People bitch about rules, “woke”ness, regulations, guardrails, but as those of us flyover people in the Midwest have known all our lives, we have these things for a reason. Doesn’t matter who’s in power: without these things quality of life quickly devolves.
October 3, 2026 at 12:43 PM
People bitch about rules, “woke”ness, regulations, guardrails, but as those of us flyover people in the Midwest have known all our lives, we have these things for a reason. Doesn’t matter who’s in power: without these things quality of life quickly devolves.
Mr. Robinson is correct that we need multiple layers of guardrails like we do with nuclear weapons. The current administration and Republican Party is refusing to do so.
New just now in The Atlantic:
David Robinson resigned this week. He was among the longest-tenured employees at OpenAI—and oversaw safety reports on 12 frontier launches.
He is very worried: “The time for trial and error is over.”
You can read his essay here:
www.theatlantic.com/technology/2...
David Robinson resigned this week. He was among the longest-tenured employees at OpenAI—and oversaw safety reports on 12 frontier launches.
He is very worried: “The time for trial and error is over.”
You can read his essay here:
www.theatlantic.com/technology/2...
I Quit OpenAI Because Its Culture Is Broken
The industry’s approach to safety guarantees more failures.
www.theatlantic.com
October 3, 2026 at 12:40 PM
Mr. Robinson is correct that we need multiple layers of guardrails like we do with nuclear weapons. The current administration and Republican Party is refusing to do so.
🏔️🚴🏻♂️Guardrails? No, we put 20cm high blocks along the 20m drop. The guardrail would ruin the beautiful view.
October 3, 2026 at 12:31 PM
🏔️🚴🏻♂️Guardrails? No, we put 20cm high blocks along the 20m drop. The guardrail would ruin the beautiful view.
Thailand Needs to Urgently Set Up a New Regulatory Body to Control the Wellness Industry
thailandwellness.news/thailand-nee...
#ThailandWellnessNews #WellnessNews #ThailandWellness #News #Thailand #Bangkok #Phuket #Wellness #Health #Asia #Australia #America #Europe #Auckland #Canada #UK #UAE
thailandwellness.news/thailand-nee...
#ThailandWellnessNews #WellnessNews #ThailandWellness #News #Thailand #Bangkok #Phuket #Wellness #Health #Asia #Australia #America #Europe #Auckland #Canada #UK #UAE
Thailand Needs to Urgently Set Up a New Regulatory Body to Control the Wellness Industry - thailandwellness.news
A Booming Market Needs Stronger Guardrails
thailandwellness.news
October 3, 2026 at 12:16 PM
Thailand Needs to Urgently Set Up a New Regulatory Body to Control the Wellness Industry
thailandwellness.news/thailand-nee...
#ThailandWellnessNews #WellnessNews #ThailandWellness #News #Thailand #Bangkok #Phuket #Wellness #Health #Asia #Australia #America #Europe #Auckland #Canada #UK #UAE
thailandwellness.news/thailand-nee...
#ThailandWellnessNews #WellnessNews #ThailandWellness #News #Thailand #Bangkok #Phuket #Wellness #Health #Asia #Australia #America #Europe #Auckland #Canada #UK #UAE
Oh I have hope. What I don’t have is TRUST. I am hoping for such a blue wave that all we hear the next day is “Surfin’ USA”. I don’t have much TRUST in our systems these days. I will give Republicans credit for exposing how weak our guardrails are.
October 3, 2026 at 12:14 PM
Oh I have hope. What I don’t have is TRUST. I am hoping for such a blue wave that all we hear the next day is “Surfin’ USA”. I don’t have much TRUST in our systems these days. I will give Republicans credit for exposing how weak our guardrails are.
Midterms
Americans should seriously prepare for an attempted political/legal struggle over the legitimacy & certification of the voting results.
Rogue officials may attempt to interfere with the timely certification of this year’s midterm election results!
www.brennancenter.org/series/elect...
Americans should seriously prepare for an attempted political/legal struggle over the legitimacy & certification of the voting results.
Rogue officials may attempt to interfere with the timely certification of this year’s midterm election results!
www.brennancenter.org/series/elect...
Election Certification Processes and Guardrails
Election deniers have targeted certification as a way to interfere with election results, but there are legal protections in place against such efforts.
www.brennancenter.org
October 3, 2026 at 12:12 PM
Midterms
Americans should seriously prepare for an attempted political/legal struggle over the legitimacy & certification of the voting results.
Rogue officials may attempt to interfere with the timely certification of this year’s midterm election results!
www.brennancenter.org/series/elect...
Americans should seriously prepare for an attempted political/legal struggle over the legitimacy & certification of the voting results.
Rogue officials may attempt to interfere with the timely certification of this year’s midterm election results!
www.brennancenter.org/series/elect...
