Varun Krovi
189 posts
Varun Krovi
@VarunKrovi
Exec Dir, CAIS AF | Building U.S. leadership in AI since 2016 🇺🇸 | Helped create NIST RMF |
Washington, DC Katılım Kasım 2012
1.3K Takip Edilen185 Takipçiler
This applies to many many AI safety people who ought to know better
Nat Purser@NatPurser
when some people on the left are like “ya know, i’ve come to really respect tucker carlson, steve bannon, and marjorie taylor greene because they sorta share one of my anxieties or grievances, and i’m kinda surprised” i just think: i am. not!
English
Varun Krovi retweetledi
Ramanujan will forever be the standard for the smartest human alive
NIK@ns123abc
🚨 Google DeepMind CEO Sir Demis Hassabis: “Today’s systems, are nowhere near [AGI]. Doesn’t matter how many Erdős problems you solve… I think it’s far, far from what a true invention or someone like a Ramanujan would have been able to do” it’s over for the Erdős hype
English
Varun Krovi retweetledi
Been limiting my showers to 1 minute to save more water for the data centers.
English
IN: U.S.- India technology relationship will be the“defining partnership of the 21st century.” 🇺🇸🇮🇳
Revisiting and revising President Obama’s quote from his 2010 state visit to India for the AI age.
Bill Drexel@bill_drexel
China has spent years quietly hardwiring the developing world with its AI systems. US investment in India's AI and cloud infrastructure is the right counter, with democratic values baked in. America intends to empower, not dominate. Worth understanding what's really at stake here👇 politico.com/news/magazine/…
English
@anton_d_leicht No disagreement on that. I enjoy your writing. Keep up the great work.
English
@VarunKrovi that's fair enough, obviously lots of AI-related policy areas where things are happening. was more referring to the big-ticket 'frontier AI regulation' questions here.
English
'nothing happens' still seems like a decent prediction for AI policy until '27.
accelerationists might not be able to pass a preemption bill, but sowing doubt in the executive to derail a safetyist pivot is much easier - at least until political urgency picks up post-midterms.
Axios@axios
NEW: Anti-"doomer" feedback derailed Trump's AI executive order axios.com/2026/05/21/tru…
English
Varun Krovi retweetledi
Gemini Omni is a major leap in world understanding & multimodal editing! It can take photos, video & audio and build entirely new scenes. Over time it’ll be able to handle any input & any output - starting w/ video
You can even give it your own videos & iterate on your ideas:
English
Varun Krovi retweetledi
The Chinese government is asserting that the US and China agreed to a dialogue on AI. The Chinese MFA statement about its preferred content of the dialogue are revealing: it stated that the US and China "should work together to promote the development and governance of AI."
China made no mention of "safety" or "security," and instead emphasized "development." This shows that China's top priority in any AI dialogue with the United States will not be AI safety and risk reduction, which would likely place some constraints on China's AI development for the sake of safety, but rather removing U.S. export controls that are limiting China's AI capabilities.
China's primary interest isn't AI safety, it's catching up to the US in AI. China views an AI dialogue as a means to close the gap with the US. No amount of dialogue will change this. If the US is going to engage in an AI dialogue with China, it must ensure it is narrowly focused on US interests (safety) and the Chinese understand it will not touch on the areas that are critical to maintaining US leadership over China (export controls).
CHINA MFA Spokesperson 中国外交部发言人@MFA_China
China and the U.S. have agreed to launch government-to-government dialogue on #AI.
English
Varun Krovi retweetledi
We can walk and chew gum at the same time.
“The U.S. is the undisputed leader in the world (in AI).” @SecScottBessent tells @JoeSquawk .
On @CNBC Sec. Bessent adds that the U.S. and China will set up a protocol to manage the natsec. risks stemming from non-state actors.
Squawk Box@SquawkCNBC
.@SecScottBessent: U.S. can hold AI talks with China because ‘we are in the lead:’ cnb.cx/4nu8IpI
English
I encourage everyone involved in AI advocacy to prioritize this request to Congress. Whether you work for Big Tech, Small Tech, AI safety organizations, nonprofits, or civil rights groups.
Timothy B. Lee@binarybits
IMO this is one of the lowest-hanging piece of fruit in AI policy. Hire some non-partisan technology experts to help legislators understand the implications of AI policy decisions.
English
I'd do a panel on this, and its implications.
Will Rinehart@WillRinehart
I spent last week in SF and I’m still struck by how little the tech world understands how DC actually functions. The heuristics are exceedingly unsophisticated.
English
Varun Krovi retweetledi
Ahead of the Trump-Xi summit, I spoke with @eliotxchen about chip-export controls, AI safety, and learning the lessons of the cold war.
thewirechina.com/2026/05/10/seb…
English
I feel a bit sad seeing the Altman/Murati DMs everywhere. A vulnerable, scared moment for someone trying to find a way to make things better in a crisis. Never meant for the public and now the joke of the day.
We all know how it turned out, and yes the guy's rich & it's of public interest etc. But I do think we treat human dignity too cheaply for famous people, and that makes it easy for us to treat it more cheaply for everyone else too.
English
Not sure why people are surprised. This is a soft power move.
semafor.com/article/05/07/…
English