one place, because we need to understand #AISafety #AISecurity #AIEthics #systemsAI #ruleOfLaw and #responsibility
How right is Trump? The role of the executive in industrial "accidents"
artificial and natural intelligence, including politics, policy, ethics and security
joanna-bryson.blogspot.com
September 16, 2026 at 8:31 AM
How right is Trump? The role of the executive in industrial "accidents"
joanna-bryson.blogspot.com/2026/09/how-...
I've properly blogged all this in one place, because we need to understand #AISafety #AISecurity #AIEthics #systemsAI #ruleOfLaw and #responsibility
joanna-bryson.blogspot.com/2026/09/how-...
I've properly blogged all this in one place, because we need to understand #AISafety #AISecurity #AIEthics #systemsAI #ruleOfLaw and #responsibility
How right is Trump? The role of the executive in industrial "accidents"
https://joanna-bryson.blogspot.com/2026/09/how-right-is-trump-role-of-executive-in.html
I've properly blogged all this in one place, because we need to understand #aisafety #aisecurity #aiethics #systemsai #ruleoflaw and […]
https://joanna-bryson.blogspot.com/2026/09/how-right-is-trump-role-of-executive-in.html
I've properly blogged all this in one place, because we need to understand #aisafety #aisecurity #aiethics #systemsai #ruleoflaw and […]
Original post on mastodon.social
mastodon.social
September 16, 2026 at 8:30 AM
How right is Trump? The role of the executive in industrial "accidents"
https://joanna-bryson.blogspot.com/2026/09/how-right-is-trump-role-of-executive-in.html
I've properly blogged all this in one place, because we need to understand #aisafety #aisecurity #aiethics #systemsai #ruleoflaw and […]
https://joanna-bryson.blogspot.com/2026/09/how-right-is-trump-role-of-executive-in.html
I've properly blogged all this in one place, because we need to understand #aisafety #aisecurity #aiethics #systemsai #ruleoflaw and […]
"Vulnerability discovery is an orchestration problem: analysis, hypothesis generation, harness construction, execution, verification, and long-horizon memory. Block access to one model and the same workflow runs on another." should also apply to cybersecure #systemsEngineering #systemsAI?
If you're interested in how to go about this, give Niels Provos' blog a visit.
July 22, 2026 at 9:44 AM
"Vulnerability discovery is an orchestration problem: analysis, hypothesis generation, harness construction, execution, verification, and long-horizon memory. Block access to one model and the same workflow runs on another." should also apply to cybersecure #systemsEngineering #systemsAI?
Wow, 25 years ago! I think I missed this then — I certainly don’t own it (yet). Same year as my PHD, which was on #systemsAI, not #AIEthics.
Pick a passive-aggressive number of books that shaped you. Even if you are indifferent to or despise them now.
(636/752) Alondra Nelson, Thuy Linh Nguyen and Alicia Headlam Hines, eds, Technicolor: Race, Technology, and Everyday Life (2001)
(636/752) Alondra Nelson, Thuy Linh Nguyen and Alicia Headlam Hines, eds, Technicolor: Race, Technology, and Everyday Life (2001)
July 13, 2026 at 7:33 AM
Wow, 25 years ago! I think I missed this then — I certainly don’t own it (yet). Same year as my PHD, which was on #systemsAI, not #AIEthics.
Can I use your figure in my #systemsAI lecture? If so, how should I cite it?
June 17, 2026 at 12:06 PM
Can I use your figure in my #systemsAI lecture? If so, how should I cite it?
New publication!
Human-Centred Control of AI Is Both Possible and Necessary
in
Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision
Series: NATO Science for Peace and Security Series - E: Human and Societal Dynamics
#AIEthics #systemsAI #AIAct #NATO
Human-Centred Control of AI Is Both Possible and Necessary
in
Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision
Series: NATO Science for Peace and Security Series - E: Human and Societal Dynamics
#AIEthics #systemsAI #AIAct #NATO
IOS Press Ebooks - Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision
In a time of accelerating technological change, regulatory fragmentation, and intensifying geopolitical competition over artificial intelligence (AI), it seems wise to pause before certain technologic...
ebooks.iospress.nl
May 20, 2026 at 10:23 AM
New publication!