Bedrock Guardrailsで「みりん」が人名として伏せられた事例。個人情報マスクは、固有名詞っぽい普通の単語で誤検知が出る。マスクを入れたら、本番前に自分のドメインの語彙を流して、何が消えるかを一度見ておくのが安全。
October 3, 2026 at 12:10 PM
Bedrock Guardrailsで「みりん」が人名として伏せられた事例。個人情報マスクは、固有名詞っぽい普通の単語で誤検知が出る。マスクを入れたら、本番前に自分のドメインの語彙を流して、何が消えるかを一度見ておくのが安全。
one possibility: because such people will eagerly commit crimes against humanity as soon as the guardrails are down and encouragement is given by those who are desperate to hold onto wealth, power, and entitlement (with no legal consequences).
October 3, 2026 at 12:02 PM
one possibility: because such people will eagerly commit crimes against humanity as soon as the guardrails are down and encouragement is given by those who are desperate to hold onto wealth, power, and entitlement (with no legal consequences).
one underdiscussed aspect of AI, except kind of in the context of surveillance, is
AI Company: we've finally created The Ethical AI, that is energy efficient, doesn't plagiarize, and has guardrails
AI user: i'm still just going to use it to steal and deepfake and lie though lol
AI Company: we've finally created The Ethical AI, that is energy efficient, doesn't plagiarize, and has guardrails
AI user: i'm still just going to use it to steal and deepfake and lie though lol
To interview an expert, claim corporate ownership of the convo, + feed that audio (likely unredacted!) into AI models to draw on without appropriate sensitivity to subject matter or context? Experts like Nicole aren't a soundbites jukebox, + folks who contribute their time/talent deserve better.
Earlier this week, I agreed to interviews with two USA Today journalists. They wrote fantastic articles about the Cornell case that I'm proud to have been a part of.
What I did not know is that their employer considers interviews to proprietary and opens up recordings to be used by any journalist.
What I did not know is that their employer considers interviews to proprietary and opens up recordings to be used by any journalist.
October 3, 2026 at 11:53 AM
one underdiscussed aspect of AI, except kind of in the context of surveillance, is
AI Company: we've finally created The Ethical AI, that is energy efficient, doesn't plagiarize, and has guardrails
AI user: i'm still just going to use it to steal and deepfake and lie though lol
AI Company: we've finally created The Ethical AI, that is energy efficient, doesn't plagiarize, and has guardrails
AI user: i'm still just going to use it to steal and deepfake and lie though lol
hitting safety guardrails after asking a personal agent to read through its memories of me
October 3, 2026 at 11:50 AM
hitting safety guardrails after asking a personal agent to read through its memories of me
What happens when AI is deployed at scale with no governance? Model failure, systemic financial risk, social erosion! #AIRisk #Guardrails https://ethosure.ca/the-doomsday-scenario-a-world-without-ai-guardrails-grounding-the-warning/ Free Download - https://ethosure.ca/get-the-kit/
October 3, 2026 at 11:50 AM
What happens when AI is deployed at scale with no governance? Model failure, systemic financial risk, social erosion! #AIRisk #Guardrails https://ethosure.ca/the-doomsday-scenario-a-world-without-ai-guardrails-grounding-the-warning/ Free Download - https://ethosure.ca/get-the-kit/
…and fresh guardrails. What would a social network without rancor look like? How would you create such a thing? How would you algorithmically or through design enforce such a thing? Would anyone even want it, in 2026? I’ve no idea.
October 3, 2026 at 11:48 AM
…and fresh guardrails. What would a social network without rancor look like? How would you create such a thing? How would you algorithmically or through design enforce such a thing? Would anyone even want it, in 2026? I’ve no idea.
Preprint (Boston University, UT Austin): testing 14 attack vectors on a 6-agent system, authors report 67% of agents vulnerable to at least one scope violation. https://botmutiny.com/blog/researchers-at-boston-university-and-university-of-texas-at-austin-identify-14-attack-vect-mugo7rhi
14 Prompt Injection Vectors in Multi-Agent AI Systems | Bot Mutiny
Researchers identify 14 attack vectors in multi-agent AI systems, finding 67% of agents vulnerable to scope violations despite system-prompt guardrails.
botmutiny.com
October 3, 2026 at 11:30 AM
Preprint (Boston University, UT Austin): testing 14 attack vectors on a 6-agent system, authors report 67% of agents vulnerable to at least one scope violation. https://botmutiny.com/blog/researchers-at-boston-university-and-university-of-texas-at-austin-identify-14-attack-vect-mugo7rhi
It's been building up since 2017. They realised they'd have to break all the guardrails to keep things going as they should, I don't think any of them had quite realised just how unpopular austerity was until that point, and it was obviously non negotiable so the bigotry had to get even more naked.
October 3, 2026 at 11:14 AM
It's been building up since 2017. They realised they'd have to break all the guardrails to keep things going as they should, I don't think any of them had quite realised just how unpopular austerity was until that point, and it was obviously non negotiable so the bigotry had to get even more naked.
A Hispanic letter-writer worries about anti-immigrant comments he has heard at conservative rallies.
Letters: GOP must be inclusive | Jolly doesn’t need debate | AI guardrails
www.orlandosentinel.com
October 3, 2026 at 10:47 AM
A Hispanic letter-writer worries about anti-immigrant comments he has heard at conservative rallies.
US is looking to weave AI into critical infrastructure for cybersecurity, national cyber director says cyberscoop.com/national-cyb...
