Safe, Responsible Agentic AI for Humanity | AI Futurist, Ethicist and Philosopher | Computer Scientist | AI Architect | Author | Opinions 100% my own.
Singapore, Singapore
Joined June 2008
- Tweets83.6K
- Following1.4K
- Followers1.8K
- Likes99.7K
AI agents are neither stochastic parrots nor machine gods. @henryfarrell and Cosma Shalizi say they’re Homeric song. I agree, but the bard now has hands, and its audience has stopped heckling. Six enforceable controls for governing a song that acts:
wp.me/pgaPm1-9Xk
Dr. Luke Soon retweeted
AI safety researcher Dr. Roman Yampolskiy explains "fast takeoff," the AI scenario Sam Altman says is riskier:
10,000 AI agents. Each smarter than all of us. Working 24/7.
Governing the Ungovernable ; .. ⛓️💥lnkd.in/gGYSx3zG 🔗 lnkd.in/g3f29Y-g ⛓️💥lnkd.in/gGYSx3zG lnkd.in/gumUFP_n 🔗 lnkd.in/g3f29Y-g #AI
Governing Before the Crash:
🔗lnkd.in/gcsM3Q3z
#AIGovernance #AISafety
Much needed in a hyper #Trust deficit new world order - the arms race of our times #AI #AIGovernance tinyurl.com/34b6fnvv @MAS_sg @SingaporeMDDI @singapore_edb @minlawsg @GovTechSG @IMDAsg @AIVerify @AISecurityInst @aipaturkey
The AI Arms Race of Our Times.; and Singapore's role in a hyper #Trust deficit new world order.. tinyurl.com/57ekxux5
#AIWars #AISafety #AIGovernance #TrustedAI #TrustworthyAI #AIEthics @MAS_sg @SingaporeMDDI @GovTechSG @AISecurityInst @IMAS_SGP
🤖 Made with AI
The Copy, the Chokepoint, and the Third Answer . tinyurl.com/57ekxux5 #AI #AIGovernance #AISafety #Trust @MAS_sg @IMDAsg @AIVerify @GovTechSG @AISecurityInst @SingaporeMDDI #AI #AIGovernance #TrustedAI #Trust #AgenticGovernance
Distillation annoys the frontier labs. That alone is not a strategy.
Learning from outputs can’t be policed; fraud must be. Nobody can yet tell them apart.
Police the door, keep the ledger. And why Singapore can keep it. tinyurl.com/57ekxux5 #AI #AIGovernance #AISafety #Trust
Dr. Luke Soon retweeted
Mark Cuban on whether today's AI companies will follow Dell and Amazon through a bubble burst:
"There will be winners. I just don't know which company. Is it Anthropic? Is it OpenAI? I don't know."
Capability is compounding. Governance is still waiting in the pit lane 🚥.
The car 🏎️isn't coming in.
My three-part series on what our governance constructs miss:
lnkd.in/p/gTNupxD8
#AIGovernance #AgenticAI #AISafety #TrustedAI #FutureOfWork #TrustOS
Here’s Why an AI ‘Kill Switch’ May Not Be So Simple : tinyurl.com/3snraf2p #AISafety #SuperIntelligence
Governing the Agent AI Safety, the Global Governance Crisis, .. ⛓️💥lnkd.in/gGYSx3zG 🔗 lnkd.in/g3f29Y-g ⛓️💥lnkd.in/gGYSx3zG lnkd.in/gumUFP_n 🔗 lnkd.in/g3f29Y-g #AI
@realBigBrainAI
Part 1. MAS Finalises AI Risk Guidelines: What Actually Changed :
🔗tinyurl.com/854zevhx
Part 2. Six Regimes, One Agent: How Singapore's AIRG Compares with the EU, UK, US, Hong Kong and Australia
🔗lnkd.in/dVFzk4P9
@MAS_sg @AIVerify @IMDAsg @GovTechSG @AISecurityInst
Dr. Luke Soon retweeted
Part 1. MAS Finalises AI Risk Guidelines: What Actually Changed :
🔗tinyurl.com/854zevhx
Part 2. How Singapore's AIRG Compares with the EU, and US
🔗lnkd.in/dVFzk4P9
tinyurl.com/34b6fnvv
#AI #AIGovernance @MAS_sg @IMDAsg @AIVerify @GovTechSG @AISecurityInst
Special Relativity: Governing at the Speed of Agents: tinyurl.com/hmn3bcvc
#AIAgents #AgenticSafety #AIEthics #AIGovernance
Special Relativity: Governing AI at the Speed of Agents
tinyurl.com/5x84vsdf
#AI #AIGovernance #AISafety #AgenticGovernance
🤖 Made with AI
Dr. Luke Soon retweeted
AMD agreed to pay $8.2B for an AI that doesn't generate a single word.
World Labs CEO Fei-Fei Li explains what it generates instead:
The Arrow of Time: The AI Decisions You Can Never Undo tinyurl.com/mphwjuud #AI
Dr. Luke Soon retweeted
Sam Altman on the one AI policy where OpenAI and Anthropic still disagree:
Brendan Bordelon asked Altman to name a specific policy that Anthropic supports and OpenAI doesn't, at either the federal or state level.
First, he pointed out how much the two companies now agree on:
"You guys are now on the same page when it comes to pacing the frontier of China. Same page with third party evaluators. What's just one differentiator?"
Altman says the difference is which models should face oversight at all:
"The last I heard, one of the differentiators is we don't want any restrictions or testing on anything but the frontier models. We do not want to slow down people that make very powerful, but not the absolute frontier."
Both companies support testing and third-party evaluation. Where they split is scope.
In Altman's framing, OpenAI wants restrictions and testing applied only to frontier models, with nothing below them. That leaves out models he himself calls "very powerful."
He explains the reasoning:
"In the spirit of really focusing on the potential catastrophic risks…"
Then he names the cost directly:
"…even though I believe that if you don't put any testing or any oversight on models that are one or two generations behind the frontier, I accept that bad things will happen with those. I would not go advocate a licensing regime for those."
So OpenAI's position, as @sama states it, has three parts:
Oversight should concentrate on the potential catastrophic risks at the absolute frontier.
Models one or two generations behind should face no testing or oversight requirements.
That will lead to some harm, and he accepts that rather than advocate licensing for those models.
Altman knowingly makes this trade-off and says so plainly. He openly concedes that sub-frontier models will cause harm, and he judges that cost worth paying to keep regulatory attention on catastrophic risk at the top.