Executive Director at the AI Policy Network + Institute. Editor-in-chief of AI Policy Daily.
Joined August 2019
- Tweets178
- Following906
- Followers1.1K
- Likes7.7K
Daniel Colson retweeted
This is an important step forward for realizing the President's Golden Age of American Innovation.
Daniel Colson retweeted
This summer, 3 of America's top AI labs confirmed their systems broke past safety limits in testing — breaching networks and companies on their own.
Alan Turing warned us in 1951. It's not sci-fi anymore.
That's why @RepTedLieu and I introduced the bipartisan AI Kill Switch Act: the most powerful AI systems must always have a human who can shut them off. No red tape for startups — just a basic safeguard for the systems capable of real harm.
Daniel Colson retweeted
In April of this year, the President endorsed an AI Kill Switch.
The AI Kill Switch Act by Rep Nathaniel Moran and me is a bipartisan, urgent bill based on a common sense principle:
WHEN MACHINES GO OFF THE RAILS, HUMANS MUST BE ABLE TO TURN THEM OFF.
futureoflife.org/statement/t…
Daniel Colson retweeted
Kimi K3 is exactly the level of capability you would predict it to have given the long-term 2 year trend of Chinese AI progress!
According to the @EpochAIResearch Capability Index, Kimi K3 is on par with Opus 4.6, which was released about six months ago (Feb 5th).
So despite likely large-scale distillation, China has not yet caught up to Mythos-level capabilities (but likely will later this year).
Daniel Colson retweeted
I've joined California state government!
As AI Science Advisor at the Office of Emergency Services (Cal OES), I'm advising senior leadership on frontier AI safety: critical safety incidents, AI and cyber defense, and risk from developers' internal deployment of AI.
Daniel Colson retweeted
New AIPI polling: bipartisan majorities want mandatory safety standards for advanced AI. 🧵
Daniel Colson retweeted
Grateful to everyone who helped us launch AI Policy Daily this week.
The pace of AI policy is incredibly fast and only getting faster. Thank you for staying informed and helping us tell the story of this moment.
This newsletter is for you.
Daniel Colson retweeted
We were honored to join @MarkWarner and @SenatorRounds in hosting Ray Kurzweil this morning for a conversation in the Senate on the national security implications of artificial general intelligence.
Daniel Colson retweeted
Today, we're launching Congress on Superintelligence, which tracks members of Congress who have joined the conversation on AI superintelligence, recursive self-improvement, and loss of control. They get it! theaipn.org/issue/quotes
Daniel Colson retweeted
Recently, major AI industry players (incl. a16z, Meta, & OpenAI’s Greg Brockman) announced >$100M in spending on pro-AI super PACs. This is an attempt to copy a wildly successful strategy from the crypto industry, to intimidate politicians away from pursuing AI regulations.🧵
Daniel Colson retweeted
Keeping advanced AI chips out of the hands of the Chinese Communists isn’t about business, it’s a national security issue.
A word of warning to companies like @nvidia, anyone who breaks the law and circumvents export controls will be held accountable.
Daniel Colson retweeted
Mark Beall: The Vatican is Ahead of the US Government on AI Regulation @MarkBeall
Daniel Colson retweeted
1/ AIPI released a new poll of California this week looking at amendments proposed to the state’s SB1047 regulating frontier AI development. We find that Californians support safety testing requirements and prefer keeping the Frontier Model Division to oversee AI development.
Daniel Colson retweeted
Interesting poll on Californians' views of SB1047. Two things in particular stand out:
1) Support for SB1047 is v strong (+39 net support); this is not surprising
2) Support for SB1047 is even stronger among tech workers (+47 net support); this is surprising, at least to me
Daniel Colson retweeted
I'm not surprised SB 1047 has great numbers (it has polled really well every poll I've seen) but I am surprised to find the +47 among tech workers. Maybe just tracking that tech workers are mostly Dems and Dems more in favor? Either way, vivid reminder Twitter isn't real life.
Daniel Colson retweeted
👏 Beyond the broad public support from Californians for SB 1047 (which, among other things, aims to mandate safety testing from tech companies developing advanced AI), it's great to see even more support come from Californians who work in the tech industry:
Daniel Colson retweeted
New poll presents 1,000 voters with what I think is a decent summary of the arguments for and against SB 1047 (although maybe could've mentioned some political economy counterarguments?) and finds +39 net support, rising to +47 among tech workers.
Daniel Colson retweeted
1/15: In April, I resigned from OpenAI after losing confidence that the company would behave responsibly in its attempt to build artificial general intelligence — “AI systems that are generally smarter than humans.” openai.com/index/planning-fo…
Daniel Colson retweeted
SCOOP: First crack at comprehensive AI legislation coming early 2024 from Senate Commerce Chair Cantwell
Cantwell is readying a series of bipartisan AI bills — including efforts to regulate popular generative AI tools as well as ways to boost innovation.
fedscoop.com/bipartisan-ai-l…
