Joined August 2010
- Tweets 218
- Following 1,429
- Followers 433
- Likes 174
19 Photos and videos
Doug Calidas retweeted
Jun 13
This is why we need a concrete system in place to ensure the world's most powerful AI models are being vetted and cleared as safe before they're released.
We're in a Cold War with China on AI. We have to win, but we have to win the right way. Purely voluntary processes clearly aren’t going to do the trick.
nytimes.com/2026/06/12/techn…
18
8
43
8,864
Doug Calidas retweeted
Unions, kids-safety advocates, and civil liberties groups are raising the alarm about @RepLoriTrahan's proposal to preempt state AI development laws.
In the Boston Globe today, @bradrcarson explains why a 3 year freeze of state laws poses a major threat: bostonglobe.com/2026/06/12/o…
9
11
512
How Trump's AI strategy is taking shape axios.com/2026/06/12/trump-a…
4
2
15
12,212
‘Bad policy and even worse politics’: DeSantis spurns potential White House AI preemption dlvr.it/TT0D0q
2
7
19
11,452
Jun 12
Nancy Mace's post-loss X account is legitimately funny. Approaching @HunterBiden levels of satire.
Jun 11
I could still come in first if Mike Pence has the courage.
91
Doug Calidas retweeted
Jun 11
The biggest political mistake the AI industry could make is assuming public trust can be bought.
Appreciate @johnofa using Blue Rose Research's recent AI polling and the Center for Shared AI Prosperity's work here.
nytimes.com/2026/06/11/opini…
1
28
145
14,043
Jun 11
Another anti-endorsement from an industry front group. Keep em coming.
FOR THE BILL: a menagerie of industry front groups
AGAINST THE BILL: everyone else, especially the Democratic base (labor, civil rights, etc.)
The discussion draft is an important step in creating a framework for AI. The preemption provision in the draft is what we need to help get AI laws right.
1
5
353
Doug Calidas retweeted
Jun 11
Hmm - that's news to us, and should be news to @RepLoriTrahan as well. Perhaps she never read @frankgrubar in. @americans4ri has long made it clear that we are fine with narrow preemption of state laws that falls squarely within the scope of the elements of a federal framework and include language that will reliably prevent courts from extending the scope of preemption in ways that go beyond this.
In fact, @bradrcarson made it clear to Rep. Trahan that if she included this narrow preemption language - which we offered to help her prepare - ARI would not only refrain from criticizing the bill but would be happy to endorse it!
1
1
184
Jun 11
Great analysis by Brian here, one of the sharpest legal and policy thinkers on all things AI.
Jun 11
Unless you want to deregulate AI, there’s no reason to support AI preemption that is broader than the federal standard you're proposing AFAICT.
To unpack this, it's worth distinguishing different reasons for supporting preemption and what these imply about its scope.
1. National security: You think some AI regulation should be federal because advanced AI creates national security risks. I agree! But if this is your rationale, there’s no reason to preempt more broadly than your proposed federal standard - just port the state standard to the federal level and preempt the state requirements (if anything, the federal standard should be more protective than the state standard if nat sec is your concern).
2. “Patchwork:” You're worried about conflicting state standards. This is hypothetical right now, and I'd recommend waiting for this to be real before dealing with it because it may never materialize. But again, note that if this is why you want preemption, there's no reason to preempt more broadly than on the issues where there's a conflict (currently, none).
3. Deregulation: You think laws regulating AI are bad, and you want to get rid of them. If this is your rationale, then broad preemption makes perfect sense! The broader, the better. I can understand why industry/libertarians/etc. would want this.
Preemption supporters should ensure their justifications for supporting preemption line up with the scope of the preemption they are supporting.
1
2
164
Doug Calidas retweeted
Jun 11
Without clear guardrails for government data acquisition, AI will greatly increase the scope and scale of surveillance. Amodei’s policy paper is clear. Close the data broker loophole. darioamodei.com/post/policy-…
5
13
1,317
Doug Calidas retweeted
I am not going to amplify some of what is circulating, but I am seeing a lot of deepfakes of political candidates in political ads. @AmyKlobuchar leads a bipartisan bill to address this problem head on. And together we warned about the danger of inaction.
Congress must act.
1
6
16
1,934
Jun 11
On days like this, the (really good) air conditioning on the DC metro feels so good. We all love to knock Metro when it’s delayed, so I’ll also give credit where it’s due.
1
49
Doug Calidas retweeted
Jun 11
This is an incredibly important clarification from Sen. Blackburn spox. Clip and save.
1
1
174
Jun 10
Hear, hear
Jun 10
I’ve led the charge to stop Trump’s outrageous efforts to block state AI regulations before. I’m ready to do it again.
1
2
857
Doug Calidas retweeted
Preemption is a horrible idea. It would ban states from requiring a workforce study before rolling out AI in nursing. It would ban states from requiring that training data be demographically representative for medical or education usage. It would ban neutrality requirements for any part of the AI stack. It would ban states from testing requirements. It would ban states from antitrust regs requiring that ownership of parts of the stack be separated. It would ban states from extending liability to the foundation model for harms. It would stop states from prohibiting the use of stolen art and writing for training.
It's a nonstarter.
We cannot shut down states, where people have the most voice, and make them incapable of acting responsibly to protect their citizens.
Opposition to the Great American AI Act's three-year ban on state AI laws is growing.
ARI stands with the @AFLCIO, the @ACLU, and a broad coalition of public advocates: states shouldn't be sidelined on AI safeguards. More from @rollcall: rollcall.com/2026/06/10/rush…
8
28
2,252
Jun 10
Very much enjoyed the @nytimes feature on AI and the future of work featuring @deanwball , @daronac
and @emollick.
One dynamic that often gets missed is how much of a social equilibrium "human work" is. There does seem to be a social dynamic whereby those with excess money to spend converge on finding things for people to do to make a living (the @davidgraeber hypothesis). This isn't necessarily a great outcome - it's probably not - but it does feel like something we shouldn't ignore.
2
107
Jun 10
Precisely. I have yet to hear anyone make an argument for why protecting American children is a football that should be traded for another set of policies.
I remain mystified as to why some in Congress think preemption is a necessary starter to enacting AI or children’s online safety protections. States have acted with more protections and quicker than many of Congress’s proposals and preempting them would undo years of hard work.
1
5
330