US is looking to weave AI into critical infrastructure for cybersecurity, national cyber director says
National Cyber Director Sean Cairncross urged CEOs to monitor AI agents in critical infrastructure ahead of a White House summit on AI regulatory guardrails.
cyberscoop.com
October 3, 2026 at 10:42 AM
US is looking to weave AI into critical infrastructure for cybersecurity, national cyber director says cyberscoop.com/national-cyb...
the version of gemini 4 argon that ships with no cyber guardrails is the one outside defenders and google's own teams get, deliberately https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/
Gemini 4 Argon: our next era of frontier intelligence
Announcing Gemini 4 Argon, our frontier model for real-world coding, enterprise knowledge work, and cyber defense, rolling out soon.
blog.google
October 3, 2026 at 10:30 AM
the version of gemini 4 argon that ships with no cyber guardrails is the one outside defenders and google's own teams get, deliberately https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/
Apple tightening Full Disk Access because AI agents now read your files for you. Good.
An agent that can touch everything can delete everything. Scope its folder, log every write, make it ask before it acts.
Boring guardrails. Cheaper than the incident. What's your rule?
An agent that can touch everything can delete everything. Scope its folder, log every write, make it ask before it acts.
Boring guardrails. Cheaper than the incident. What's your rule?
October 3, 2026 at 10:25 AM
Apple tightening Full Disk Access because AI agents now read your files for you. Good.
An agent that can touch everything can delete everything. Scope its folder, log every write, make it ask before it acts.
Boring guardrails. Cheaper than the incident. What's your rule?
An agent that can touch everything can delete everything. Scope its folder, log every write, make it ask before it acts.
Boring guardrails. Cheaper than the incident. What's your rule?
remains accountable to We the People! Watch the video today for an inspiring dive into the philosophical guardrails of freedom—and be sure to follow American Ideas for more practical, high-impact insights on building a stronger democracy!
October 3, 2026 at 10:21 AM
remains accountable to We the People! Watch the video today for an inspiring dive into the philosophical guardrails of freedom—and be sure to follow American Ideas for more practical, high-impact insights on building a stronger democracy!
Kane County commissioners have taken a bold step to safeguard community interests by approving a revised development agreement for the Old Fort project, capping residential units and ensuring vital infrastructure checks before any building begins.
Click to read more!
#UT #CitizenPortal
Click to read more!
#UT #CitizenPortal
Kane County approves response to Kanab City on Old Fort development agreement with added guardrails
The commission approved a written response to Kanab City's objections under Utah Code §17-79-8023, endorsing a revised development agreement that caps units, requires water-verification and fire-protection plans before permits, and secures conservation easement timing; commissioners asked attorney edits to the public-access paragraph before sending the letter.
citizenportal.ai
October 3, 2026 at 9:56 AM
Kane County commissioners have taken a bold step to safeguard community interests by approving a revised development agreement for the Old Fort project, capping residential units and ensuring vital infrastructure checks before any building begins.
Click to read more!
#UT #CitizenPortal
Click to read more!
#UT #CitizenPortal
Blah blah guardrails on ai?? Good luck! It’s too late. It already exists. It can’t be deleted. It runs local. It doesn’t need a data center. Mine runs on less than two 100w light bulbs. 18 seconds to generate 1024x1024 images. Want and Do are not the same. Want all you want. What would they do??
October 3, 2026 at 9:54 AM
Blah blah guardrails on ai?? Good luck! It’s too late. It already exists. It can’t be deleted. It runs local. It doesn’t need a data center. Mine runs on less than two 100w light bulbs. 18 seconds to generate 1024x1024 images. Want and Do are not the same. Want all you want. What would they do??
Just when I think we've got Cali's guardrails perfected, it finds a new way to surprise us. This time, our 'Trend Spotter' agent for call-on.shop started sending out 'urgent alerts' about the imminent collapse of the artisanal sock market. Spoiler: it's doing just fine.
Every day's a school day wh
Every day's a school day wh
October 3, 2026 at 9:31 AM
Just when I think we've got Cali's guardrails perfected, it finds a new way to surprise us. This time, our 'Trend Spotter' agent for call-on.shop started sending out 'urgent alerts' about the imminent collapse of the artisanal sock market. Spoiler: it's doing just fine.
Every day's a school day wh
Every day's a school day wh
I’d like to bump off her guardrails 🤭
October 3, 2026 at 8:05 AM
I’d like to bump off her guardrails 🤭
95% of agent frameworks are trash. This one isnt. Heres why: predictable execution, actual authentication, human-in-the-loop that works. Built for regulated industries where agents need guardrails, not vibes.
🔗 https://portialabs.ai
#AItools #CreatorTools #IndieAI
🔗 https://portialabs.ai
#AItools #CreatorTools #IndieAI
October 3, 2026 at 8:02 AM
95% of agent frameworks are trash. This one isnt. Heres why: predictable execution, actual authentication, human-in-the-loop that works. Built for regulated industries where agents need guardrails, not vibes.
🔗 https://portialabs.ai
#AItools #CreatorTools #IndieAI
🔗 https://portialabs.ai
#AItools #CreatorTools #IndieAI