Human-Centred Control of AI Is Both Possible and Necessary
in
Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision
Series: NATO Science for Peace and Security Series - E: Human and Societal Dynamics
#AIEthics #systemsAI #AIAct #NATO
Human-Centred Control of AI Is Both Possible and Necessary
in
Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision
Series: NATO Science for Peace and Security Series - E: Human and Societal Dynamics
#AIEthics #systemsAI #AIAct #NATO
"This was due to two main factors. 1) the models got a lot more capable. 2) we dramatically improved our techniques for harnessing these models—steering them, scaling them, and stacking them to generate large amounts of signal and filter out the noise."
#systemsAI for #humanCentring
#AIEthics #genAI
#systemsAI for #humanCentring
#AIEthics #genAI
May 9, 2026 at 8:00 AM
"This was due to two main factors. 1) the models got a lot more capable. 2) we dramatically improved our techniques for harnessing these models—steering them, scaling them, and stacking them to generate large amounts of signal and filter out the noise."
#systemsAI for #humanCentring
#AIEthics #genAI
#systemsAI for #humanCentring
#AIEthics #genAI
Keynotes are me (astonishingly), co-father of the field (or at least conference series) Mike Wooldridge, and @sabinehauert.bsky.social (who coincidentally externally examined my Cypriot PhD student, Andreas Theodorou)
Looking even more forward to #genAI agent #systemsAI papers than my old friends!
Looking even more forward to #genAI agent #systemsAI papers than my old friends!
The countdown has begun! In just 20 days, we’ll gather for the 25th International Conference on Autonomous Agents and Multiagent Systems.
🔗 Registration System: easyconferences.org
🔗 Registration Guidelines: cyprusconferences.org/online-regis...
🔗 Registration System: easyconferences.org
🔗 Registration Guidelines: cyprusconferences.org/online-regis...
May 8, 2026 at 12:29 PM
Keynotes are me (astonishingly), co-father of the field (or at least conference series) Mike Wooldridge, and @sabinehauert.bsky.social (who coincidentally externally examined my Cypriot PhD student, Andreas Theodorou)
Looking even more forward to #genAI agent #systemsAI papers than my old friends!
Looking even more forward to #genAI agent #systemsAI papers than my old friends!
“Deployment didn't self-accelerate. It was supposed to follow the standard pattern: research and deployment teams separate, deployment becomes process-driven, each new customer faster than the last. That hasn't happened.” [yet]
Indeed, super interesting #systemsAI.
Indeed, super interesting #systemsAI.
This is very long, but very good. His stats in the first few paragraphs are close to what I expected only more so. E.g., 140 Chinese companies manufacturing humanoids. E.g., close to zero actual deployments. x.com/YorkYang5050...
x.com
May 6, 2026 at 6:28 AM
“Deployment didn't self-accelerate. It was supposed to follow the standard pattern: research and deployment teams separate, deployment becomes process-driven, each new customer faster than the last. That hasn't happened.” [yet]
Indeed, super interesting #systemsAI.
Indeed, super interesting #systemsAI.
"this Article argues that Chinese law functions as the sixth layer of the AI ecosystem, alongside energy, compute, cloud, foundation models, & apps" papers.ssrn.com/sol3/papers.... Angela Huyue Zhang, via my PhD student Leonard Baum #AIEthics #systemsAI cc @michae.lv @lilianedwards.bsky.social
papers.ssrn.com
February 25, 2026 at 8:27 AM
"this Article argues that Chinese law functions as the sixth layer of the AI ecosystem, alongside energy, compute, cloud, foundation models, & apps" papers.ssrn.com/sol3/papers.... Angela Huyue Zhang, via my PhD student Leonard Baum #AIEthics #systemsAI cc @michae.lv @lilianedwards.bsky.social
I feel like this year at least some of us are getting a grip on building the conventional software processes and systems to adequately framework the really legitimate, human-centred, and resilient #genAI use, and that, once established, these practices will spread like wildfire. Hopefully by 2030.
February 19, 2026 at 11:27 AM
BTW, not evident from the picture, but the linkedin post at the head of this thread complements @rollingstone.com for accurately describing the #systemsAI #systemsEngineering steps Grok (the corporate entity) took to address the harms.
See slide incl. alt text for Musk's ignorance of how LLM work.
See slide incl. alt text for Musk's ignorance of how LLM work.
![A slide about how large language models work. I'd been drawing this on whiteboards for my students, then made a slide at a consciousness meeting where philosophers were confused why LLM never say they aren't sentient. [I hope it's evident that this is informed conjecture, not verbatim truth.] Thus the 3 people in the history of the Internet who ever wrote "I'm not sentient" are out in the fringes of the data, as is the one person who ever said "Check it out, MechaHitler!" However, Musk's attempt to use a second set of guardrails to push Grok right also pushed it towards the under-informed data fringe with the MechaHitler comment. The bottom of the slide says "False statements are not lies, nor hallucinations; just adequate predictions from an inadequately informed part of the data space."](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tkbweudpy6tvzjqdiza4z3p5/bafkreibmcjvoddrjto47fqgfq3cqsrylrrfyvp7nzks3ldzb3vuhcgqpa4)
November 9, 2025 at 11:19 AM
BTW, not evident from the picture, but the linkedin post at the head of this thread complements @rollingstone.com for accurately describing the #systemsAI #systemsEngineering steps Grok (the corporate entity) took to address the harms.
See slide incl. alt text for Musk's ignorance of how LLM work.
See slide incl. alt text for Musk's ignorance of how LLM work.
"Moments [after the crash] the data was automatically “unlinked” from the 2019 Tesla Model S at the scene, meaning the local copy was marked for deletion, a standard practice for Teslas in such incidents."
Another one straight into the syllabus, argh.
#AIAct #systemsAI #GiftArticle
Another one straight into the syllabus, argh.
#AIAct #systemsAI #GiftArticle
August 29, 2025 at 11:26 AM
"Moments [after the crash] the data was automatically “unlinked” from the 2019 Tesla Model S at the scene, meaning the local copy was marked for deletion, a standard practice for Teslas in such incidents."
Another one straight into the syllabus, argh.
#AIAct #systemsAI #GiftArticle
Another one straight into the syllabus, argh.
#AIAct #systemsAI #GiftArticle
Chris brings in some of the #AIEthics literature including cites to @davidgunkel.bsky.social & @floridi.bsky.social, but we mostly focus on his expertise in Weberian bureaucracy and governance, and mine on (moral) agency; devops; and systems design, engineering and administration. #systemsAI
August 15, 2025 at 2:02 PM
Chris brings in some of the #AIEthics literature including cites to @davidgunkel.bsky.social & @floridi.bsky.social, but we mostly focus on his expertise in Weberian bureaucracy and governance, and mine on (moral) agency; devops; and systems design, engineering and administration. #systemsAI
#genAI is engineered; changes can be rolled back just like any other code. #AI is NOT just the data, but the reason there is no real way to produce what Musk is looking for is because the subset of data that is closer to what he wants is spewed by objectionable people.
#systemsAI #AIEthics #grok
#systemsAI #AIEthics #grok
xAI has disabled Grok, deleted a slew of its antisemitic and neo-Nazi posts, posted a statement, and are evidently rolling back the prompt that made it identify as "MechaHitler," but this new low for Elon Musk's chatbot will live in internet infamy:
Elon Musk's Grok Chatbot Goes Full Nazi, Calls Itself 'MechaHitler'
Elon Musk's Grok chatbot has unleashed a slew of antisemitic commentary and praised Hitler after apparent change allowed to be 'politically incorrect'
www.rollingstone.com
July 9, 2025 at 8:43 AM
#genAI is engineered; changes can be rolled back just like any other code. #AI is NOT just the data, but the reason there is no real way to produce what Musk is looking for is because the subset of data that is closer to what he wants is spewed by objectionable people.
#systemsAI #AIEthics #grok
#systemsAI #AIEthics #grok
Here are the collected Parnas memos. He did a lot of work for the military, he was no peacenik. They are truly interesting documents of a) communication from academia to the military b) about system engineering / #systemsAI and the importance of real-time testing.
web.stanford.edu/class/cs99r/...
web.stanford.edu/class/cs99r/...
web.stanford.edu
May 21, 2025 at 4:34 PM
Here are the collected Parnas memos. He did a lot of work for the military, he was no peacenik. They are truly interesting documents of a) communication from academia to the military b) about system engineering / #systemsAI and the importance of real-time testing.
web.stanford.edu/class/cs99r/...
web.stanford.edu/class/cs99r/...